J Biomater Sci Polymer Edn Vol 16 No 12 pp 1521ndash1535 (2005) VSP 2005Also available online - wwwvsppubcom

Improved phenotype of rat islets in a macrocapsuleby co-encapsulation with cross-linked Hb

YONG YEON KIM 1 SU YOUNG CHAE 1 SUNGWON KIM 1YOUNGRO BYUN 1 and YOU HAN BAE 2lowast1 Center for Biomaterials and Biotechnology Department of Materials Science and Engineering

Gwangju Institute of Science and Technology 1 Oryong-dong Puk-guGwangju 500-712 South Korea

2 Department of Pharmaceutics and Pharmaceutical Chemistry University of Utah421 Wakara Way Suite 315 Salt Lake City UT 84108 USA

Received 21 September 2004 accepted 14 May 2005

AbstractmdashA number of rat islets were co-encapsulated in a diffusion chamber-type device iemacrocapsule with a thermoreversible polymeric extracellular matrix (ECM) and bioactive ingredientof cross-linked hemoglobin (Hb-C) The ECM was formed from an aqueous solution of N-isopropyl-acrylamide co-polymers with a small amount of acrylic acid which exhibited unique solndashgel transitionin a temperature range of 30ndash34C without noticeable hysteresis The incorporation of Hb-C inthe islet macrocapsule showed a concentration-dependent effect on insulin secretion and viabilityof the entrapped islets Insulin secretion stimulation by glucose and cell viability were more thandoubled when compared with a control group (without Hb-C) at an optimum Hb-C concentrationof 025 mM due to its unique oxygen transporting capacity Furthermore 025 mM Hb-C in themacrocapsule was able to support islet density up to 1000 isletsdevice in a 154 microl total volumewithout negative effects on islet functionality and viability Hb-C incorporation is an effective strategyfor a macrocapsule-type biohybrid artificial pancreas for Type-I diabetes treatment which can befurther developed to a rechargeable system by employing the thermoreversible ECM and designinga proper macrocapsule

Key words Islets artificial pancreas macroencapsulation hemoglobin thermoreversible gel

INTRODUCTION

Transplantation of biohybrid artificial pancreas (BAP) physically immunoisolatedislets or insulinoma cells has been considered as a potential approach for the treat-ment of insulin-dependent (Type I) diabetes mellitus [1ndash4] Despite considerable

lowastTo whom correspondence should be addressed Tel (1-801) 585-1518 Fax (1-801) 585-3614E-mail youbaemccutahedu

1522 Y Y Kim et al

academic and industrial efforts in developing a BAP there still remain unsolvedand well-recognized obstacles such as short islet supply incomplete immunoiso-lation and biocompatibility of immunoprotecting membrane for its clinical appli-cations [5 6] Further hurdles include the large volume of the implant impairedinsulin secretion from encapsulated islets and the decrease in islet functions andviabilities over time after transplantation [5ndash7]

Deteriorated viability and functionality of encapsulated islets have been consid-ered as serious problems for their long-term clinical application [5 6] Due to theloss of blood supply by destruction of microvascular structure which penetratesinto islets and supply oxygen and nutrient in pancreas the isolated islets rely on theoxygen and nutrient supply largely by passive diffusion from surrounding Further-more hypoxic condition becomes worse when multiple islets are encapsulated in amacrocapsule This often causes central necrosis of the islets [8ndash11]

To overcome these problems several approaches have been attempted To enhanceinsulin secretion insulinotropic agents such as hypoglycemic drugs or peptidessuch as glucagon-like peptide-1 and their polymeric derivatives have been intro-duced in cultured andor encapsulated islets [12ndash14] In other attempts enhancedislet or cell viability and functionality have been achieved by increasing oxygensupply using hemoglobin (Hb) cross-linked Hb (Hb-C) perfluorocarbon emulsionand water electro-hydrolysis [15ndash18] From published animal and clinical results itis approximated that the functionality of isolated and encapsulated islets is at most15ndash20 of the islets in healthy pancreas in the body Therefore large transplan-tation volume resulting from large islet mass is required to normalize the bloodglucose levels of diabetic animals or patients Furthermore limited life span of thetransplanted islets restricts successful application of the BAP

To address such issues a rechargeable BAP system composed with immuno-protecting membrane thermoreversible polymeric extracellular matrix (ECM) in-sulinotropic agents and oxygen transporting molecules (Hb-C) have been proposedby our group [6 19ndash21] The rechargeable system employs a unique strategy toreplenish degenerated islets after graft failure with fresh ones under mild conditionswithout risky surgical operations [6 20] We have developed a novel temperature-sensitive N-isopropylacrylamide-based physical gel Owing to the unique inversethermosensitivity around physiological temperature poly(N-isopropylacrylamide)and related co-polymers have frequently been employed for biomedical fields suchas polymeric ECM drug delivery and tissue engineering [19ndash24] Aqueous solu-tions of high-molecular-weight poly(N-isopropylacrylamide-co-acylic acid (AAc))(5 mol AAc) at a concentration range of 5ndash10 wt showed sol-to-gel transitionby increasing temperature to slightly below physiological temperature (30ndash34C)without hysterisis and volume change [19] With the absence of a cytotoxic effecton the immobilized cells the co-polymer solution has been considered as an ECMfor a rechargeable BAP For facilitated oxygen transport Hb-C was introduced tomicrocapsules of islet or insulinoma cells Due to its unique oxygen transportingcapacity more than 5-fold enhanced insulin secretion and 4-fold enhanced cell via-

Enhanced function of islet macrocapsule 1523

bility of microencapsulated islets were observed after long-term static culture [18]Hb-C also showed great potential for in vivo animal experiment Implantation ofHb-C-containing islet microcapsules showed prolonged glucose normalization ofstreptozotocin-induced diabetic mice up to 8 weeks with normal glucose clearancekinetics whereas the normal islet microcapsule transplanted mice showed reoccur-rence of diabetes at 4 weeks after transplantation [25] Furthermore the incor-porated Hb-C successfully protected the encapsulated cells from oxidative stressesespecially nitric-oxide-induced cellular damage due to the unique nitric oxide scav-enging effect of Hb-C [26]

In this study we introduced polymeric ECM and Hb-C into an islet macrocapsulesystem which was enveloped by permeability-selective membranes Optimum Hb-Cconcentration and islet density for effective insulin secretion and viability weredetermined

MATERIALS AND METHODS

Materials

Poly(ethylene glycol) (PEG 2 kDa) 2-hydroxyethyl methacrylate (HEMA) methylmethacrylate (MMA) and methacrylic acid (MAA) were obtained from Aldrich(Milwaukee WI USA) Collagenase (Type V) Hanksrsquo balanced salt solution(HBSS) Ficoll 400DL (400 kDa) D-glucose RPMI-1640 acridine orange (AO)and propidium iodide (PI) were purchased from Sigma (St Louis MO USA) Fe-tal bovine serum (FBS) was obtained from Biofluid (Rockville MD USA) Hy-drophilic Duroporereg (pore size 010 microm) was obtained from Millipore (BedfordMA USA) N-isopropylacrylamide (NiPAAm) tetramethylene diamine (TEMED)ammonium persulfate (APS) and 22prime-azobis-isobutyronitrile (AIBN) were pur-chased from Acros NiPAAm was recrystallized in n-hexane and AIBN was recrys-tallized in methanol Acrylic acid (AAc) obtained from Acros (Fairlawn NJ USA)was purified by vacuum distillation (40C10 mmHg) Outdated human red bloodcells were obtained from the Korean Red Cross CellTiter 96 AQueous one solutioncell proliferation assay kit (containing 3-[45-dimethylthiazol-2-yl]-5-[3-carboyl-methoxyphenyl]-2-[4-sulfophenyl]-2H-tetrazolium inner salt (MTS) reagent) waspurchased from Promega (Madison WI USA) Rat insulin radioimmunoassay kitswere purchased from ICN (Costa Mesa CA USA)

Preparation of cross-linked hemoglobin (Hb-C)

Fresh oxy-Hb was isolated from outdated human red blood cells (RBC) as previ-ously reported [18] Briefly after washing twice with isotonic saline RBC lysiswas conducted by organic solvent treatment (dichloromethane) Cell debris was re-moved by centrifugation and filtration The purified and concentrated Hb solutionwas obtained by ultrafiltration (50 times 103 molecular weight cut-off (MWCO) mem-

1524 Y Y Kim et al

brane) The Hb solution was sterilized by filtration though a 022 microm pore sizedsyringe filter and stored in a refrigerator at 4C

Hb-C used in this study was prepared by the same method reported in our previouspublication [18] Hb cross-linking was conducted by a reaction between residualamino groups of Hb with activated PEG diacid (PEG N-hydroxysuccinimide ester(NHS-PEG) 2 kDa) In brief predetermined amounts of PEG and Hb (PEGHbmol ratio 10 1) were added to a one-neck round-bottom flask and stirred at 4Cfor 4 h After the cross-linking reaction Hb-C was purified from the mixture byultrafiltration with a 100times103 MWCO membrane to remove monomeric Hb and by-products The sterilized Hb-C solution was obtained by filtration through a 022 micrompore size syringe filter and stored at 4C The resulting Hb-C was characterized bySDSndashPAGE

Preparation of extracelluar matrix (ECM) and semi-permeable membrane

As a polymeric ECM the co-polymers of NiPAAm with AAc (98 2 feed mol ratio)were synthesized by free radical polymerization in benzene (10 wt monomerconcentration) with AIBN as an initiator (7 times 10minus3 mol AIBNmol monomer) asreported previously [19]

The semi-permeable immunoprotecting membrane was fabricated by in situpolymerization with spin-coating of poly(HEMA-co-MMA-co-MAA) (69 30 molratio in feed of HEMA to MMA and 1 wt of MAA of HEMA and MMA) hydrogelon Duraporereg membranes (20 mm times 20 mm) Briefly before loading monomersolutions (HEMA MMA MAA and cross-linking agent of 15-hexadiene-34-diol)and initiator (10 APS and TEMED) were mixed and then after the appropriateviscosity was obtained the mixture was loaded (6 min after adding the initiator)After loading the first spin-coating and gel formation occurred simultaneously Inthe second coating loading time and loading amount were reduced compared to thefirst coating because the purpose of double coating is to improve surface smoothnessand not to increase thickness of the coating layer [27]

Islets isolation and encapsulation

Islets of Langerhans were isolated from pancreata of male SpraguendashDawley rats(200ndash250 g) by collagenase digestion and discontinuous Ficoll density gradientcentrifugation [18] The isolated islets were incubated in RPMI-1640 mediasupplemented with 10 heat-inactivated FBS for 1 day at 37C in a 95 air5CO2 atmosphere Highly purified (more than 98) islets were obtained byhandpicking The handpicked islets were cultured for 24 h prior to encapsulation

The isolated islets were macroencapsulated in the planar diffusion chamber withtwo different loading conditions Firstly 1000 islets were macroencapsulatedwith ECM solutions containing varying amounts of Hb-C (0 0125 025 and05 mM based on Hb in HBSS) without changing the number of islets fora macrocapsule device Secondly different numbers of islets (500 1000 and

Enhanced function of islet macrocapsule 1525

Figure 1 Schematic diagram of islet macroencapsulation device A typical device is composedof three Teflon rings (top middle and bottom) and two immunoisolating membranes (upper andlower) For macroencapsulation the lower membrane was fixed with the bottom and middle rings theisletECM mixture was poured into the central space and then the device was sealed with the uppermembrane and the top ring

2000 islets per device) were encapsulated using 025 mM Hb-C containing ECMsolution The macrocapsule devices (volume 154 microl) were composed of threecomponents supports made of three polytetrafluoroethylene (Teflon) rings twosemi-permeable membranes that formed the walls of the encapsulation chamberand a thermoreversible polymeric ECM that immobilized the islets at physiologicaltemperature A schematic illustration of the macrocapsule device and its dimensionsare shown in Fig 1 For encapsulation the bottom ring lower immunoisolatingmembrane and middle ring were assembled The islets were mixed with thepolymeric ECM solution (7 wt concentration in HBSS with and without Hb-C)at 4C and placed on the device which was closed with a upper membrane and topring Then each device was placed in a 6-well plate with 6 ml culture medium andincubated in humidified atmosphere (5 CO237C) Culture media (RPMI-1640with 10 FBS) were changed every other day

Static culture of encapsulated islets

To determine an optimum Hb-C concentration various amounts of Hb-C (00125 025 and 05 mM Hb equivalents) were applied to evaluate concentration-dependent Hb-C effects on macroencapsulated islets within polymeric ECM Toinvestigate basal and glucose stimulated insulin secretion of the encapsulatedislets the islet macrocapsules were incubated at different glucose concentrationsof 100 mgdl and 300 mgdl for 4 h Samples (250 microl) were taken from the plateand insulin concentration was measured by insulin radioimmunoassay (RIA) Thestatic glucose stimulation tests were performed on day 3 (1st week) and week 4

1526 Y Y Kim et al

after encapsulation To investigate cell viability the islets were retrieved from themacrocapsules on day 3 (1st week) and week 4 after encapsulation and the cellviability was measured using MTS reagent

After optimization of Hb-C concentration (025 mM Hb equivalent) various num-bers of islets (2000 1000 and 500 isletsdevice) were encapsulated in macroencap-sulation devices with polymeric ECM (with and without Hb-C) and statically cul-tured for 4 weeks with glucose-stimulated insulin secretion tests and viability assaysby MTS assays and AOPI staining

Cell viability tests

MTS assay The viability of islets retrieved from macrocapsules was measuredby MTS assay Briefly the 30 retrieved islets were placed into a 96-well plate with20 microl of assay solution and 100 microl cell culture medium and incubated at 37C for4 h After incubation the absorbance at 490 nm was measured by using a platereader The obtained optical density values were normalized by subtraction of theblank reading (without islets)

AOPI staining After static culture the retrieved islets were stained with AOPIto investigate cell viability [28] Briefly the retrieved islets were placed into 96wells and stained with AOPI solution (100 microl AOPI solution 067 microgml AO and75 microgml PI in HBSS) for 10 min in a dark place After removing the remainingAOPI by washing with PBS the islets were observed with fluorescence microscopeThrough the staining viable cells are stained bright green by the incorporation ofAO in nucleic acids Due to the impermeability for PI of a viable cell membraneonly dead cells were stained red by intercalation of PI with DNA andor RNA

RESULTS

Preparation of Hb-C

Purified Hb and Hb-C showed a typical oxy-hemoglobin UV-Vis spectrum withnegligible met-hemoglobin content as previously reported [18] The high oxy-hemoglobin content in Hb-C is desirable for facilitated oxygen transport TheSDSndashPAGE chromatogram clearly revealed that the isolated Hb solution containeda minimal amount of impurities The chromatogram also qualitatively showedthe separation of high-molecular-mass Hb-C by ultrafiltration Compared tothe reaction mixture and ultrafiltration filtrate the final Hb-C displayed delayedretardation through the polyacrylamide gel owing to the high cross-linking densityThe results of SDSndashPAGE are presented in Fig 2

Preparation of ECM and semi-permeable membrane

The polymeric ECM of poly(NIPAAm-co-AAc) had a high molecular mass (morethan 1000 kDa) and an aqueous solution of the polymer showed novel concentration

Enhanced function of islet macrocapsule 1527

Figure 2 Characterization of PEG-Hb conjugate using SDSndashPAGE (12 gel) Lane 1 standardmarkers lane 2 isolated Hb lane 3 after conjugation lane 4 purified Hb-C lane 5 ultrafiltrationfiltrate

and temperature-dependent solndashgel transition phenomena without hysterisis asreported by Han et al [19] The solndashgel transition temperature of the 7 aqueoussolution was 33C After double spin-coating a smooth dense and permeability-selective thin poly(HEMA-MMA-MAA) hydrogel top layer was anchored on thetop layer of a porous Duroporereg membrane SEM results showed that the thicknessof the coating layer was less than 5 microm and 30ndash60 of the coating layer anchoredthrough the Duroporereg membrane (data not shown) Serum and protein mixturepermeability test showed that the MWCO of the spin-coated membrane was about70 times 103 and IgG and other immunoglobulin were completely prevented frompenetrating through the membrane [27]

Optimization of Hb-C concentration for macroencapsulation

Hb-C-concentration-dependent glucose stimulation in insulin secretion and cellviability were observed with macroencapsulated rat islets (1000 islets per pouch154 microl loading volume) As demonstrated in Fig 3A and 3B the basal andglucose-stimulated insulin secretion functions of macroencapsulated islets werepreserved by the incorporation of Hb-C The insulin-secreting functions wereincreased with increasing Hb-C concentration up to 025 mM and then slightlydecreased with a higher Hb-C concentration (05 mM Hb-C) Compared to control(macroencapsulated islets without Hb-C) 213 and 303 enhanced insulin

1528 Y Y Kim et al

Figure 3 Static insulin secretion of macroencapsulated islets at basal and stimulated glucoseconcentration (A) Basal glucose at 100 mgdl (B) stimulated glucose at 300 mgdl in RPMI-1640medium Macrocapsules (1000 islets per pouch) were incubated with 6 ml RPMI-1640 medium for4 h The results are expressed as mean plusmn SD (n = 5) Statistical differences between 1st week and4th week data were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

Enhanced function of islet macrocapsule 1529

secretion was observed in macrocapsules containing 025 mM Hb-C for basal(100 mgdl) and high glucose (300 mgdl) stimulation respectively on day 3after encapsulation (hereafter referred to as 1st week) The values increased to704 (100 mgdl glucose) and 994 (300 mgdl glucose) at week 4 after staticincubation Compared to the 1st week results the basal and stimulated insulinsecretion of the control group dropped to 61 and 51 after 4 weeks culturerespectively In case of the 025 mM Hb-C containing group these values were 86and 78 for basal and glucose-stimulated conditions respectively The stimulationindex (ratio of glucose-stimulated to basal insulin secretion) also showed a similartrend Enhanced insulin secretion (154 times) by glucose stimulation in the controlgroup was decreased to 128 times after 4 weeks culture On the other hand thestimulation indexes of 025 mM Hb-C containing group were 154 and 150 for 1stand 4th week culture respectively

The cell viability data obtained by MTS assay also showed a similar tendencyfor the incorporated Hb-C as shown in Fig 4 The group containing 025 mM Hb-Cshowed the best cell viability after 1 and 4 weeks culture Compared to the 1st weekvalue 825 relative viability was obtained in 025 mM Hb-C group after 4 weeksculture whereas only 376 cells maintained were viable after 4 weeks withoutHb-C

Figure 4 Effect of Hb-C concentration on islet viability Islets (1000 islets per pouch) wereencapsulated with different concentrations of Hb-C and the viability was measured by MTS assayafter different time (3 days and 4 weeks after encapsulation) Data are mean plusmn SD (n = 5) Statisticaldifferences between 1st week and 4th week data were calculated using unpaired Studentrsquos t-tests(lowastP lt 005 lowastlowastP lt 001)

1530 Y Y Kim et al

Optimization of islet density for macroencapsulation

The number of islets per macrocapsule device also strongly affected the insulin-secreting function and viability of islets Basal (100 mgdl) and glucose-stimulated(300 mgdl glucose) insulin secretion functions of the control group (without Hb-C)and Hb-C group with different islet density are shown in Fig 5A and 5B Thebasal and glucose insulin secretion tests were performed on day 3 and week 4after encapsulation The tests performed on day 3 after encapsulation did not showa significant dependence on number of islets encapsulated Therefore the dataobtained with 1000 isletsdevice were employed as control

In the case of the experimental control group (without Hb-C) a dramaticdecrease of basal and glucose-stimulated insulin secretion was observed with 2000isletsdevice group (Fig 5A) Compared with control (1st week data) only 81and 150 insulin was secreted by basal and high glucose stimulation respectivelyWhen decreasing islet density to 1000 and 500 isletsdevice the insulin-secretionfunctions were enhanced The 1000 isletsdevice group showed 635 and 630of relative insulin secretion by basal and high glucose stimulation respectivelyAlthough slightly decreased insulin secretion was observed with 500 isletsdevicegroup compared to the 1000 isletsdevice group they did not show statisticaldifferences between the two groups

Introduction of the Hb-C into the macrocapsule device significantly enhancedthe insulin-secretion function measured at 4 weeks after static culture The threeHb-C groups showed a level of insulin secretion superior to the experimentalcontrol groups (without Hb-C) In case of the 2000 isletsdevice group 650and 730 relative (in comparison with 1st week control) insulin secretion wasobserved by basal and high glucose stimulation respectively The situations wereenhanced by decreasing the islet density to 500 and 1000 isletsdevice More than80 of initial insulin secretion was measured in the 1000 and 500 isletsdevicegroup independent of islet density and glucose concentration Reduced cellviability with rising islet density and enhanced cell viability by incorporation ofHb-C into islet macrocapsules data were obtained from cell viability tests usingMTS reagents (Fig 6) In the case of Hb-C-containing groups the relative cellviabilities (compared to week 1 data of 1000 isletspouch) were maintained at morethan 80 in the 1000 and 500 isletsdevice groups after 4 weeks static cultureSlightly reduced cell viability (relative viability of 655) was observed in the2000 isletsdevice group In the case of control groups (without Hb-C) remarkabledifferences in cell viability were observed The relative cell viabilities of around50 for the 1000 and 500 isletsdevice groups dropped to 225 by increasing theislet density to 2000 isletsdevice

Cell viability staining using AOPI showed similar trends As shown in Fig 7 theretrieved islets from the Hb-C-containing groups showed a green colored viablecell state A negligible difference of cell staining was observed between 1000and 500 isletsdevice Increasing the islet density to 2000 isletsdevice showed anegative effect on cell viability because of a slight development of central necrosis

Enhanced function of islet macrocapsule 1531

Figure 5 Static insulin secretion of macroencapsulated islets cultured in medium with basal andhigh glucose levels (A) Without Hb-C (B) with 025 mM Hb-C Macrocapsules with different isletdensity (2000 1000 and 500 isletsdevice) were incubated with 6 ml RPMI-1640 medium for 4 h andsecreted insulin contents were measured by insulin radioimmunoassay The unit of G is mgdl Thedata are expressed as mean plusmn SD (n = 4) Statistical differences between 1st week and 4th week datawere calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

1532 Y Y Kim et al

Figure 6 Cell viability of macroencapsulated islets with different islet density after 4 weeks staticculture Data are shown as meanplusmnSD (n = 4) Statistical differences between 1st week and 4th weekdata were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

as shown in Fig 7 However the results for the control groups (without Hb-C)showed a remarkable difference between the 2000 and 1000 isletsdevice groupsThe islets from the 2000 isletsdevice group showed a typical red colored deadcell staining with destruction of integrated islet structure due to severe cell necrosisandor apoptosis Compared to the 2000 isletsdevice group slightly enhanced cellviability but with large necrotic core and viable rim of islets was observed in the1000 isletsdevice group as shown in Fig 7

DISCUSSION

In this study we investigated the optimum conditions of Hb-C concentration andislet density for islet macrocapsules Islets were immobilized by a thermore-versible polymeric ECM Among several devices proposed for islet immunoisola-tion macrocapsule devices possess advantages such as ease of implantation and ex-plantation with minimal surgical risk variety of implantation site and reduced trans-plantation volume by increasing islet density The thermosensitive polymeric ECMwhich showed a unique sol-to-gel transition may allow one to replace degeneratedislets after graft failure in the macroencapsulation devices by fresh islets by slightlydecreasing the temperature without surgical operations [6 19ndash21] Moreover the

Enhanced function of islet macrocapsule 1533

Figure 7 Fluorescence images of islets in the macrocapsules after 4 weeks of culture (A) 2000isletspouch with 025 mM Hb-C (B) 2000 isletspouch without Hb-C (C) 1000 isletspouch with025 mM Hb-C and (D) 1000 isletspouch without Hb-C

introduction of a bioactive ingredient (Hb-C) may enhance the cell functionality andviability as shown in our previous islet microencapsulation system [18 25]

As shown in Figs 3 and 4 insulin-secreting functions and cell viability weregreatly affected by the incorporated Hb-C in a concentration-dependent fashionThe optimum concentration of Hb-C as obtained from the experiments was025 mM (based on only Hb) The enhancing effect of the incorporated Hb-C oncell functionality and viability could be explained by the oxygen supply and otherreported bioactivities such as radical scavenging buffering and cell-stabilizingeffect [15 18 25] Most importantly increased oxygen supply may play acritical role in the enhanced cell functionality and viability Many studies havefocused on the limited oxygen supply on isolated and encapsulated islets Insulinsecretion especially glucose-stimulated insulin secretion was greatly affected byoxygen partial pressure [9] and the malfunction of islets or insulinoma cellswas overcome by increasing oxygen supply using hemoglobin Hb-C and waterelectrohydrolysis [15 16 18] However increasing the Hb-C concentration overthe optimum concentration resulted in decreased functionality and viability ofencapsulated islets as shown in Figs 3 and 4 The phenomena may be attributed todeteriorated mass transport by increased viscosity andor density of the polymericECM Although the polymeric ECM itself did not act as a diffusional barrier forsmall molecules such as insulin nutrients and wastes [29] a high concentrationof Hb-C in polymeric ECM may act as a mass transport barrier which resulted inreduced cell functionality and viability

Size or volume issue of a BAP is another important factor for its successfulapplication In the case of islet macroencapsulation the volume issue could be

1534 Y Y Kim et al

overcome by increasing islet density However cell viability is significantly affectedby cell density Figures 6 and 7 clearly show the decreased cell viability byincreasing islet density The enhanced cell viability by incorporation of Hb-Cwas caused by Hb-C groups however a high islet density (2000 isletsdevice)reduced viability by developing central necrosis By increasing islet density strongcompetition between encapsulated islets for nutrients and oxygen and seriousaccumulation of waste may occur in the device which results in reduced cellviability by increasing cell necrosis andor apoptosis On the other hand decreasingthe islet density increases the device volume which may evoke the volume issue

CONCLUSIONS

In this report we investigated the effect of Hb-C on the functionality and viabilityof an islet macroencapsulation device with a thermally reversible polymeric ECMBy introduction of 025 mM Hb-C into the islet macrocapsule more than 100enhanced insulin secretion and cell viability were observed after 4 weeks cultureFurthermore increase of islet density up to 1000 isletsdevice (154 microl volume)without any negative effect on insulin secretion and viability was possible byintroduction of 025 mM Hb into islet macrocapsules Although the practical in vivostudy remains as further study the incorporation of Hb-C into islet macrocapsulemight prolong the life span of encapsulated islets and reduce the implant volume byincreasing the functionality of islets and islet density

Acknowledgements

This work was supported by NIH DK56884 and BK 21 program (South Korea)

REFERENCES

1 G Reach Diabet Med 10 105 (1993)2 R P Lanza and W L Chick Ann N Y Acad Sci 831 323 (1997)3 R Calafiore Diabet Metab Rev 14 315 (1998)4 M K Lee and Y H Bae Adv Drug Deliv Rev 42 103 (2000)5 C K Colton and E S Avgoustiniators J Biomech Eng 113 152 (1992)6 Q P Hou and Y H Bae Adv Drug Deliv Rev 35 271 (1999)7 D Brandhorst H Brandhorst A Zwolinski F Nahidi and R G Bretzel Transplantation 71

179 (2001)8 J C Hutton and W J Malaisse Diabetologia 18 395 (1980)9 K E Dionne C K Colton and M L Yarmush Diabetes 42 12 (1993)

10 J Schrezenmeir L Gero M Solhdju J Kirchgessner C Laue J Beyer H Stier and W Muller-Klieser Transplant Proc 26 809 (1994)

11 M Ohta D Nelson M D Meglasson and M Erecinska J Biol Chem 265 17525 (1990)12 J S Whang S Y Chae M K Lee and Y H Bae Biomaterials 19 1198 (1998)13 K H Park S W Kim and Y H Bae J Biomed Mater Res 55 72 (2001)14 H Gappa M Baudys J J Koh S W Kim and Y H Bae Tissue Eng 7 35 (2001)

Enhanced function of islet macrocapsule 1523

bility of microencapsulated islets were observed after long-term static culture [18]Hb-C also showed great potential for in vivo animal experiment Implantation ofHb-C-containing islet microcapsules showed prolonged glucose normalization ofstreptozotocin-induced diabetic mice up to 8 weeks with normal glucose clearancekinetics whereas the normal islet microcapsule transplanted mice showed reoccur-rence of diabetes at 4 weeks after transplantation [25] Furthermore the incor-porated Hb-C successfully protected the encapsulated cells from oxidative stressesespecially nitric-oxide-induced cellular damage due to the unique nitric oxide scav-enging effect of Hb-C [26]

In this study we introduced polymeric ECM and Hb-C into an islet macrocapsulesystem which was enveloped by permeability-selective membranes Optimum Hb-Cconcentration and islet density for effective insulin secretion and viability weredetermined

MATERIALS AND METHODS

Materials

Poly(ethylene glycol) (PEG 2 kDa) 2-hydroxyethyl methacrylate (HEMA) methylmethacrylate (MMA) and methacrylic acid (MAA) were obtained from Aldrich(Milwaukee WI USA) Collagenase (Type V) Hanksrsquo balanced salt solution(HBSS) Ficoll 400DL (400 kDa) D-glucose RPMI-1640 acridine orange (AO)and propidium iodide (PI) were purchased from Sigma (St Louis MO USA) Fe-tal bovine serum (FBS) was obtained from Biofluid (Rockville MD USA) Hy-drophilic Duroporereg (pore size 010 microm) was obtained from Millipore (BedfordMA USA) N-isopropylacrylamide (NiPAAm) tetramethylene diamine (TEMED)ammonium persulfate (APS) and 22prime-azobis-isobutyronitrile (AIBN) were pur-chased from Acros NiPAAm was recrystallized in n-hexane and AIBN was recrys-tallized in methanol Acrylic acid (AAc) obtained from Acros (Fairlawn NJ USA)was purified by vacuum distillation (40C10 mmHg) Outdated human red bloodcells were obtained from the Korean Red Cross CellTiter 96 AQueous one solutioncell proliferation assay kit (containing 3-[45-dimethylthiazol-2-yl]-5-[3-carboyl-methoxyphenyl]-2-[4-sulfophenyl]-2H-tetrazolium inner salt (MTS) reagent) waspurchased from Promega (Madison WI USA) Rat insulin radioimmunoassay kitswere purchased from ICN (Costa Mesa CA USA)

Preparation of cross-linked hemoglobin (Hb-C)

Fresh oxy-Hb was isolated from outdated human red blood cells (RBC) as previ-ously reported [18] Briefly after washing twice with isotonic saline RBC lysiswas conducted by organic solvent treatment (dichloromethane) Cell debris was re-moved by centrifugation and filtration The purified and concentrated Hb solutionwas obtained by ultrafiltration (50 times 103 molecular weight cut-off (MWCO) mem-

1524 Y Y Kim et al

brane) The Hb solution was sterilized by filtration though a 022 microm pore sizedsyringe filter and stored in a refrigerator at 4C

Hb-C used in this study was prepared by the same method reported in our previouspublication [18] Hb cross-linking was conducted by a reaction between residualamino groups of Hb with activated PEG diacid (PEG N-hydroxysuccinimide ester(NHS-PEG) 2 kDa) In brief predetermined amounts of PEG and Hb (PEGHbmol ratio 10 1) were added to a one-neck round-bottom flask and stirred at 4Cfor 4 h After the cross-linking reaction Hb-C was purified from the mixture byultrafiltration with a 100times103 MWCO membrane to remove monomeric Hb and by-products The sterilized Hb-C solution was obtained by filtration through a 022 micrompore size syringe filter and stored at 4C The resulting Hb-C was characterized bySDSndashPAGE

Preparation of extracelluar matrix (ECM) and semi-permeable membrane

As a polymeric ECM the co-polymers of NiPAAm with AAc (98 2 feed mol ratio)were synthesized by free radical polymerization in benzene (10 wt monomerconcentration) with AIBN as an initiator (7 times 10minus3 mol AIBNmol monomer) asreported previously [19]

The semi-permeable immunoprotecting membrane was fabricated by in situpolymerization with spin-coating of poly(HEMA-co-MMA-co-MAA) (69 30 molratio in feed of HEMA to MMA and 1 wt of MAA of HEMA and MMA) hydrogelon Duraporereg membranes (20 mm times 20 mm) Briefly before loading monomersolutions (HEMA MMA MAA and cross-linking agent of 15-hexadiene-34-diol)and initiator (10 APS and TEMED) were mixed and then after the appropriateviscosity was obtained the mixture was loaded (6 min after adding the initiator)After loading the first spin-coating and gel formation occurred simultaneously Inthe second coating loading time and loading amount were reduced compared to thefirst coating because the purpose of double coating is to improve surface smoothnessand not to increase thickness of the coating layer [27]

Islets isolation and encapsulation

Islets of Langerhans were isolated from pancreata of male SpraguendashDawley rats(200ndash250 g) by collagenase digestion and discontinuous Ficoll density gradientcentrifugation [18] The isolated islets were incubated in RPMI-1640 mediasupplemented with 10 heat-inactivated FBS for 1 day at 37C in a 95 air5CO2 atmosphere Highly purified (more than 98) islets were obtained byhandpicking The handpicked islets were cultured for 24 h prior to encapsulation

The isolated islets were macroencapsulated in the planar diffusion chamber withtwo different loading conditions Firstly 1000 islets were macroencapsulatedwith ECM solutions containing varying amounts of Hb-C (0 0125 025 and05 mM based on Hb in HBSS) without changing the number of islets fora macrocapsule device Secondly different numbers of islets (500 1000 and

Enhanced function of islet macrocapsule 1525

Figure 1 Schematic diagram of islet macroencapsulation device A typical device is composedof three Teflon rings (top middle and bottom) and two immunoisolating membranes (upper andlower) For macroencapsulation the lower membrane was fixed with the bottom and middle rings theisletECM mixture was poured into the central space and then the device was sealed with the uppermembrane and the top ring

2000 islets per device) were encapsulated using 025 mM Hb-C containing ECMsolution The macrocapsule devices (volume 154 microl) were composed of threecomponents supports made of three polytetrafluoroethylene (Teflon) rings twosemi-permeable membranes that formed the walls of the encapsulation chamberand a thermoreversible polymeric ECM that immobilized the islets at physiologicaltemperature A schematic illustration of the macrocapsule device and its dimensionsare shown in Fig 1 For encapsulation the bottom ring lower immunoisolatingmembrane and middle ring were assembled The islets were mixed with thepolymeric ECM solution (7 wt concentration in HBSS with and without Hb-C)at 4C and placed on the device which was closed with a upper membrane and topring Then each device was placed in a 6-well plate with 6 ml culture medium andincubated in humidified atmosphere (5 CO237C) Culture media (RPMI-1640with 10 FBS) were changed every other day

Static culture of encapsulated islets

To determine an optimum Hb-C concentration various amounts of Hb-C (00125 025 and 05 mM Hb equivalents) were applied to evaluate concentration-dependent Hb-C effects on macroencapsulated islets within polymeric ECM Toinvestigate basal and glucose stimulated insulin secretion of the encapsulatedislets the islet macrocapsules were incubated at different glucose concentrationsof 100 mgdl and 300 mgdl for 4 h Samples (250 microl) were taken from the plateand insulin concentration was measured by insulin radioimmunoassay (RIA) Thestatic glucose stimulation tests were performed on day 3 (1st week) and week 4

1526 Y Y Kim et al

after encapsulation To investigate cell viability the islets were retrieved from themacrocapsules on day 3 (1st week) and week 4 after encapsulation and the cellviability was measured using MTS reagent

After optimization of Hb-C concentration (025 mM Hb equivalent) various num-bers of islets (2000 1000 and 500 isletsdevice) were encapsulated in macroencap-sulation devices with polymeric ECM (with and without Hb-C) and statically cul-tured for 4 weeks with glucose-stimulated insulin secretion tests and viability assaysby MTS assays and AOPI staining

Cell viability tests

MTS assay The viability of islets retrieved from macrocapsules was measuredby MTS assay Briefly the 30 retrieved islets were placed into a 96-well plate with20 microl of assay solution and 100 microl cell culture medium and incubated at 37C for4 h After incubation the absorbance at 490 nm was measured by using a platereader The obtained optical density values were normalized by subtraction of theblank reading (without islets)

AOPI staining After static culture the retrieved islets were stained with AOPIto investigate cell viability [28] Briefly the retrieved islets were placed into 96wells and stained with AOPI solution (100 microl AOPI solution 067 microgml AO and75 microgml PI in HBSS) for 10 min in a dark place After removing the remainingAOPI by washing with PBS the islets were observed with fluorescence microscopeThrough the staining viable cells are stained bright green by the incorporation ofAO in nucleic acids Due to the impermeability for PI of a viable cell membraneonly dead cells were stained red by intercalation of PI with DNA andor RNA

RESULTS

Preparation of Hb-C

Purified Hb and Hb-C showed a typical oxy-hemoglobin UV-Vis spectrum withnegligible met-hemoglobin content as previously reported [18] The high oxy-hemoglobin content in Hb-C is desirable for facilitated oxygen transport TheSDSndashPAGE chromatogram clearly revealed that the isolated Hb solution containeda minimal amount of impurities The chromatogram also qualitatively showedthe separation of high-molecular-mass Hb-C by ultrafiltration Compared tothe reaction mixture and ultrafiltration filtrate the final Hb-C displayed delayedretardation through the polyacrylamide gel owing to the high cross-linking densityThe results of SDSndashPAGE are presented in Fig 2

Preparation of ECM and semi-permeable membrane

The polymeric ECM of poly(NIPAAm-co-AAc) had a high molecular mass (morethan 1000 kDa) and an aqueous solution of the polymer showed novel concentration

Enhanced function of islet macrocapsule 1527

Figure 2 Characterization of PEG-Hb conjugate using SDSndashPAGE (12 gel) Lane 1 standardmarkers lane 2 isolated Hb lane 3 after conjugation lane 4 purified Hb-C lane 5 ultrafiltrationfiltrate

and temperature-dependent solndashgel transition phenomena without hysterisis asreported by Han et al [19] The solndashgel transition temperature of the 7 aqueoussolution was 33C After double spin-coating a smooth dense and permeability-selective thin poly(HEMA-MMA-MAA) hydrogel top layer was anchored on thetop layer of a porous Duroporereg membrane SEM results showed that the thicknessof the coating layer was less than 5 microm and 30ndash60 of the coating layer anchoredthrough the Duroporereg membrane (data not shown) Serum and protein mixturepermeability test showed that the MWCO of the spin-coated membrane was about70 times 103 and IgG and other immunoglobulin were completely prevented frompenetrating through the membrane [27]

Optimization of Hb-C concentration for macroencapsulation

Hb-C-concentration-dependent glucose stimulation in insulin secretion and cellviability were observed with macroencapsulated rat islets (1000 islets per pouch154 microl loading volume) As demonstrated in Fig 3A and 3B the basal andglucose-stimulated insulin secretion functions of macroencapsulated islets werepreserved by the incorporation of Hb-C The insulin-secreting functions wereincreased with increasing Hb-C concentration up to 025 mM and then slightlydecreased with a higher Hb-C concentration (05 mM Hb-C) Compared to control(macroencapsulated islets without Hb-C) 213 and 303 enhanced insulin

1528 Y Y Kim et al

Figure 3 Static insulin secretion of macroencapsulated islets at basal and stimulated glucoseconcentration (A) Basal glucose at 100 mgdl (B) stimulated glucose at 300 mgdl in RPMI-1640medium Macrocapsules (1000 islets per pouch) were incubated with 6 ml RPMI-1640 medium for4 h The results are expressed as mean plusmn SD (n = 5) Statistical differences between 1st week and4th week data were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

Enhanced function of islet macrocapsule 1529

secretion was observed in macrocapsules containing 025 mM Hb-C for basal(100 mgdl) and high glucose (300 mgdl) stimulation respectively on day 3after encapsulation (hereafter referred to as 1st week) The values increased to704 (100 mgdl glucose) and 994 (300 mgdl glucose) at week 4 after staticincubation Compared to the 1st week results the basal and stimulated insulinsecretion of the control group dropped to 61 and 51 after 4 weeks culturerespectively In case of the 025 mM Hb-C containing group these values were 86and 78 for basal and glucose-stimulated conditions respectively The stimulationindex (ratio of glucose-stimulated to basal insulin secretion) also showed a similartrend Enhanced insulin secretion (154 times) by glucose stimulation in the controlgroup was decreased to 128 times after 4 weeks culture On the other hand thestimulation indexes of 025 mM Hb-C containing group were 154 and 150 for 1stand 4th week culture respectively

The cell viability data obtained by MTS assay also showed a similar tendencyfor the incorporated Hb-C as shown in Fig 4 The group containing 025 mM Hb-Cshowed the best cell viability after 1 and 4 weeks culture Compared to the 1st weekvalue 825 relative viability was obtained in 025 mM Hb-C group after 4 weeksculture whereas only 376 cells maintained were viable after 4 weeks withoutHb-C

Figure 4 Effect of Hb-C concentration on islet viability Islets (1000 islets per pouch) wereencapsulated with different concentrations of Hb-C and the viability was measured by MTS assayafter different time (3 days and 4 weeks after encapsulation) Data are mean plusmn SD (n = 5) Statisticaldifferences between 1st week and 4th week data were calculated using unpaired Studentrsquos t-tests(lowastP lt 005 lowastlowastP lt 001)

1530 Y Y Kim et al

Optimization of islet density for macroencapsulation

The number of islets per macrocapsule device also strongly affected the insulin-secreting function and viability of islets Basal (100 mgdl) and glucose-stimulated(300 mgdl glucose) insulin secretion functions of the control group (without Hb-C)and Hb-C group with different islet density are shown in Fig 5A and 5B Thebasal and glucose insulin secretion tests were performed on day 3 and week 4after encapsulation The tests performed on day 3 after encapsulation did not showa significant dependence on number of islets encapsulated Therefore the dataobtained with 1000 isletsdevice were employed as control

In the case of the experimental control group (without Hb-C) a dramaticdecrease of basal and glucose-stimulated insulin secretion was observed with 2000isletsdevice group (Fig 5A) Compared with control (1st week data) only 81and 150 insulin was secreted by basal and high glucose stimulation respectivelyWhen decreasing islet density to 1000 and 500 isletsdevice the insulin-secretionfunctions were enhanced The 1000 isletsdevice group showed 635 and 630of relative insulin secretion by basal and high glucose stimulation respectivelyAlthough slightly decreased insulin secretion was observed with 500 isletsdevicegroup compared to the 1000 isletsdevice group they did not show statisticaldifferences between the two groups

Introduction of the Hb-C into the macrocapsule device significantly enhancedthe insulin-secretion function measured at 4 weeks after static culture The threeHb-C groups showed a level of insulin secretion superior to the experimentalcontrol groups (without Hb-C) In case of the 2000 isletsdevice group 650and 730 relative (in comparison with 1st week control) insulin secretion wasobserved by basal and high glucose stimulation respectively The situations wereenhanced by decreasing the islet density to 500 and 1000 isletsdevice More than80 of initial insulin secretion was measured in the 1000 and 500 isletsdevicegroup independent of islet density and glucose concentration Reduced cellviability with rising islet density and enhanced cell viability by incorporation ofHb-C into islet macrocapsules data were obtained from cell viability tests usingMTS reagents (Fig 6) In the case of Hb-C-containing groups the relative cellviabilities (compared to week 1 data of 1000 isletspouch) were maintained at morethan 80 in the 1000 and 500 isletsdevice groups after 4 weeks static cultureSlightly reduced cell viability (relative viability of 655) was observed in the2000 isletsdevice group In the case of control groups (without Hb-C) remarkabledifferences in cell viability were observed The relative cell viabilities of around50 for the 1000 and 500 isletsdevice groups dropped to 225 by increasing theislet density to 2000 isletsdevice

Cell viability staining using AOPI showed similar trends As shown in Fig 7 theretrieved islets from the Hb-C-containing groups showed a green colored viablecell state A negligible difference of cell staining was observed between 1000and 500 isletsdevice Increasing the islet density to 2000 isletsdevice showed anegative effect on cell viability because of a slight development of central necrosis

Enhanced function of islet macrocapsule 1531

Figure 5 Static insulin secretion of macroencapsulated islets cultured in medium with basal andhigh glucose levels (A) Without Hb-C (B) with 025 mM Hb-C Macrocapsules with different isletdensity (2000 1000 and 500 isletsdevice) were incubated with 6 ml RPMI-1640 medium for 4 h andsecreted insulin contents were measured by insulin radioimmunoassay The unit of G is mgdl Thedata are expressed as mean plusmn SD (n = 4) Statistical differences between 1st week and 4th week datawere calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

1532 Y Y Kim et al

Figure 6 Cell viability of macroencapsulated islets with different islet density after 4 weeks staticculture Data are shown as meanplusmnSD (n = 4) Statistical differences between 1st week and 4th weekdata were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

as shown in Fig 7 However the results for the control groups (without Hb-C)showed a remarkable difference between the 2000 and 1000 isletsdevice groupsThe islets from the 2000 isletsdevice group showed a typical red colored deadcell staining with destruction of integrated islet structure due to severe cell necrosisandor apoptosis Compared to the 2000 isletsdevice group slightly enhanced cellviability but with large necrotic core and viable rim of islets was observed in the1000 isletsdevice group as shown in Fig 7

DISCUSSION

In this study we investigated the optimum conditions of Hb-C concentration andislet density for islet macrocapsules Islets were immobilized by a thermore-versible polymeric ECM Among several devices proposed for islet immunoisola-tion macrocapsule devices possess advantages such as ease of implantation and ex-plantation with minimal surgical risk variety of implantation site and reduced trans-plantation volume by increasing islet density The thermosensitive polymeric ECMwhich showed a unique sol-to-gel transition may allow one to replace degeneratedislets after graft failure in the macroencapsulation devices by fresh islets by slightlydecreasing the temperature without surgical operations [6 19ndash21] Moreover the

Enhanced function of islet macrocapsule 1533

Figure 7 Fluorescence images of islets in the macrocapsules after 4 weeks of culture (A) 2000isletspouch with 025 mM Hb-C (B) 2000 isletspouch without Hb-C (C) 1000 isletspouch with025 mM Hb-C and (D) 1000 isletspouch without Hb-C

introduction of a bioactive ingredient (Hb-C) may enhance the cell functionality andviability as shown in our previous islet microencapsulation system [18 25]

As shown in Figs 3 and 4 insulin-secreting functions and cell viability weregreatly affected by the incorporated Hb-C in a concentration-dependent fashionThe optimum concentration of Hb-C as obtained from the experiments was025 mM (based on only Hb) The enhancing effect of the incorporated Hb-C oncell functionality and viability could be explained by the oxygen supply and otherreported bioactivities such as radical scavenging buffering and cell-stabilizingeffect [15 18 25] Most importantly increased oxygen supply may play acritical role in the enhanced cell functionality and viability Many studies havefocused on the limited oxygen supply on isolated and encapsulated islets Insulinsecretion especially glucose-stimulated insulin secretion was greatly affected byoxygen partial pressure [9] and the malfunction of islets or insulinoma cellswas overcome by increasing oxygen supply using hemoglobin Hb-C and waterelectrohydrolysis [15 16 18] However increasing the Hb-C concentration overthe optimum concentration resulted in decreased functionality and viability ofencapsulated islets as shown in Figs 3 and 4 The phenomena may be attributed todeteriorated mass transport by increased viscosity andor density of the polymericECM Although the polymeric ECM itself did not act as a diffusional barrier forsmall molecules such as insulin nutrients and wastes [29] a high concentrationof Hb-C in polymeric ECM may act as a mass transport barrier which resulted inreduced cell functionality and viability

Size or volume issue of a BAP is another important factor for its successfulapplication In the case of islet macroencapsulation the volume issue could be

1534 Y Y Kim et al

overcome by increasing islet density However cell viability is significantly affectedby cell density Figures 6 and 7 clearly show the decreased cell viability byincreasing islet density The enhanced cell viability by incorporation of Hb-Cwas caused by Hb-C groups however a high islet density (2000 isletsdevice)reduced viability by developing central necrosis By increasing islet density strongcompetition between encapsulated islets for nutrients and oxygen and seriousaccumulation of waste may occur in the device which results in reduced cellviability by increasing cell necrosis andor apoptosis On the other hand decreasingthe islet density increases the device volume which may evoke the volume issue

CONCLUSIONS

In this report we investigated the effect of Hb-C on the functionality and viabilityof an islet macroencapsulation device with a thermally reversible polymeric ECMBy introduction of 025 mM Hb-C into the islet macrocapsule more than 100enhanced insulin secretion and cell viability were observed after 4 weeks cultureFurthermore increase of islet density up to 1000 isletsdevice (154 microl volume)without any negative effect on insulin secretion and viability was possible byintroduction of 025 mM Hb into islet macrocapsules Although the practical in vivostudy remains as further study the incorporation of Hb-C into islet macrocapsulemight prolong the life span of encapsulated islets and reduce the implant volume byincreasing the functionality of islets and islet density

Acknowledgements

This work was supported by NIH DK56884 and BK 21 program (South Korea)

REFERENCES

1 G Reach Diabet Med 10 105 (1993)2 R P Lanza and W L Chick Ann N Y Acad Sci 831 323 (1997)3 R Calafiore Diabet Metab Rev 14 315 (1998)4 M K Lee and Y H Bae Adv Drug Deliv Rev 42 103 (2000)5 C K Colton and E S Avgoustiniators J Biomech Eng 113 152 (1992)6 Q P Hou and Y H Bae Adv Drug Deliv Rev 35 271 (1999)7 D Brandhorst H Brandhorst A Zwolinski F Nahidi and R G Bretzel Transplantation 71

179 (2001)8 J C Hutton and W J Malaisse Diabetologia 18 395 (1980)9 K E Dionne C K Colton and M L Yarmush Diabetes 42 12 (1993)

10 J Schrezenmeir L Gero M Solhdju J Kirchgessner C Laue J Beyer H Stier and W Muller-Klieser Transplant Proc 26 809 (1994)

11 M Ohta D Nelson M D Meglasson and M Erecinska J Biol Chem 265 17525 (1990)12 J S Whang S Y Chae M K Lee and Y H Bae Biomaterials 19 1198 (1998)13 K H Park S W Kim and Y H Bae J Biomed Mater Res 55 72 (2001)14 H Gappa M Baudys J J Koh S W Kim and Y H Bae Tissue Eng 7 35 (2001)

1524 Y Y Kim et al

brane) The Hb solution was sterilized by filtration though a 022 microm pore sizedsyringe filter and stored in a refrigerator at 4C

Hb-C used in this study was prepared by the same method reported in our previouspublication [18] Hb cross-linking was conducted by a reaction between residualamino groups of Hb with activated PEG diacid (PEG N-hydroxysuccinimide ester(NHS-PEG) 2 kDa) In brief predetermined amounts of PEG and Hb (PEGHbmol ratio 10 1) were added to a one-neck round-bottom flask and stirred at 4Cfor 4 h After the cross-linking reaction Hb-C was purified from the mixture byultrafiltration with a 100times103 MWCO membrane to remove monomeric Hb and by-products The sterilized Hb-C solution was obtained by filtration through a 022 micrompore size syringe filter and stored at 4C The resulting Hb-C was characterized bySDSndashPAGE

Preparation of extracelluar matrix (ECM) and semi-permeable membrane

As a polymeric ECM the co-polymers of NiPAAm with AAc (98 2 feed mol ratio)were synthesized by free radical polymerization in benzene (10 wt monomerconcentration) with AIBN as an initiator (7 times 10minus3 mol AIBNmol monomer) asreported previously [19]

The semi-permeable immunoprotecting membrane was fabricated by in situpolymerization with spin-coating of poly(HEMA-co-MMA-co-MAA) (69 30 molratio in feed of HEMA to MMA and 1 wt of MAA of HEMA and MMA) hydrogelon Duraporereg membranes (20 mm times 20 mm) Briefly before loading monomersolutions (HEMA MMA MAA and cross-linking agent of 15-hexadiene-34-diol)and initiator (10 APS and TEMED) were mixed and then after the appropriateviscosity was obtained the mixture was loaded (6 min after adding the initiator)After loading the first spin-coating and gel formation occurred simultaneously Inthe second coating loading time and loading amount were reduced compared to thefirst coating because the purpose of double coating is to improve surface smoothnessand not to increase thickness of the coating layer [27]

Islets isolation and encapsulation

Islets of Langerhans were isolated from pancreata of male SpraguendashDawley rats(200ndash250 g) by collagenase digestion and discontinuous Ficoll density gradientcentrifugation [18] The isolated islets were incubated in RPMI-1640 mediasupplemented with 10 heat-inactivated FBS for 1 day at 37C in a 95 air5CO2 atmosphere Highly purified (more than 98) islets were obtained byhandpicking The handpicked islets were cultured for 24 h prior to encapsulation

The isolated islets were macroencapsulated in the planar diffusion chamber withtwo different loading conditions Firstly 1000 islets were macroencapsulatedwith ECM solutions containing varying amounts of Hb-C (0 0125 025 and05 mM based on Hb in HBSS) without changing the number of islets fora macrocapsule device Secondly different numbers of islets (500 1000 and

Enhanced function of islet macrocapsule 1525

Figure 1 Schematic diagram of islet macroencapsulation device A typical device is composedof three Teflon rings (top middle and bottom) and two immunoisolating membranes (upper andlower) For macroencapsulation the lower membrane was fixed with the bottom and middle rings theisletECM mixture was poured into the central space and then the device was sealed with the uppermembrane and the top ring

2000 islets per device) were encapsulated using 025 mM Hb-C containing ECMsolution The macrocapsule devices (volume 154 microl) were composed of threecomponents supports made of three polytetrafluoroethylene (Teflon) rings twosemi-permeable membranes that formed the walls of the encapsulation chamberand a thermoreversible polymeric ECM that immobilized the islets at physiologicaltemperature A schematic illustration of the macrocapsule device and its dimensionsare shown in Fig 1 For encapsulation the bottom ring lower immunoisolatingmembrane and middle ring were assembled The islets were mixed with thepolymeric ECM solution (7 wt concentration in HBSS with and without Hb-C)at 4C and placed on the device which was closed with a upper membrane and topring Then each device was placed in a 6-well plate with 6 ml culture medium andincubated in humidified atmosphere (5 CO237C) Culture media (RPMI-1640with 10 FBS) were changed every other day

Static culture of encapsulated islets

To determine an optimum Hb-C concentration various amounts of Hb-C (00125 025 and 05 mM Hb equivalents) were applied to evaluate concentration-dependent Hb-C effects on macroencapsulated islets within polymeric ECM Toinvestigate basal and glucose stimulated insulin secretion of the encapsulatedislets the islet macrocapsules were incubated at different glucose concentrationsof 100 mgdl and 300 mgdl for 4 h Samples (250 microl) were taken from the plateand insulin concentration was measured by insulin radioimmunoassay (RIA) Thestatic glucose stimulation tests were performed on day 3 (1st week) and week 4

1526 Y Y Kim et al

after encapsulation To investigate cell viability the islets were retrieved from themacrocapsules on day 3 (1st week) and week 4 after encapsulation and the cellviability was measured using MTS reagent

After optimization of Hb-C concentration (025 mM Hb equivalent) various num-bers of islets (2000 1000 and 500 isletsdevice) were encapsulated in macroencap-sulation devices with polymeric ECM (with and without Hb-C) and statically cul-tured for 4 weeks with glucose-stimulated insulin secretion tests and viability assaysby MTS assays and AOPI staining

Cell viability tests

MTS assay The viability of islets retrieved from macrocapsules was measuredby MTS assay Briefly the 30 retrieved islets were placed into a 96-well plate with20 microl of assay solution and 100 microl cell culture medium and incubated at 37C for4 h After incubation the absorbance at 490 nm was measured by using a platereader The obtained optical density values were normalized by subtraction of theblank reading (without islets)

AOPI staining After static culture the retrieved islets were stained with AOPIto investigate cell viability [28] Briefly the retrieved islets were placed into 96wells and stained with AOPI solution (100 microl AOPI solution 067 microgml AO and75 microgml PI in HBSS) for 10 min in a dark place After removing the remainingAOPI by washing with PBS the islets were observed with fluorescence microscopeThrough the staining viable cells are stained bright green by the incorporation ofAO in nucleic acids Due to the impermeability for PI of a viable cell membraneonly dead cells were stained red by intercalation of PI with DNA andor RNA

RESULTS

Preparation of Hb-C

Purified Hb and Hb-C showed a typical oxy-hemoglobin UV-Vis spectrum withnegligible met-hemoglobin content as previously reported [18] The high oxy-hemoglobin content in Hb-C is desirable for facilitated oxygen transport TheSDSndashPAGE chromatogram clearly revealed that the isolated Hb solution containeda minimal amount of impurities The chromatogram also qualitatively showedthe separation of high-molecular-mass Hb-C by ultrafiltration Compared tothe reaction mixture and ultrafiltration filtrate the final Hb-C displayed delayedretardation through the polyacrylamide gel owing to the high cross-linking densityThe results of SDSndashPAGE are presented in Fig 2

Preparation of ECM and semi-permeable membrane

The polymeric ECM of poly(NIPAAm-co-AAc) had a high molecular mass (morethan 1000 kDa) and an aqueous solution of the polymer showed novel concentration

Enhanced function of islet macrocapsule 1527

Figure 2 Characterization of PEG-Hb conjugate using SDSndashPAGE (12 gel) Lane 1 standardmarkers lane 2 isolated Hb lane 3 after conjugation lane 4 purified Hb-C lane 5 ultrafiltrationfiltrate

and temperature-dependent solndashgel transition phenomena without hysterisis asreported by Han et al [19] The solndashgel transition temperature of the 7 aqueoussolution was 33C After double spin-coating a smooth dense and permeability-selective thin poly(HEMA-MMA-MAA) hydrogel top layer was anchored on thetop layer of a porous Duroporereg membrane SEM results showed that the thicknessof the coating layer was less than 5 microm and 30ndash60 of the coating layer anchoredthrough the Duroporereg membrane (data not shown) Serum and protein mixturepermeability test showed that the MWCO of the spin-coated membrane was about70 times 103 and IgG and other immunoglobulin were completely prevented frompenetrating through the membrane [27]

Optimization of Hb-C concentration for macroencapsulation

Hb-C-concentration-dependent glucose stimulation in insulin secretion and cellviability were observed with macroencapsulated rat islets (1000 islets per pouch154 microl loading volume) As demonstrated in Fig 3A and 3B the basal andglucose-stimulated insulin secretion functions of macroencapsulated islets werepreserved by the incorporation of Hb-C The insulin-secreting functions wereincreased with increasing Hb-C concentration up to 025 mM and then slightlydecreased with a higher Hb-C concentration (05 mM Hb-C) Compared to control(macroencapsulated islets without Hb-C) 213 and 303 enhanced insulin

1528 Y Y Kim et al

Figure 3 Static insulin secretion of macroencapsulated islets at basal and stimulated glucoseconcentration (A) Basal glucose at 100 mgdl (B) stimulated glucose at 300 mgdl in RPMI-1640medium Macrocapsules (1000 islets per pouch) were incubated with 6 ml RPMI-1640 medium for4 h The results are expressed as mean plusmn SD (n = 5) Statistical differences between 1st week and4th week data were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

Enhanced function of islet macrocapsule 1529

secretion was observed in macrocapsules containing 025 mM Hb-C for basal(100 mgdl) and high glucose (300 mgdl) stimulation respectively on day 3after encapsulation (hereafter referred to as 1st week) The values increased to704 (100 mgdl glucose) and 994 (300 mgdl glucose) at week 4 after staticincubation Compared to the 1st week results the basal and stimulated insulinsecretion of the control group dropped to 61 and 51 after 4 weeks culturerespectively In case of the 025 mM Hb-C containing group these values were 86and 78 for basal and glucose-stimulated conditions respectively The stimulationindex (ratio of glucose-stimulated to basal insulin secretion) also showed a similartrend Enhanced insulin secretion (154 times) by glucose stimulation in the controlgroup was decreased to 128 times after 4 weeks culture On the other hand thestimulation indexes of 025 mM Hb-C containing group were 154 and 150 for 1stand 4th week culture respectively

The cell viability data obtained by MTS assay also showed a similar tendencyfor the incorporated Hb-C as shown in Fig 4 The group containing 025 mM Hb-Cshowed the best cell viability after 1 and 4 weeks culture Compared to the 1st weekvalue 825 relative viability was obtained in 025 mM Hb-C group after 4 weeksculture whereas only 376 cells maintained were viable after 4 weeks withoutHb-C

Figure 4 Effect of Hb-C concentration on islet viability Islets (1000 islets per pouch) wereencapsulated with different concentrations of Hb-C and the viability was measured by MTS assayafter different time (3 days and 4 weeks after encapsulation) Data are mean plusmn SD (n = 5) Statisticaldifferences between 1st week and 4th week data were calculated using unpaired Studentrsquos t-tests(lowastP lt 005 lowastlowastP lt 001)

1530 Y Y Kim et al

Optimization of islet density for macroencapsulation

The number of islets per macrocapsule device also strongly affected the insulin-secreting function and viability of islets Basal (100 mgdl) and glucose-stimulated(300 mgdl glucose) insulin secretion functions of the control group (without Hb-C)and Hb-C group with different islet density are shown in Fig 5A and 5B Thebasal and glucose insulin secretion tests were performed on day 3 and week 4after encapsulation The tests performed on day 3 after encapsulation did not showa significant dependence on number of islets encapsulated Therefore the dataobtained with 1000 isletsdevice were employed as control

In the case of the experimental control group (without Hb-C) a dramaticdecrease of basal and glucose-stimulated insulin secretion was observed with 2000isletsdevice group (Fig 5A) Compared with control (1st week data) only 81and 150 insulin was secreted by basal and high glucose stimulation respectivelyWhen decreasing islet density to 1000 and 500 isletsdevice the insulin-secretionfunctions were enhanced The 1000 isletsdevice group showed 635 and 630of relative insulin secretion by basal and high glucose stimulation respectivelyAlthough slightly decreased insulin secretion was observed with 500 isletsdevicegroup compared to the 1000 isletsdevice group they did not show statisticaldifferences between the two groups

Introduction of the Hb-C into the macrocapsule device significantly enhancedthe insulin-secretion function measured at 4 weeks after static culture The threeHb-C groups showed a level of insulin secretion superior to the experimentalcontrol groups (without Hb-C) In case of the 2000 isletsdevice group 650and 730 relative (in comparison with 1st week control) insulin secretion wasobserved by basal and high glucose stimulation respectively The situations wereenhanced by decreasing the islet density to 500 and 1000 isletsdevice More than80 of initial insulin secretion was measured in the 1000 and 500 isletsdevicegroup independent of islet density and glucose concentration Reduced cellviability with rising islet density and enhanced cell viability by incorporation ofHb-C into islet macrocapsules data were obtained from cell viability tests usingMTS reagents (Fig 6) In the case of Hb-C-containing groups the relative cellviabilities (compared to week 1 data of 1000 isletspouch) were maintained at morethan 80 in the 1000 and 500 isletsdevice groups after 4 weeks static cultureSlightly reduced cell viability (relative viability of 655) was observed in the2000 isletsdevice group In the case of control groups (without Hb-C) remarkabledifferences in cell viability were observed The relative cell viabilities of around50 for the 1000 and 500 isletsdevice groups dropped to 225 by increasing theislet density to 2000 isletsdevice

Cell viability staining using AOPI showed similar trends As shown in Fig 7 theretrieved islets from the Hb-C-containing groups showed a green colored viablecell state A negligible difference of cell staining was observed between 1000and 500 isletsdevice Increasing the islet density to 2000 isletsdevice showed anegative effect on cell viability because of a slight development of central necrosis

Enhanced function of islet macrocapsule 1531

Figure 5 Static insulin secretion of macroencapsulated islets cultured in medium with basal andhigh glucose levels (A) Without Hb-C (B) with 025 mM Hb-C Macrocapsules with different isletdensity (2000 1000 and 500 isletsdevice) were incubated with 6 ml RPMI-1640 medium for 4 h andsecreted insulin contents were measured by insulin radioimmunoassay The unit of G is mgdl Thedata are expressed as mean plusmn SD (n = 4) Statistical differences between 1st week and 4th week datawere calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

1532 Y Y Kim et al

Figure 6 Cell viability of macroencapsulated islets with different islet density after 4 weeks staticculture Data are shown as meanplusmnSD (n = 4) Statistical differences between 1st week and 4th weekdata were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

as shown in Fig 7 However the results for the control groups (without Hb-C)showed a remarkable difference between the 2000 and 1000 isletsdevice groupsThe islets from the 2000 isletsdevice group showed a typical red colored deadcell staining with destruction of integrated islet structure due to severe cell necrosisandor apoptosis Compared to the 2000 isletsdevice group slightly enhanced cellviability but with large necrotic core and viable rim of islets was observed in the1000 isletsdevice group as shown in Fig 7

DISCUSSION

In this study we investigated the optimum conditions of Hb-C concentration andislet density for islet macrocapsules Islets were immobilized by a thermore-versible polymeric ECM Among several devices proposed for islet immunoisola-tion macrocapsule devices possess advantages such as ease of implantation and ex-plantation with minimal surgical risk variety of implantation site and reduced trans-plantation volume by increasing islet density The thermosensitive polymeric ECMwhich showed a unique sol-to-gel transition may allow one to replace degeneratedislets after graft failure in the macroencapsulation devices by fresh islets by slightlydecreasing the temperature without surgical operations [6 19ndash21] Moreover the

Enhanced function of islet macrocapsule 1533

Figure 7 Fluorescence images of islets in the macrocapsules after 4 weeks of culture (A) 2000isletspouch with 025 mM Hb-C (B) 2000 isletspouch without Hb-C (C) 1000 isletspouch with025 mM Hb-C and (D) 1000 isletspouch without Hb-C

introduction of a bioactive ingredient (Hb-C) may enhance the cell functionality andviability as shown in our previous islet microencapsulation system [18 25]

As shown in Figs 3 and 4 insulin-secreting functions and cell viability weregreatly affected by the incorporated Hb-C in a concentration-dependent fashionThe optimum concentration of Hb-C as obtained from the experiments was025 mM (based on only Hb) The enhancing effect of the incorporated Hb-C oncell functionality and viability could be explained by the oxygen supply and otherreported bioactivities such as radical scavenging buffering and cell-stabilizingeffect [15 18 25] Most importantly increased oxygen supply may play acritical role in the enhanced cell functionality and viability Many studies havefocused on the limited oxygen supply on isolated and encapsulated islets Insulinsecretion especially glucose-stimulated insulin secretion was greatly affected byoxygen partial pressure [9] and the malfunction of islets or insulinoma cellswas overcome by increasing oxygen supply using hemoglobin Hb-C and waterelectrohydrolysis [15 16 18] However increasing the Hb-C concentration overthe optimum concentration resulted in decreased functionality and viability ofencapsulated islets as shown in Figs 3 and 4 The phenomena may be attributed todeteriorated mass transport by increased viscosity andor density of the polymericECM Although the polymeric ECM itself did not act as a diffusional barrier forsmall molecules such as insulin nutrients and wastes [29] a high concentrationof Hb-C in polymeric ECM may act as a mass transport barrier which resulted inreduced cell functionality and viability

Size or volume issue of a BAP is another important factor for its successfulapplication In the case of islet macroencapsulation the volume issue could be

1534 Y Y Kim et al

overcome by increasing islet density However cell viability is significantly affectedby cell density Figures 6 and 7 clearly show the decreased cell viability byincreasing islet density The enhanced cell viability by incorporation of Hb-Cwas caused by Hb-C groups however a high islet density (2000 isletsdevice)reduced viability by developing central necrosis By increasing islet density strongcompetition between encapsulated islets for nutrients and oxygen and seriousaccumulation of waste may occur in the device which results in reduced cellviability by increasing cell necrosis andor apoptosis On the other hand decreasingthe islet density increases the device volume which may evoke the volume issue

CONCLUSIONS

In this report we investigated the effect of Hb-C on the functionality and viabilityof an islet macroencapsulation device with a thermally reversible polymeric ECMBy introduction of 025 mM Hb-C into the islet macrocapsule more than 100enhanced insulin secretion and cell viability were observed after 4 weeks cultureFurthermore increase of islet density up to 1000 isletsdevice (154 microl volume)without any negative effect on insulin secretion and viability was possible byintroduction of 025 mM Hb into islet macrocapsules Although the practical in vivostudy remains as further study the incorporation of Hb-C into islet macrocapsulemight prolong the life span of encapsulated islets and reduce the implant volume byincreasing the functionality of islets and islet density

Acknowledgements

This work was supported by NIH DK56884 and BK 21 program (South Korea)

REFERENCES

1 G Reach Diabet Med 10 105 (1993)2 R P Lanza and W L Chick Ann N Y Acad Sci 831 323 (1997)3 R Calafiore Diabet Metab Rev 14 315 (1998)4 M K Lee and Y H Bae Adv Drug Deliv Rev 42 103 (2000)5 C K Colton and E S Avgoustiniators J Biomech Eng 113 152 (1992)6 Q P Hou and Y H Bae Adv Drug Deliv Rev 35 271 (1999)7 D Brandhorst H Brandhorst A Zwolinski F Nahidi and R G Bretzel Transplantation 71

179 (2001)8 J C Hutton and W J Malaisse Diabetologia 18 395 (1980)9 K E Dionne C K Colton and M L Yarmush Diabetes 42 12 (1993)

10 J Schrezenmeir L Gero M Solhdju J Kirchgessner C Laue J Beyer H Stier and W Muller-Klieser Transplant Proc 26 809 (1994)

11 M Ohta D Nelson M D Meglasson and M Erecinska J Biol Chem 265 17525 (1990)12 J S Whang S Y Chae M K Lee and Y H Bae Biomaterials 19 1198 (1998)13 K H Park S W Kim and Y H Bae J Biomed Mater Res 55 72 (2001)14 H Gappa M Baudys J J Koh S W Kim and Y H Bae Tissue Eng 7 35 (2001)

Enhanced function of islet macrocapsule 1525

Figure 1 Schematic diagram of islet macroencapsulation device A typical device is composedof three Teflon rings (top middle and bottom) and two immunoisolating membranes (upper andlower) For macroencapsulation the lower membrane was fixed with the bottom and middle rings theisletECM mixture was poured into the central space and then the device was sealed with the uppermembrane and the top ring

2000 islets per device) were encapsulated using 025 mM Hb-C containing ECMsolution The macrocapsule devices (volume 154 microl) were composed of threecomponents supports made of three polytetrafluoroethylene (Teflon) rings twosemi-permeable membranes that formed the walls of the encapsulation chamberand a thermoreversible polymeric ECM that immobilized the islets at physiologicaltemperature A schematic illustration of the macrocapsule device and its dimensionsare shown in Fig 1 For encapsulation the bottom ring lower immunoisolatingmembrane and middle ring were assembled The islets were mixed with thepolymeric ECM solution (7 wt concentration in HBSS with and without Hb-C)at 4C and placed on the device which was closed with a upper membrane and topring Then each device was placed in a 6-well plate with 6 ml culture medium andincubated in humidified atmosphere (5 CO237C) Culture media (RPMI-1640with 10 FBS) were changed every other day

Static culture of encapsulated islets

To determine an optimum Hb-C concentration various amounts of Hb-C (00125 025 and 05 mM Hb equivalents) were applied to evaluate concentration-dependent Hb-C effects on macroencapsulated islets within polymeric ECM Toinvestigate basal and glucose stimulated insulin secretion of the encapsulatedislets the islet macrocapsules were incubated at different glucose concentrationsof 100 mgdl and 300 mgdl for 4 h Samples (250 microl) were taken from the plateand insulin concentration was measured by insulin radioimmunoassay (RIA) Thestatic glucose stimulation tests were performed on day 3 (1st week) and week 4

1526 Y Y Kim et al

after encapsulation To investigate cell viability the islets were retrieved from themacrocapsules on day 3 (1st week) and week 4 after encapsulation and the cellviability was measured using MTS reagent

After optimization of Hb-C concentration (025 mM Hb equivalent) various num-bers of islets (2000 1000 and 500 isletsdevice) were encapsulated in macroencap-sulation devices with polymeric ECM (with and without Hb-C) and statically cul-tured for 4 weeks with glucose-stimulated insulin secretion tests and viability assaysby MTS assays and AOPI staining

Cell viability tests

MTS assay The viability of islets retrieved from macrocapsules was measuredby MTS assay Briefly the 30 retrieved islets were placed into a 96-well plate with20 microl of assay solution and 100 microl cell culture medium and incubated at 37C for4 h After incubation the absorbance at 490 nm was measured by using a platereader The obtained optical density values were normalized by subtraction of theblank reading (without islets)

AOPI staining After static culture the retrieved islets were stained with AOPIto investigate cell viability [28] Briefly the retrieved islets were placed into 96wells and stained with AOPI solution (100 microl AOPI solution 067 microgml AO and75 microgml PI in HBSS) for 10 min in a dark place After removing the remainingAOPI by washing with PBS the islets were observed with fluorescence microscopeThrough the staining viable cells are stained bright green by the incorporation ofAO in nucleic acids Due to the impermeability for PI of a viable cell membraneonly dead cells were stained red by intercalation of PI with DNA andor RNA

RESULTS

Preparation of Hb-C

Purified Hb and Hb-C showed a typical oxy-hemoglobin UV-Vis spectrum withnegligible met-hemoglobin content as previously reported [18] The high oxy-hemoglobin content in Hb-C is desirable for facilitated oxygen transport TheSDSndashPAGE chromatogram clearly revealed that the isolated Hb solution containeda minimal amount of impurities The chromatogram also qualitatively showedthe separation of high-molecular-mass Hb-C by ultrafiltration Compared tothe reaction mixture and ultrafiltration filtrate the final Hb-C displayed delayedretardation through the polyacrylamide gel owing to the high cross-linking densityThe results of SDSndashPAGE are presented in Fig 2

Preparation of ECM and semi-permeable membrane

The polymeric ECM of poly(NIPAAm-co-AAc) had a high molecular mass (morethan 1000 kDa) and an aqueous solution of the polymer showed novel concentration

Enhanced function of islet macrocapsule 1527

Figure 2 Characterization of PEG-Hb conjugate using SDSndashPAGE (12 gel) Lane 1 standardmarkers lane 2 isolated Hb lane 3 after conjugation lane 4 purified Hb-C lane 5 ultrafiltrationfiltrate

and temperature-dependent solndashgel transition phenomena without hysterisis asreported by Han et al [19] The solndashgel transition temperature of the 7 aqueoussolution was 33C After double spin-coating a smooth dense and permeability-selective thin poly(HEMA-MMA-MAA) hydrogel top layer was anchored on thetop layer of a porous Duroporereg membrane SEM results showed that the thicknessof the coating layer was less than 5 microm and 30ndash60 of the coating layer anchoredthrough the Duroporereg membrane (data not shown) Serum and protein mixturepermeability test showed that the MWCO of the spin-coated membrane was about70 times 103 and IgG and other immunoglobulin were completely prevented frompenetrating through the membrane [27]

Optimization of Hb-C concentration for macroencapsulation

Hb-C-concentration-dependent glucose stimulation in insulin secretion and cellviability were observed with macroencapsulated rat islets (1000 islets per pouch154 microl loading volume) As demonstrated in Fig 3A and 3B the basal andglucose-stimulated insulin secretion functions of macroencapsulated islets werepreserved by the incorporation of Hb-C The insulin-secreting functions wereincreased with increasing Hb-C concentration up to 025 mM and then slightlydecreased with a higher Hb-C concentration (05 mM Hb-C) Compared to control(macroencapsulated islets without Hb-C) 213 and 303 enhanced insulin

1528 Y Y Kim et al

Figure 3 Static insulin secretion of macroencapsulated islets at basal and stimulated glucoseconcentration (A) Basal glucose at 100 mgdl (B) stimulated glucose at 300 mgdl in RPMI-1640medium Macrocapsules (1000 islets per pouch) were incubated with 6 ml RPMI-1640 medium for4 h The results are expressed as mean plusmn SD (n = 5) Statistical differences between 1st week and4th week data were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

Enhanced function of islet macrocapsule 1529

secretion was observed in macrocapsules containing 025 mM Hb-C for basal(100 mgdl) and high glucose (300 mgdl) stimulation respectively on day 3after encapsulation (hereafter referred to as 1st week) The values increased to704 (100 mgdl glucose) and 994 (300 mgdl glucose) at week 4 after staticincubation Compared to the 1st week results the basal and stimulated insulinsecretion of the control group dropped to 61 and 51 after 4 weeks culturerespectively In case of the 025 mM Hb-C containing group these values were 86and 78 for basal and glucose-stimulated conditions respectively The stimulationindex (ratio of glucose-stimulated to basal insulin secretion) also showed a similartrend Enhanced insulin secretion (154 times) by glucose stimulation in the controlgroup was decreased to 128 times after 4 weeks culture On the other hand thestimulation indexes of 025 mM Hb-C containing group were 154 and 150 for 1stand 4th week culture respectively

The cell viability data obtained by MTS assay also showed a similar tendencyfor the incorporated Hb-C as shown in Fig 4 The group containing 025 mM Hb-Cshowed the best cell viability after 1 and 4 weeks culture Compared to the 1st weekvalue 825 relative viability was obtained in 025 mM Hb-C group after 4 weeksculture whereas only 376 cells maintained were viable after 4 weeks withoutHb-C

Figure 4 Effect of Hb-C concentration on islet viability Islets (1000 islets per pouch) wereencapsulated with different concentrations of Hb-C and the viability was measured by MTS assayafter different time (3 days and 4 weeks after encapsulation) Data are mean plusmn SD (n = 5) Statisticaldifferences between 1st week and 4th week data were calculated using unpaired Studentrsquos t-tests(lowastP lt 005 lowastlowastP lt 001)

1530 Y Y Kim et al

Optimization of islet density for macroencapsulation

The number of islets per macrocapsule device also strongly affected the insulin-secreting function and viability of islets Basal (100 mgdl) and glucose-stimulated(300 mgdl glucose) insulin secretion functions of the control group (without Hb-C)and Hb-C group with different islet density are shown in Fig 5A and 5B Thebasal and glucose insulin secretion tests were performed on day 3 and week 4after encapsulation The tests performed on day 3 after encapsulation did not showa significant dependence on number of islets encapsulated Therefore the dataobtained with 1000 isletsdevice were employed as control

In the case of the experimental control group (without Hb-C) a dramaticdecrease of basal and glucose-stimulated insulin secretion was observed with 2000isletsdevice group (Fig 5A) Compared with control (1st week data) only 81and 150 insulin was secreted by basal and high glucose stimulation respectivelyWhen decreasing islet density to 1000 and 500 isletsdevice the insulin-secretionfunctions were enhanced The 1000 isletsdevice group showed 635 and 630of relative insulin secretion by basal and high glucose stimulation respectivelyAlthough slightly decreased insulin secretion was observed with 500 isletsdevicegroup compared to the 1000 isletsdevice group they did not show statisticaldifferences between the two groups

Introduction of the Hb-C into the macrocapsule device significantly enhancedthe insulin-secretion function measured at 4 weeks after static culture The threeHb-C groups showed a level of insulin secretion superior to the experimentalcontrol groups (without Hb-C) In case of the 2000 isletsdevice group 650and 730 relative (in comparison with 1st week control) insulin secretion wasobserved by basal and high glucose stimulation respectively The situations wereenhanced by decreasing the islet density to 500 and 1000 isletsdevice More than80 of initial insulin secretion was measured in the 1000 and 500 isletsdevicegroup independent of islet density and glucose concentration Reduced cellviability with rising islet density and enhanced cell viability by incorporation ofHb-C into islet macrocapsules data were obtained from cell viability tests usingMTS reagents (Fig 6) In the case of Hb-C-containing groups the relative cellviabilities (compared to week 1 data of 1000 isletspouch) were maintained at morethan 80 in the 1000 and 500 isletsdevice groups after 4 weeks static cultureSlightly reduced cell viability (relative viability of 655) was observed in the2000 isletsdevice group In the case of control groups (without Hb-C) remarkabledifferences in cell viability were observed The relative cell viabilities of around50 for the 1000 and 500 isletsdevice groups dropped to 225 by increasing theislet density to 2000 isletsdevice

Cell viability staining using AOPI showed similar trends As shown in Fig 7 theretrieved islets from the Hb-C-containing groups showed a green colored viablecell state A negligible difference of cell staining was observed between 1000and 500 isletsdevice Increasing the islet density to 2000 isletsdevice showed anegative effect on cell viability because of a slight development of central necrosis

Enhanced function of islet macrocapsule 1531

Figure 5 Static insulin secretion of macroencapsulated islets cultured in medium with basal andhigh glucose levels (A) Without Hb-C (B) with 025 mM Hb-C Macrocapsules with different isletdensity (2000 1000 and 500 isletsdevice) were incubated with 6 ml RPMI-1640 medium for 4 h andsecreted insulin contents were measured by insulin radioimmunoassay The unit of G is mgdl Thedata are expressed as mean plusmn SD (n = 4) Statistical differences between 1st week and 4th week datawere calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

1532 Y Y Kim et al

Figure 6 Cell viability of macroencapsulated islets with different islet density after 4 weeks staticculture Data are shown as meanplusmnSD (n = 4) Statistical differences between 1st week and 4th weekdata were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

as shown in Fig 7 However the results for the control groups (without Hb-C)showed a remarkable difference between the 2000 and 1000 isletsdevice groupsThe islets from the 2000 isletsdevice group showed a typical red colored deadcell staining with destruction of integrated islet structure due to severe cell necrosisandor apoptosis Compared to the 2000 isletsdevice group slightly enhanced cellviability but with large necrotic core and viable rim of islets was observed in the1000 isletsdevice group as shown in Fig 7

DISCUSSION

In this study we investigated the optimum conditions of Hb-C concentration andislet density for islet macrocapsules Islets were immobilized by a thermore-versible polymeric ECM Among several devices proposed for islet immunoisola-tion macrocapsule devices possess advantages such as ease of implantation and ex-plantation with minimal surgical risk variety of implantation site and reduced trans-plantation volume by increasing islet density The thermosensitive polymeric ECMwhich showed a unique sol-to-gel transition may allow one to replace degeneratedislets after graft failure in the macroencapsulation devices by fresh islets by slightlydecreasing the temperature without surgical operations [6 19ndash21] Moreover the

Enhanced function of islet macrocapsule 1533

Figure 7 Fluorescence images of islets in the macrocapsules after 4 weeks of culture (A) 2000isletspouch with 025 mM Hb-C (B) 2000 isletspouch without Hb-C (C) 1000 isletspouch with025 mM Hb-C and (D) 1000 isletspouch without Hb-C

introduction of a bioactive ingredient (Hb-C) may enhance the cell functionality andviability as shown in our previous islet microencapsulation system [18 25]

As shown in Figs 3 and 4 insulin-secreting functions and cell viability weregreatly affected by the incorporated Hb-C in a concentration-dependent fashionThe optimum concentration of Hb-C as obtained from the experiments was025 mM (based on only Hb) The enhancing effect of the incorporated Hb-C oncell functionality and viability could be explained by the oxygen supply and otherreported bioactivities such as radical scavenging buffering and cell-stabilizingeffect [15 18 25] Most importantly increased oxygen supply may play acritical role in the enhanced cell functionality and viability Many studies havefocused on the limited oxygen supply on isolated and encapsulated islets Insulinsecretion especially glucose-stimulated insulin secretion was greatly affected byoxygen partial pressure [9] and the malfunction of islets or insulinoma cellswas overcome by increasing oxygen supply using hemoglobin Hb-C and waterelectrohydrolysis [15 16 18] However increasing the Hb-C concentration overthe optimum concentration resulted in decreased functionality and viability ofencapsulated islets as shown in Figs 3 and 4 The phenomena may be attributed todeteriorated mass transport by increased viscosity andor density of the polymericECM Although the polymeric ECM itself did not act as a diffusional barrier forsmall molecules such as insulin nutrients and wastes [29] a high concentrationof Hb-C in polymeric ECM may act as a mass transport barrier which resulted inreduced cell functionality and viability

Size or volume issue of a BAP is another important factor for its successfulapplication In the case of islet macroencapsulation the volume issue could be

1534 Y Y Kim et al

overcome by increasing islet density However cell viability is significantly affectedby cell density Figures 6 and 7 clearly show the decreased cell viability byincreasing islet density The enhanced cell viability by incorporation of Hb-Cwas caused by Hb-C groups however a high islet density (2000 isletsdevice)reduced viability by developing central necrosis By increasing islet density strongcompetition between encapsulated islets for nutrients and oxygen and seriousaccumulation of waste may occur in the device which results in reduced cellviability by increasing cell necrosis andor apoptosis On the other hand decreasingthe islet density increases the device volume which may evoke the volume issue

CONCLUSIONS

In this report we investigated the effect of Hb-C on the functionality and viabilityof an islet macroencapsulation device with a thermally reversible polymeric ECMBy introduction of 025 mM Hb-C into the islet macrocapsule more than 100enhanced insulin secretion and cell viability were observed after 4 weeks cultureFurthermore increase of islet density up to 1000 isletsdevice (154 microl volume)without any negative effect on insulin secretion and viability was possible byintroduction of 025 mM Hb into islet macrocapsules Although the practical in vivostudy remains as further study the incorporation of Hb-C into islet macrocapsulemight prolong the life span of encapsulated islets and reduce the implant volume byincreasing the functionality of islets and islet density

Acknowledgements

This work was supported by NIH DK56884 and BK 21 program (South Korea)

REFERENCES

1 G Reach Diabet Med 10 105 (1993)2 R P Lanza and W L Chick Ann N Y Acad Sci 831 323 (1997)3 R Calafiore Diabet Metab Rev 14 315 (1998)4 M K Lee and Y H Bae Adv Drug Deliv Rev 42 103 (2000)5 C K Colton and E S Avgoustiniators J Biomech Eng 113 152 (1992)6 Q P Hou and Y H Bae Adv Drug Deliv Rev 35 271 (1999)7 D Brandhorst H Brandhorst A Zwolinski F Nahidi and R G Bretzel Transplantation 71

179 (2001)8 J C Hutton and W J Malaisse Diabetologia 18 395 (1980)9 K E Dionne C K Colton and M L Yarmush Diabetes 42 12 (1993)

10 J Schrezenmeir L Gero M Solhdju J Kirchgessner C Laue J Beyer H Stier and W Muller-Klieser Transplant Proc 26 809 (1994)

11 M Ohta D Nelson M D Meglasson and M Erecinska J Biol Chem 265 17525 (1990)12 J S Whang S Y Chae M K Lee and Y H Bae Biomaterials 19 1198 (1998)13 K H Park S W Kim and Y H Bae J Biomed Mater Res 55 72 (2001)14 H Gappa M Baudys J J Koh S W Kim and Y H Bae Tissue Eng 7 35 (2001)

1526 Y Y Kim et al

after encapsulation To investigate cell viability the islets were retrieved from themacrocapsules on day 3 (1st week) and week 4 after encapsulation and the cellviability was measured using MTS reagent

After optimization of Hb-C concentration (025 mM Hb equivalent) various num-bers of islets (2000 1000 and 500 isletsdevice) were encapsulated in macroencap-sulation devices with polymeric ECM (with and without Hb-C) and statically cul-tured for 4 weeks with glucose-stimulated insulin secretion tests and viability assaysby MTS assays and AOPI staining

Cell viability tests

MTS assay The viability of islets retrieved from macrocapsules was measuredby MTS assay Briefly the 30 retrieved islets were placed into a 96-well plate with20 microl of assay solution and 100 microl cell culture medium and incubated at 37C for4 h After incubation the absorbance at 490 nm was measured by using a platereader The obtained optical density values were normalized by subtraction of theblank reading (without islets)

AOPI staining After static culture the retrieved islets were stained with AOPIto investigate cell viability [28] Briefly the retrieved islets were placed into 96wells and stained with AOPI solution (100 microl AOPI solution 067 microgml AO and75 microgml PI in HBSS) for 10 min in a dark place After removing the remainingAOPI by washing with PBS the islets were observed with fluorescence microscopeThrough the staining viable cells are stained bright green by the incorporation ofAO in nucleic acids Due to the impermeability for PI of a viable cell membraneonly dead cells were stained red by intercalation of PI with DNA andor RNA

RESULTS

Preparation of Hb-C

Purified Hb and Hb-C showed a typical oxy-hemoglobin UV-Vis spectrum withnegligible met-hemoglobin content as previously reported [18] The high oxy-hemoglobin content in Hb-C is desirable for facilitated oxygen transport TheSDSndashPAGE chromatogram clearly revealed that the isolated Hb solution containeda minimal amount of impurities The chromatogram also qualitatively showedthe separation of high-molecular-mass Hb-C by ultrafiltration Compared tothe reaction mixture and ultrafiltration filtrate the final Hb-C displayed delayedretardation through the polyacrylamide gel owing to the high cross-linking densityThe results of SDSndashPAGE are presented in Fig 2

Preparation of ECM and semi-permeable membrane

The polymeric ECM of poly(NIPAAm-co-AAc) had a high molecular mass (morethan 1000 kDa) and an aqueous solution of the polymer showed novel concentration

Enhanced function of islet macrocapsule 1527

Figure 2 Characterization of PEG-Hb conjugate using SDSndashPAGE (12 gel) Lane 1 standardmarkers lane 2 isolated Hb lane 3 after conjugation lane 4 purified Hb-C lane 5 ultrafiltrationfiltrate

and temperature-dependent solndashgel transition phenomena without hysterisis asreported by Han et al [19] The solndashgel transition temperature of the 7 aqueoussolution was 33C After double spin-coating a smooth dense and permeability-selective thin poly(HEMA-MMA-MAA) hydrogel top layer was anchored on thetop layer of a porous Duroporereg membrane SEM results showed that the thicknessof the coating layer was less than 5 microm and 30ndash60 of the coating layer anchoredthrough the Duroporereg membrane (data not shown) Serum and protein mixturepermeability test showed that the MWCO of the spin-coated membrane was about70 times 103 and IgG and other immunoglobulin were completely prevented frompenetrating through the membrane [27]

Optimization of Hb-C concentration for macroencapsulation

Hb-C-concentration-dependent glucose stimulation in insulin secretion and cellviability were observed with macroencapsulated rat islets (1000 islets per pouch154 microl loading volume) As demonstrated in Fig 3A and 3B the basal andglucose-stimulated insulin secretion functions of macroencapsulated islets werepreserved by the incorporation of Hb-C The insulin-secreting functions wereincreased with increasing Hb-C concentration up to 025 mM and then slightlydecreased with a higher Hb-C concentration (05 mM Hb-C) Compared to control(macroencapsulated islets without Hb-C) 213 and 303 enhanced insulin

1528 Y Y Kim et al

Figure 3 Static insulin secretion of macroencapsulated islets at basal and stimulated glucoseconcentration (A) Basal glucose at 100 mgdl (B) stimulated glucose at 300 mgdl in RPMI-1640medium Macrocapsules (1000 islets per pouch) were incubated with 6 ml RPMI-1640 medium for4 h The results are expressed as mean plusmn SD (n = 5) Statistical differences between 1st week and4th week data were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

Enhanced function of islet macrocapsule 1529

secretion was observed in macrocapsules containing 025 mM Hb-C for basal(100 mgdl) and high glucose (300 mgdl) stimulation respectively on day 3after encapsulation (hereafter referred to as 1st week) The values increased to704 (100 mgdl glucose) and 994 (300 mgdl glucose) at week 4 after staticincubation Compared to the 1st week results the basal and stimulated insulinsecretion of the control group dropped to 61 and 51 after 4 weeks culturerespectively In case of the 025 mM Hb-C containing group these values were 86and 78 for basal and glucose-stimulated conditions respectively The stimulationindex (ratio of glucose-stimulated to basal insulin secretion) also showed a similartrend Enhanced insulin secretion (154 times) by glucose stimulation in the controlgroup was decreased to 128 times after 4 weeks culture On the other hand thestimulation indexes of 025 mM Hb-C containing group were 154 and 150 for 1stand 4th week culture respectively

The cell viability data obtained by MTS assay also showed a similar tendencyfor the incorporated Hb-C as shown in Fig 4 The group containing 025 mM Hb-Cshowed the best cell viability after 1 and 4 weeks culture Compared to the 1st weekvalue 825 relative viability was obtained in 025 mM Hb-C group after 4 weeksculture whereas only 376 cells maintained were viable after 4 weeks withoutHb-C

Figure 4 Effect of Hb-C concentration on islet viability Islets (1000 islets per pouch) wereencapsulated with different concentrations of Hb-C and the viability was measured by MTS assayafter different time (3 days and 4 weeks after encapsulation) Data are mean plusmn SD (n = 5) Statisticaldifferences between 1st week and 4th week data were calculated using unpaired Studentrsquos t-tests(lowastP lt 005 lowastlowastP lt 001)

1530 Y Y Kim et al

Optimization of islet density for macroencapsulation

The number of islets per macrocapsule device also strongly affected the insulin-secreting function and viability of islets Basal (100 mgdl) and glucose-stimulated(300 mgdl glucose) insulin secretion functions of the control group (without Hb-C)and Hb-C group with different islet density are shown in Fig 5A and 5B Thebasal and glucose insulin secretion tests were performed on day 3 and week 4after encapsulation The tests performed on day 3 after encapsulation did not showa significant dependence on number of islets encapsulated Therefore the dataobtained with 1000 isletsdevice were employed as control

In the case of the experimental control group (without Hb-C) a dramaticdecrease of basal and glucose-stimulated insulin secretion was observed with 2000isletsdevice group (Fig 5A) Compared with control (1st week data) only 81and 150 insulin was secreted by basal and high glucose stimulation respectivelyWhen decreasing islet density to 1000 and 500 isletsdevice the insulin-secretionfunctions were enhanced The 1000 isletsdevice group showed 635 and 630of relative insulin secretion by basal and high glucose stimulation respectivelyAlthough slightly decreased insulin secretion was observed with 500 isletsdevicegroup compared to the 1000 isletsdevice group they did not show statisticaldifferences between the two groups

Introduction of the Hb-C into the macrocapsule device significantly enhancedthe insulin-secretion function measured at 4 weeks after static culture The threeHb-C groups showed a level of insulin secretion superior to the experimentalcontrol groups (without Hb-C) In case of the 2000 isletsdevice group 650and 730 relative (in comparison with 1st week control) insulin secretion wasobserved by basal and high glucose stimulation respectively The situations wereenhanced by decreasing the islet density to 500 and 1000 isletsdevice More than80 of initial insulin secretion was measured in the 1000 and 500 isletsdevicegroup independent of islet density and glucose concentration Reduced cellviability with rising islet density and enhanced cell viability by incorporation ofHb-C into islet macrocapsules data were obtained from cell viability tests usingMTS reagents (Fig 6) In the case of Hb-C-containing groups the relative cellviabilities (compared to week 1 data of 1000 isletspouch) were maintained at morethan 80 in the 1000 and 500 isletsdevice groups after 4 weeks static cultureSlightly reduced cell viability (relative viability of 655) was observed in the2000 isletsdevice group In the case of control groups (without Hb-C) remarkabledifferences in cell viability were observed The relative cell viabilities of around50 for the 1000 and 500 isletsdevice groups dropped to 225 by increasing theislet density to 2000 isletsdevice

Cell viability staining using AOPI showed similar trends As shown in Fig 7 theretrieved islets from the Hb-C-containing groups showed a green colored viablecell state A negligible difference of cell staining was observed between 1000and 500 isletsdevice Increasing the islet density to 2000 isletsdevice showed anegative effect on cell viability because of a slight development of central necrosis

Enhanced function of islet macrocapsule 1531

Figure 5 Static insulin secretion of macroencapsulated islets cultured in medium with basal andhigh glucose levels (A) Without Hb-C (B) with 025 mM Hb-C Macrocapsules with different isletdensity (2000 1000 and 500 isletsdevice) were incubated with 6 ml RPMI-1640 medium for 4 h andsecreted insulin contents were measured by insulin radioimmunoassay The unit of G is mgdl Thedata are expressed as mean plusmn SD (n = 4) Statistical differences between 1st week and 4th week datawere calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

1532 Y Y Kim et al

Figure 6 Cell viability of macroencapsulated islets with different islet density after 4 weeks staticculture Data are shown as meanplusmnSD (n = 4) Statistical differences between 1st week and 4th weekdata were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

as shown in Fig 7 However the results for the control groups (without Hb-C)showed a remarkable difference between the 2000 and 1000 isletsdevice groupsThe islets from the 2000 isletsdevice group showed a typical red colored deadcell staining with destruction of integrated islet structure due to severe cell necrosisandor apoptosis Compared to the 2000 isletsdevice group slightly enhanced cellviability but with large necrotic core and viable rim of islets was observed in the1000 isletsdevice group as shown in Fig 7

DISCUSSION

In this study we investigated the optimum conditions of Hb-C concentration andislet density for islet macrocapsules Islets were immobilized by a thermore-versible polymeric ECM Among several devices proposed for islet immunoisola-tion macrocapsule devices possess advantages such as ease of implantation and ex-plantation with minimal surgical risk variety of implantation site and reduced trans-plantation volume by increasing islet density The thermosensitive polymeric ECMwhich showed a unique sol-to-gel transition may allow one to replace degeneratedislets after graft failure in the macroencapsulation devices by fresh islets by slightlydecreasing the temperature without surgical operations [6 19ndash21] Moreover the

Enhanced function of islet macrocapsule 1533

Figure 7 Fluorescence images of islets in the macrocapsules after 4 weeks of culture (A) 2000isletspouch with 025 mM Hb-C (B) 2000 isletspouch without Hb-C (C) 1000 isletspouch with025 mM Hb-C and (D) 1000 isletspouch without Hb-C

introduction of a bioactive ingredient (Hb-C) may enhance the cell functionality andviability as shown in our previous islet microencapsulation system [18 25]

As shown in Figs 3 and 4 insulin-secreting functions and cell viability weregreatly affected by the incorporated Hb-C in a concentration-dependent fashionThe optimum concentration of Hb-C as obtained from the experiments was025 mM (based on only Hb) The enhancing effect of the incorporated Hb-C oncell functionality and viability could be explained by the oxygen supply and otherreported bioactivities such as radical scavenging buffering and cell-stabilizingeffect [15 18 25] Most importantly increased oxygen supply may play acritical role in the enhanced cell functionality and viability Many studies havefocused on the limited oxygen supply on isolated and encapsulated islets Insulinsecretion especially glucose-stimulated insulin secretion was greatly affected byoxygen partial pressure [9] and the malfunction of islets or insulinoma cellswas overcome by increasing oxygen supply using hemoglobin Hb-C and waterelectrohydrolysis [15 16 18] However increasing the Hb-C concentration overthe optimum concentration resulted in decreased functionality and viability ofencapsulated islets as shown in Figs 3 and 4 The phenomena may be attributed todeteriorated mass transport by increased viscosity andor density of the polymericECM Although the polymeric ECM itself did not act as a diffusional barrier forsmall molecules such as insulin nutrients and wastes [29] a high concentrationof Hb-C in polymeric ECM may act as a mass transport barrier which resulted inreduced cell functionality and viability

Size or volume issue of a BAP is another important factor for its successfulapplication In the case of islet macroencapsulation the volume issue could be

1534 Y Y Kim et al

overcome by increasing islet density However cell viability is significantly affectedby cell density Figures 6 and 7 clearly show the decreased cell viability byincreasing islet density The enhanced cell viability by incorporation of Hb-Cwas caused by Hb-C groups however a high islet density (2000 isletsdevice)reduced viability by developing central necrosis By increasing islet density strongcompetition between encapsulated islets for nutrients and oxygen and seriousaccumulation of waste may occur in the device which results in reduced cellviability by increasing cell necrosis andor apoptosis On the other hand decreasingthe islet density increases the device volume which may evoke the volume issue

CONCLUSIONS

In this report we investigated the effect of Hb-C on the functionality and viabilityof an islet macroencapsulation device with a thermally reversible polymeric ECMBy introduction of 025 mM Hb-C into the islet macrocapsule more than 100enhanced insulin secretion and cell viability were observed after 4 weeks cultureFurthermore increase of islet density up to 1000 isletsdevice (154 microl volume)without any negative effect on insulin secretion and viability was possible byintroduction of 025 mM Hb into islet macrocapsules Although the practical in vivostudy remains as further study the incorporation of Hb-C into islet macrocapsulemight prolong the life span of encapsulated islets and reduce the implant volume byincreasing the functionality of islets and islet density

Acknowledgements

This work was supported by NIH DK56884 and BK 21 program (South Korea)

REFERENCES

1 G Reach Diabet Med 10 105 (1993)2 R P Lanza and W L Chick Ann N Y Acad Sci 831 323 (1997)3 R Calafiore Diabet Metab Rev 14 315 (1998)4 M K Lee and Y H Bae Adv Drug Deliv Rev 42 103 (2000)5 C K Colton and E S Avgoustiniators J Biomech Eng 113 152 (1992)6 Q P Hou and Y H Bae Adv Drug Deliv Rev 35 271 (1999)7 D Brandhorst H Brandhorst A Zwolinski F Nahidi and R G Bretzel Transplantation 71

179 (2001)8 J C Hutton and W J Malaisse Diabetologia 18 395 (1980)9 K E Dionne C K Colton and M L Yarmush Diabetes 42 12 (1993)

10 J Schrezenmeir L Gero M Solhdju J Kirchgessner C Laue J Beyer H Stier and W Muller-Klieser Transplant Proc 26 809 (1994)

11 M Ohta D Nelson M D Meglasson and M Erecinska J Biol Chem 265 17525 (1990)12 J S Whang S Y Chae M K Lee and Y H Bae Biomaterials 19 1198 (1998)13 K H Park S W Kim and Y H Bae J Biomed Mater Res 55 72 (2001)14 H Gappa M Baudys J J Koh S W Kim and Y H Bae Tissue Eng 7 35 (2001)

Enhanced function of islet macrocapsule 1527

Figure 2 Characterization of PEG-Hb conjugate using SDSndashPAGE (12 gel) Lane 1 standardmarkers lane 2 isolated Hb lane 3 after conjugation lane 4 purified Hb-C lane 5 ultrafiltrationfiltrate

and temperature-dependent solndashgel transition phenomena without hysterisis asreported by Han et al [19] The solndashgel transition temperature of the 7 aqueoussolution was 33C After double spin-coating a smooth dense and permeability-selective thin poly(HEMA-MMA-MAA) hydrogel top layer was anchored on thetop layer of a porous Duroporereg membrane SEM results showed that the thicknessof the coating layer was less than 5 microm and 30ndash60 of the coating layer anchoredthrough the Duroporereg membrane (data not shown) Serum and protein mixturepermeability test showed that the MWCO of the spin-coated membrane was about70 times 103 and IgG and other immunoglobulin were completely prevented frompenetrating through the membrane [27]

Optimization of Hb-C concentration for macroencapsulation

Hb-C-concentration-dependent glucose stimulation in insulin secretion and cellviability were observed with macroencapsulated rat islets (1000 islets per pouch154 microl loading volume) As demonstrated in Fig 3A and 3B the basal andglucose-stimulated insulin secretion functions of macroencapsulated islets werepreserved by the incorporation of Hb-C The insulin-secreting functions wereincreased with increasing Hb-C concentration up to 025 mM and then slightlydecreased with a higher Hb-C concentration (05 mM Hb-C) Compared to control(macroencapsulated islets without Hb-C) 213 and 303 enhanced insulin

1528 Y Y Kim et al

Figure 3 Static insulin secretion of macroencapsulated islets at basal and stimulated glucoseconcentration (A) Basal glucose at 100 mgdl (B) stimulated glucose at 300 mgdl in RPMI-1640medium Macrocapsules (1000 islets per pouch) were incubated with 6 ml RPMI-1640 medium for4 h The results are expressed as mean plusmn SD (n = 5) Statistical differences between 1st week and4th week data were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

Enhanced function of islet macrocapsule 1529

secretion was observed in macrocapsules containing 025 mM Hb-C for basal(100 mgdl) and high glucose (300 mgdl) stimulation respectively on day 3after encapsulation (hereafter referred to as 1st week) The values increased to704 (100 mgdl glucose) and 994 (300 mgdl glucose) at week 4 after staticincubation Compared to the 1st week results the basal and stimulated insulinsecretion of the control group dropped to 61 and 51 after 4 weeks culturerespectively In case of the 025 mM Hb-C containing group these values were 86and 78 for basal and glucose-stimulated conditions respectively The stimulationindex (ratio of glucose-stimulated to basal insulin secretion) also showed a similartrend Enhanced insulin secretion (154 times) by glucose stimulation in the controlgroup was decreased to 128 times after 4 weeks culture On the other hand thestimulation indexes of 025 mM Hb-C containing group were 154 and 150 for 1stand 4th week culture respectively

The cell viability data obtained by MTS assay also showed a similar tendencyfor the incorporated Hb-C as shown in Fig 4 The group containing 025 mM Hb-Cshowed the best cell viability after 1 and 4 weeks culture Compared to the 1st weekvalue 825 relative viability was obtained in 025 mM Hb-C group after 4 weeksculture whereas only 376 cells maintained were viable after 4 weeks withoutHb-C

Figure 4 Effect of Hb-C concentration on islet viability Islets (1000 islets per pouch) wereencapsulated with different concentrations of Hb-C and the viability was measured by MTS assayafter different time (3 days and 4 weeks after encapsulation) Data are mean plusmn SD (n = 5) Statisticaldifferences between 1st week and 4th week data were calculated using unpaired Studentrsquos t-tests(lowastP lt 005 lowastlowastP lt 001)

1530 Y Y Kim et al

Optimization of islet density for macroencapsulation

The number of islets per macrocapsule device also strongly affected the insulin-secreting function and viability of islets Basal (100 mgdl) and glucose-stimulated(300 mgdl glucose) insulin secretion functions of the control group (without Hb-C)and Hb-C group with different islet density are shown in Fig 5A and 5B Thebasal and glucose insulin secretion tests were performed on day 3 and week 4after encapsulation The tests performed on day 3 after encapsulation did not showa significant dependence on number of islets encapsulated Therefore the dataobtained with 1000 isletsdevice were employed as control

In the case of the experimental control group (without Hb-C) a dramaticdecrease of basal and glucose-stimulated insulin secretion was observed with 2000isletsdevice group (Fig 5A) Compared with control (1st week data) only 81and 150 insulin was secreted by basal and high glucose stimulation respectivelyWhen decreasing islet density to 1000 and 500 isletsdevice the insulin-secretionfunctions were enhanced The 1000 isletsdevice group showed 635 and 630of relative insulin secretion by basal and high glucose stimulation respectivelyAlthough slightly decreased insulin secretion was observed with 500 isletsdevicegroup compared to the 1000 isletsdevice group they did not show statisticaldifferences between the two groups

Introduction of the Hb-C into the macrocapsule device significantly enhancedthe insulin-secretion function measured at 4 weeks after static culture The threeHb-C groups showed a level of insulin secretion superior to the experimentalcontrol groups (without Hb-C) In case of the 2000 isletsdevice group 650and 730 relative (in comparison with 1st week control) insulin secretion wasobserved by basal and high glucose stimulation respectively The situations wereenhanced by decreasing the islet density to 500 and 1000 isletsdevice More than80 of initial insulin secretion was measured in the 1000 and 500 isletsdevicegroup independent of islet density and glucose concentration Reduced cellviability with rising islet density and enhanced cell viability by incorporation ofHb-C into islet macrocapsules data were obtained from cell viability tests usingMTS reagents (Fig 6) In the case of Hb-C-containing groups the relative cellviabilities (compared to week 1 data of 1000 isletspouch) were maintained at morethan 80 in the 1000 and 500 isletsdevice groups after 4 weeks static cultureSlightly reduced cell viability (relative viability of 655) was observed in the2000 isletsdevice group In the case of control groups (without Hb-C) remarkabledifferences in cell viability were observed The relative cell viabilities of around50 for the 1000 and 500 isletsdevice groups dropped to 225 by increasing theislet density to 2000 isletsdevice

Cell viability staining using AOPI showed similar trends As shown in Fig 7 theretrieved islets from the Hb-C-containing groups showed a green colored viablecell state A negligible difference of cell staining was observed between 1000and 500 isletsdevice Increasing the islet density to 2000 isletsdevice showed anegative effect on cell viability because of a slight development of central necrosis

Enhanced function of islet macrocapsule 1531

Figure 5 Static insulin secretion of macroencapsulated islets cultured in medium with basal andhigh glucose levels (A) Without Hb-C (B) with 025 mM Hb-C Macrocapsules with different isletdensity (2000 1000 and 500 isletsdevice) were incubated with 6 ml RPMI-1640 medium for 4 h andsecreted insulin contents were measured by insulin radioimmunoassay The unit of G is mgdl Thedata are expressed as mean plusmn SD (n = 4) Statistical differences between 1st week and 4th week datawere calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

1532 Y Y Kim et al

Figure 6 Cell viability of macroencapsulated islets with different islet density after 4 weeks staticculture Data are shown as meanplusmnSD (n = 4) Statistical differences between 1st week and 4th weekdata were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

as shown in Fig 7 However the results for the control groups (without Hb-C)showed a remarkable difference between the 2000 and 1000 isletsdevice groupsThe islets from the 2000 isletsdevice group showed a typical red colored deadcell staining with destruction of integrated islet structure due to severe cell necrosisandor apoptosis Compared to the 2000 isletsdevice group slightly enhanced cellviability but with large necrotic core and viable rim of islets was observed in the1000 isletsdevice group as shown in Fig 7

DISCUSSION

In this study we investigated the optimum conditions of Hb-C concentration andislet density for islet macrocapsules Islets were immobilized by a thermore-versible polymeric ECM Among several devices proposed for islet immunoisola-tion macrocapsule devices possess advantages such as ease of implantation and ex-plantation with minimal surgical risk variety of implantation site and reduced trans-plantation volume by increasing islet density The thermosensitive polymeric ECMwhich showed a unique sol-to-gel transition may allow one to replace degeneratedislets after graft failure in the macroencapsulation devices by fresh islets by slightlydecreasing the temperature without surgical operations [6 19ndash21] Moreover the

Enhanced function of islet macrocapsule 1533

Figure 7 Fluorescence images of islets in the macrocapsules after 4 weeks of culture (A) 2000isletspouch with 025 mM Hb-C (B) 2000 isletspouch without Hb-C (C) 1000 isletspouch with025 mM Hb-C and (D) 1000 isletspouch without Hb-C

introduction of a bioactive ingredient (Hb-C) may enhance the cell functionality andviability as shown in our previous islet microencapsulation system [18 25]

As shown in Figs 3 and 4 insulin-secreting functions and cell viability weregreatly affected by the incorporated Hb-C in a concentration-dependent fashionThe optimum concentration of Hb-C as obtained from the experiments was025 mM (based on only Hb) The enhancing effect of the incorporated Hb-C oncell functionality and viability could be explained by the oxygen supply and otherreported bioactivities such as radical scavenging buffering and cell-stabilizingeffect [15 18 25] Most importantly increased oxygen supply may play acritical role in the enhanced cell functionality and viability Many studies havefocused on the limited oxygen supply on isolated and encapsulated islets Insulinsecretion especially glucose-stimulated insulin secretion was greatly affected byoxygen partial pressure [9] and the malfunction of islets or insulinoma cellswas overcome by increasing oxygen supply using hemoglobin Hb-C and waterelectrohydrolysis [15 16 18] However increasing the Hb-C concentration overthe optimum concentration resulted in decreased functionality and viability ofencapsulated islets as shown in Figs 3 and 4 The phenomena may be attributed todeteriorated mass transport by increased viscosity andor density of the polymericECM Although the polymeric ECM itself did not act as a diffusional barrier forsmall molecules such as insulin nutrients and wastes [29] a high concentrationof Hb-C in polymeric ECM may act as a mass transport barrier which resulted inreduced cell functionality and viability

Size or volume issue of a BAP is another important factor for its successfulapplication In the case of islet macroencapsulation the volume issue could be

1534 Y Y Kim et al

overcome by increasing islet density However cell viability is significantly affectedby cell density Figures 6 and 7 clearly show the decreased cell viability byincreasing islet density The enhanced cell viability by incorporation of Hb-Cwas caused by Hb-C groups however a high islet density (2000 isletsdevice)reduced viability by developing central necrosis By increasing islet density strongcompetition between encapsulated islets for nutrients and oxygen and seriousaccumulation of waste may occur in the device which results in reduced cellviability by increasing cell necrosis andor apoptosis On the other hand decreasingthe islet density increases the device volume which may evoke the volume issue

CONCLUSIONS

In this report we investigated the effect of Hb-C on the functionality and viabilityof an islet macroencapsulation device with a thermally reversible polymeric ECMBy introduction of 025 mM Hb-C into the islet macrocapsule more than 100enhanced insulin secretion and cell viability were observed after 4 weeks cultureFurthermore increase of islet density up to 1000 isletsdevice (154 microl volume)without any negative effect on insulin secretion and viability was possible byintroduction of 025 mM Hb into islet macrocapsules Although the practical in vivostudy remains as further study the incorporation of Hb-C into islet macrocapsulemight prolong the life span of encapsulated islets and reduce the implant volume byincreasing the functionality of islets and islet density

Acknowledgements

This work was supported by NIH DK56884 and BK 21 program (South Korea)

REFERENCES

1 G Reach Diabet Med 10 105 (1993)2 R P Lanza and W L Chick Ann N Y Acad Sci 831 323 (1997)3 R Calafiore Diabet Metab Rev 14 315 (1998)4 M K Lee and Y H Bae Adv Drug Deliv Rev 42 103 (2000)5 C K Colton and E S Avgoustiniators J Biomech Eng 113 152 (1992)6 Q P Hou and Y H Bae Adv Drug Deliv Rev 35 271 (1999)7 D Brandhorst H Brandhorst A Zwolinski F Nahidi and R G Bretzel Transplantation 71

179 (2001)8 J C Hutton and W J Malaisse Diabetologia 18 395 (1980)9 K E Dionne C K Colton and M L Yarmush Diabetes 42 12 (1993)

10 J Schrezenmeir L Gero M Solhdju J Kirchgessner C Laue J Beyer H Stier and W Muller-Klieser Transplant Proc 26 809 (1994)

11 M Ohta D Nelson M D Meglasson and M Erecinska J Biol Chem 265 17525 (1990)12 J S Whang S Y Chae M K Lee and Y H Bae Biomaterials 19 1198 (1998)13 K H Park S W Kim and Y H Bae J Biomed Mater Res 55 72 (2001)14 H Gappa M Baudys J J Koh S W Kim and Y H Bae Tissue Eng 7 35 (2001)

1528 Y Y Kim et al

Figure 3 Static insulin secretion of macroencapsulated islets at basal and stimulated glucoseconcentration (A) Basal glucose at 100 mgdl (B) stimulated glucose at 300 mgdl in RPMI-1640medium Macrocapsules (1000 islets per pouch) were incubated with 6 ml RPMI-1640 medium for4 h The results are expressed as mean plusmn SD (n = 5) Statistical differences between 1st week and4th week data were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

Enhanced function of islet macrocapsule 1529

secretion was observed in macrocapsules containing 025 mM Hb-C for basal(100 mgdl) and high glucose (300 mgdl) stimulation respectively on day 3after encapsulation (hereafter referred to as 1st week) The values increased to704 (100 mgdl glucose) and 994 (300 mgdl glucose) at week 4 after staticincubation Compared to the 1st week results the basal and stimulated insulinsecretion of the control group dropped to 61 and 51 after 4 weeks culturerespectively In case of the 025 mM Hb-C containing group these values were 86and 78 for basal and glucose-stimulated conditions respectively The stimulationindex (ratio of glucose-stimulated to basal insulin secretion) also showed a similartrend Enhanced insulin secretion (154 times) by glucose stimulation in the controlgroup was decreased to 128 times after 4 weeks culture On the other hand thestimulation indexes of 025 mM Hb-C containing group were 154 and 150 for 1stand 4th week culture respectively

The cell viability data obtained by MTS assay also showed a similar tendencyfor the incorporated Hb-C as shown in Fig 4 The group containing 025 mM Hb-Cshowed the best cell viability after 1 and 4 weeks culture Compared to the 1st weekvalue 825 relative viability was obtained in 025 mM Hb-C group after 4 weeksculture whereas only 376 cells maintained were viable after 4 weeks withoutHb-C

Figure 4 Effect of Hb-C concentration on islet viability Islets (1000 islets per pouch) wereencapsulated with different concentrations of Hb-C and the viability was measured by MTS assayafter different time (3 days and 4 weeks after encapsulation) Data are mean plusmn SD (n = 5) Statisticaldifferences between 1st week and 4th week data were calculated using unpaired Studentrsquos t-tests(lowastP lt 005 lowastlowastP lt 001)

1530 Y Y Kim et al

Optimization of islet density for macroencapsulation

The number of islets per macrocapsule device also strongly affected the insulin-secreting function and viability of islets Basal (100 mgdl) and glucose-stimulated(300 mgdl glucose) insulin secretion functions of the control group (without Hb-C)and Hb-C group with different islet density are shown in Fig 5A and 5B Thebasal and glucose insulin secretion tests were performed on day 3 and week 4after encapsulation The tests performed on day 3 after encapsulation did not showa significant dependence on number of islets encapsulated Therefore the dataobtained with 1000 isletsdevice were employed as control

In the case of the experimental control group (without Hb-C) a dramaticdecrease of basal and glucose-stimulated insulin secretion was observed with 2000isletsdevice group (Fig 5A) Compared with control (1st week data) only 81and 150 insulin was secreted by basal and high glucose stimulation respectivelyWhen decreasing islet density to 1000 and 500 isletsdevice the insulin-secretionfunctions were enhanced The 1000 isletsdevice group showed 635 and 630of relative insulin secretion by basal and high glucose stimulation respectivelyAlthough slightly decreased insulin secretion was observed with 500 isletsdevicegroup compared to the 1000 isletsdevice group they did not show statisticaldifferences between the two groups

Introduction of the Hb-C into the macrocapsule device significantly enhancedthe insulin-secretion function measured at 4 weeks after static culture The threeHb-C groups showed a level of insulin secretion superior to the experimentalcontrol groups (without Hb-C) In case of the 2000 isletsdevice group 650and 730 relative (in comparison with 1st week control) insulin secretion wasobserved by basal and high glucose stimulation respectively The situations wereenhanced by decreasing the islet density to 500 and 1000 isletsdevice More than80 of initial insulin secretion was measured in the 1000 and 500 isletsdevicegroup independent of islet density and glucose concentration Reduced cellviability with rising islet density and enhanced cell viability by incorporation ofHb-C into islet macrocapsules data were obtained from cell viability tests usingMTS reagents (Fig 6) In the case of Hb-C-containing groups the relative cellviabilities (compared to week 1 data of 1000 isletspouch) were maintained at morethan 80 in the 1000 and 500 isletsdevice groups after 4 weeks static cultureSlightly reduced cell viability (relative viability of 655) was observed in the2000 isletsdevice group In the case of control groups (without Hb-C) remarkabledifferences in cell viability were observed The relative cell viabilities of around50 for the 1000 and 500 isletsdevice groups dropped to 225 by increasing theislet density to 2000 isletsdevice

Cell viability staining using AOPI showed similar trends As shown in Fig 7 theretrieved islets from the Hb-C-containing groups showed a green colored viablecell state A negligible difference of cell staining was observed between 1000and 500 isletsdevice Increasing the islet density to 2000 isletsdevice showed anegative effect on cell viability because of a slight development of central necrosis

Enhanced function of islet macrocapsule 1531

Figure 5 Static insulin secretion of macroencapsulated islets cultured in medium with basal andhigh glucose levels (A) Without Hb-C (B) with 025 mM Hb-C Macrocapsules with different isletdensity (2000 1000 and 500 isletsdevice) were incubated with 6 ml RPMI-1640 medium for 4 h andsecreted insulin contents were measured by insulin radioimmunoassay The unit of G is mgdl Thedata are expressed as mean plusmn SD (n = 4) Statistical differences between 1st week and 4th week datawere calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

1532 Y Y Kim et al

Figure 6 Cell viability of macroencapsulated islets with different islet density after 4 weeks staticculture Data are shown as meanplusmnSD (n = 4) Statistical differences between 1st week and 4th weekdata were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

as shown in Fig 7 However the results for the control groups (without Hb-C)showed a remarkable difference between the 2000 and 1000 isletsdevice groupsThe islets from the 2000 isletsdevice group showed a typical red colored deadcell staining with destruction of integrated islet structure due to severe cell necrosisandor apoptosis Compared to the 2000 isletsdevice group slightly enhanced cellviability but with large necrotic core and viable rim of islets was observed in the1000 isletsdevice group as shown in Fig 7

DISCUSSION

In this study we investigated the optimum conditions of Hb-C concentration andislet density for islet macrocapsules Islets were immobilized by a thermore-versible polymeric ECM Among several devices proposed for islet immunoisola-tion macrocapsule devices possess advantages such as ease of implantation and ex-plantation with minimal surgical risk variety of implantation site and reduced trans-plantation volume by increasing islet density The thermosensitive polymeric ECMwhich showed a unique sol-to-gel transition may allow one to replace degeneratedislets after graft failure in the macroencapsulation devices by fresh islets by slightlydecreasing the temperature without surgical operations [6 19ndash21] Moreover the

Enhanced function of islet macrocapsule 1533

Figure 7 Fluorescence images of islets in the macrocapsules after 4 weeks of culture (A) 2000isletspouch with 025 mM Hb-C (B) 2000 isletspouch without Hb-C (C) 1000 isletspouch with025 mM Hb-C and (D) 1000 isletspouch without Hb-C

introduction of a bioactive ingredient (Hb-C) may enhance the cell functionality andviability as shown in our previous islet microencapsulation system [18 25]

As shown in Figs 3 and 4 insulin-secreting functions and cell viability weregreatly affected by the incorporated Hb-C in a concentration-dependent fashionThe optimum concentration of Hb-C as obtained from the experiments was025 mM (based on only Hb) The enhancing effect of the incorporated Hb-C oncell functionality and viability could be explained by the oxygen supply and otherreported bioactivities such as radical scavenging buffering and cell-stabilizingeffect [15 18 25] Most importantly increased oxygen supply may play acritical role in the enhanced cell functionality and viability Many studies havefocused on the limited oxygen supply on isolated and encapsulated islets Insulinsecretion especially glucose-stimulated insulin secretion was greatly affected byoxygen partial pressure [9] and the malfunction of islets or insulinoma cellswas overcome by increasing oxygen supply using hemoglobin Hb-C and waterelectrohydrolysis [15 16 18] However increasing the Hb-C concentration overthe optimum concentration resulted in decreased functionality and viability ofencapsulated islets as shown in Figs 3 and 4 The phenomena may be attributed todeteriorated mass transport by increased viscosity andor density of the polymericECM Although the polymeric ECM itself did not act as a diffusional barrier forsmall molecules such as insulin nutrients and wastes [29] a high concentrationof Hb-C in polymeric ECM may act as a mass transport barrier which resulted inreduced cell functionality and viability

Size or volume issue of a BAP is another important factor for its successfulapplication In the case of islet macroencapsulation the volume issue could be

1534 Y Y Kim et al

overcome by increasing islet density However cell viability is significantly affectedby cell density Figures 6 and 7 clearly show the decreased cell viability byincreasing islet density The enhanced cell viability by incorporation of Hb-Cwas caused by Hb-C groups however a high islet density (2000 isletsdevice)reduced viability by developing central necrosis By increasing islet density strongcompetition between encapsulated islets for nutrients and oxygen and seriousaccumulation of waste may occur in the device which results in reduced cellviability by increasing cell necrosis andor apoptosis On the other hand decreasingthe islet density increases the device volume which may evoke the volume issue

CONCLUSIONS

In this report we investigated the effect of Hb-C on the functionality and viabilityof an islet macroencapsulation device with a thermally reversible polymeric ECMBy introduction of 025 mM Hb-C into the islet macrocapsule more than 100enhanced insulin secretion and cell viability were observed after 4 weeks cultureFurthermore increase of islet density up to 1000 isletsdevice (154 microl volume)without any negative effect on insulin secretion and viability was possible byintroduction of 025 mM Hb into islet macrocapsules Although the practical in vivostudy remains as further study the incorporation of Hb-C into islet macrocapsulemight prolong the life span of encapsulated islets and reduce the implant volume byincreasing the functionality of islets and islet density

Acknowledgements

This work was supported by NIH DK56884 and BK 21 program (South Korea)

REFERENCES

1 G Reach Diabet Med 10 105 (1993)2 R P Lanza and W L Chick Ann N Y Acad Sci 831 323 (1997)3 R Calafiore Diabet Metab Rev 14 315 (1998)4 M K Lee and Y H Bae Adv Drug Deliv Rev 42 103 (2000)5 C K Colton and E S Avgoustiniators J Biomech Eng 113 152 (1992)6 Q P Hou and Y H Bae Adv Drug Deliv Rev 35 271 (1999)7 D Brandhorst H Brandhorst A Zwolinski F Nahidi and R G Bretzel Transplantation 71

179 (2001)8 J C Hutton and W J Malaisse Diabetologia 18 395 (1980)9 K E Dionne C K Colton and M L Yarmush Diabetes 42 12 (1993)

10 J Schrezenmeir L Gero M Solhdju J Kirchgessner C Laue J Beyer H Stier and W Muller-Klieser Transplant Proc 26 809 (1994)

11 M Ohta D Nelson M D Meglasson and M Erecinska J Biol Chem 265 17525 (1990)12 J S Whang S Y Chae M K Lee and Y H Bae Biomaterials 19 1198 (1998)13 K H Park S W Kim and Y H Bae J Biomed Mater Res 55 72 (2001)14 H Gappa M Baudys J J Koh S W Kim and Y H Bae Tissue Eng 7 35 (2001)

Enhanced function of islet macrocapsule 1529

secretion was observed in macrocapsules containing 025 mM Hb-C for basal(100 mgdl) and high glucose (300 mgdl) stimulation respectively on day 3after encapsulation (hereafter referred to as 1st week) The values increased to704 (100 mgdl glucose) and 994 (300 mgdl glucose) at week 4 after staticincubation Compared to the 1st week results the basal and stimulated insulinsecretion of the control group dropped to 61 and 51 after 4 weeks culturerespectively In case of the 025 mM Hb-C containing group these values were 86and 78 for basal and glucose-stimulated conditions respectively The stimulationindex (ratio of glucose-stimulated to basal insulin secretion) also showed a similartrend Enhanced insulin secretion (154 times) by glucose stimulation in the controlgroup was decreased to 128 times after 4 weeks culture On the other hand thestimulation indexes of 025 mM Hb-C containing group were 154 and 150 for 1stand 4th week culture respectively

The cell viability data obtained by MTS assay also showed a similar tendencyfor the incorporated Hb-C as shown in Fig 4 The group containing 025 mM Hb-Cshowed the best cell viability after 1 and 4 weeks culture Compared to the 1st weekvalue 825 relative viability was obtained in 025 mM Hb-C group after 4 weeksculture whereas only 376 cells maintained were viable after 4 weeks withoutHb-C

Figure 4 Effect of Hb-C concentration on islet viability Islets (1000 islets per pouch) wereencapsulated with different concentrations of Hb-C and the viability was measured by MTS assayafter different time (3 days and 4 weeks after encapsulation) Data are mean plusmn SD (n = 5) Statisticaldifferences between 1st week and 4th week data were calculated using unpaired Studentrsquos t-tests(lowastP lt 005 lowastlowastP lt 001)

1530 Y Y Kim et al

Optimization of islet density for macroencapsulation

The number of islets per macrocapsule device also strongly affected the insulin-secreting function and viability of islets Basal (100 mgdl) and glucose-stimulated(300 mgdl glucose) insulin secretion functions of the control group (without Hb-C)and Hb-C group with different islet density are shown in Fig 5A and 5B Thebasal and glucose insulin secretion tests were performed on day 3 and week 4after encapsulation The tests performed on day 3 after encapsulation did not showa significant dependence on number of islets encapsulated Therefore the dataobtained with 1000 isletsdevice were employed as control

In the case of the experimental control group (without Hb-C) a dramaticdecrease of basal and glucose-stimulated insulin secretion was observed with 2000isletsdevice group (Fig 5A) Compared with control (1st week data) only 81and 150 insulin was secreted by basal and high glucose stimulation respectivelyWhen decreasing islet density to 1000 and 500 isletsdevice the insulin-secretionfunctions were enhanced The 1000 isletsdevice group showed 635 and 630of relative insulin secretion by basal and high glucose stimulation respectivelyAlthough slightly decreased insulin secretion was observed with 500 isletsdevicegroup compared to the 1000 isletsdevice group they did not show statisticaldifferences between the two groups

Introduction of the Hb-C into the macrocapsule device significantly enhancedthe insulin-secretion function measured at 4 weeks after static culture The threeHb-C groups showed a level of insulin secretion superior to the experimentalcontrol groups (without Hb-C) In case of the 2000 isletsdevice group 650and 730 relative (in comparison with 1st week control) insulin secretion wasobserved by basal and high glucose stimulation respectively The situations wereenhanced by decreasing the islet density to 500 and 1000 isletsdevice More than80 of initial insulin secretion was measured in the 1000 and 500 isletsdevicegroup independent of islet density and glucose concentration Reduced cellviability with rising islet density and enhanced cell viability by incorporation ofHb-C into islet macrocapsules data were obtained from cell viability tests usingMTS reagents (Fig 6) In the case of Hb-C-containing groups the relative cellviabilities (compared to week 1 data of 1000 isletspouch) were maintained at morethan 80 in the 1000 and 500 isletsdevice groups after 4 weeks static cultureSlightly reduced cell viability (relative viability of 655) was observed in the2000 isletsdevice group In the case of control groups (without Hb-C) remarkabledifferences in cell viability were observed The relative cell viabilities of around50 for the 1000 and 500 isletsdevice groups dropped to 225 by increasing theislet density to 2000 isletsdevice

Cell viability staining using AOPI showed similar trends As shown in Fig 7 theretrieved islets from the Hb-C-containing groups showed a green colored viablecell state A negligible difference of cell staining was observed between 1000and 500 isletsdevice Increasing the islet density to 2000 isletsdevice showed anegative effect on cell viability because of a slight development of central necrosis

Enhanced function of islet macrocapsule 1531

Figure 5 Static insulin secretion of macroencapsulated islets cultured in medium with basal andhigh glucose levels (A) Without Hb-C (B) with 025 mM Hb-C Macrocapsules with different isletdensity (2000 1000 and 500 isletsdevice) were incubated with 6 ml RPMI-1640 medium for 4 h andsecreted insulin contents were measured by insulin radioimmunoassay The unit of G is mgdl Thedata are expressed as mean plusmn SD (n = 4) Statistical differences between 1st week and 4th week datawere calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

1532 Y Y Kim et al

Figure 6 Cell viability of macroencapsulated islets with different islet density after 4 weeks staticculture Data are shown as meanplusmnSD (n = 4) Statistical differences between 1st week and 4th weekdata were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

as shown in Fig 7 However the results for the control groups (without Hb-C)showed a remarkable difference between the 2000 and 1000 isletsdevice groupsThe islets from the 2000 isletsdevice group showed a typical red colored deadcell staining with destruction of integrated islet structure due to severe cell necrosisandor apoptosis Compared to the 2000 isletsdevice group slightly enhanced cellviability but with large necrotic core and viable rim of islets was observed in the1000 isletsdevice group as shown in Fig 7

DISCUSSION

In this study we investigated the optimum conditions of Hb-C concentration andislet density for islet macrocapsules Islets were immobilized by a thermore-versible polymeric ECM Among several devices proposed for islet immunoisola-tion macrocapsule devices possess advantages such as ease of implantation and ex-plantation with minimal surgical risk variety of implantation site and reduced trans-plantation volume by increasing islet density The thermosensitive polymeric ECMwhich showed a unique sol-to-gel transition may allow one to replace degeneratedislets after graft failure in the macroencapsulation devices by fresh islets by slightlydecreasing the temperature without surgical operations [6 19ndash21] Moreover the

Enhanced function of islet macrocapsule 1533

Figure 7 Fluorescence images of islets in the macrocapsules after 4 weeks of culture (A) 2000isletspouch with 025 mM Hb-C (B) 2000 isletspouch without Hb-C (C) 1000 isletspouch with025 mM Hb-C and (D) 1000 isletspouch without Hb-C

introduction of a bioactive ingredient (Hb-C) may enhance the cell functionality andviability as shown in our previous islet microencapsulation system [18 25]

As shown in Figs 3 and 4 insulin-secreting functions and cell viability weregreatly affected by the incorporated Hb-C in a concentration-dependent fashionThe optimum concentration of Hb-C as obtained from the experiments was025 mM (based on only Hb) The enhancing effect of the incorporated Hb-C oncell functionality and viability could be explained by the oxygen supply and otherreported bioactivities such as radical scavenging buffering and cell-stabilizingeffect [15 18 25] Most importantly increased oxygen supply may play acritical role in the enhanced cell functionality and viability Many studies havefocused on the limited oxygen supply on isolated and encapsulated islets Insulinsecretion especially glucose-stimulated insulin secretion was greatly affected byoxygen partial pressure [9] and the malfunction of islets or insulinoma cellswas overcome by increasing oxygen supply using hemoglobin Hb-C and waterelectrohydrolysis [15 16 18] However increasing the Hb-C concentration overthe optimum concentration resulted in decreased functionality and viability ofencapsulated islets as shown in Figs 3 and 4 The phenomena may be attributed todeteriorated mass transport by increased viscosity andor density of the polymericECM Although the polymeric ECM itself did not act as a diffusional barrier forsmall molecules such as insulin nutrients and wastes [29] a high concentrationof Hb-C in polymeric ECM may act as a mass transport barrier which resulted inreduced cell functionality and viability

Size or volume issue of a BAP is another important factor for its successfulapplication In the case of islet macroencapsulation the volume issue could be

1534 Y Y Kim et al

overcome by increasing islet density However cell viability is significantly affectedby cell density Figures 6 and 7 clearly show the decreased cell viability byincreasing islet density The enhanced cell viability by incorporation of Hb-Cwas caused by Hb-C groups however a high islet density (2000 isletsdevice)reduced viability by developing central necrosis By increasing islet density strongcompetition between encapsulated islets for nutrients and oxygen and seriousaccumulation of waste may occur in the device which results in reduced cellviability by increasing cell necrosis andor apoptosis On the other hand decreasingthe islet density increases the device volume which may evoke the volume issue

CONCLUSIONS

In this report we investigated the effect of Hb-C on the functionality and viabilityof an islet macroencapsulation device with a thermally reversible polymeric ECMBy introduction of 025 mM Hb-C into the islet macrocapsule more than 100enhanced insulin secretion and cell viability were observed after 4 weeks cultureFurthermore increase of islet density up to 1000 isletsdevice (154 microl volume)without any negative effect on insulin secretion and viability was possible byintroduction of 025 mM Hb into islet macrocapsules Although the practical in vivostudy remains as further study the incorporation of Hb-C into islet macrocapsulemight prolong the life span of encapsulated islets and reduce the implant volume byincreasing the functionality of islets and islet density

Acknowledgements

This work was supported by NIH DK56884 and BK 21 program (South Korea)

REFERENCES

1 G Reach Diabet Med 10 105 (1993)2 R P Lanza and W L Chick Ann N Y Acad Sci 831 323 (1997)3 R Calafiore Diabet Metab Rev 14 315 (1998)4 M K Lee and Y H Bae Adv Drug Deliv Rev 42 103 (2000)5 C K Colton and E S Avgoustiniators J Biomech Eng 113 152 (1992)6 Q P Hou and Y H Bae Adv Drug Deliv Rev 35 271 (1999)7 D Brandhorst H Brandhorst A Zwolinski F Nahidi and R G Bretzel Transplantation 71

179 (2001)8 J C Hutton and W J Malaisse Diabetologia 18 395 (1980)9 K E Dionne C K Colton and M L Yarmush Diabetes 42 12 (1993)

10 J Schrezenmeir L Gero M Solhdju J Kirchgessner C Laue J Beyer H Stier and W Muller-Klieser Transplant Proc 26 809 (1994)

11 M Ohta D Nelson M D Meglasson and M Erecinska J Biol Chem 265 17525 (1990)12 J S Whang S Y Chae M K Lee and Y H Bae Biomaterials 19 1198 (1998)13 K H Park S W Kim and Y H Bae J Biomed Mater Res 55 72 (2001)14 H Gappa M Baudys J J Koh S W Kim and Y H Bae Tissue Eng 7 35 (2001)

1530 Y Y Kim et al

Optimization of islet density for macroencapsulation

The number of islets per macrocapsule device also strongly affected the insulin-secreting function and viability of islets Basal (100 mgdl) and glucose-stimulated(300 mgdl glucose) insulin secretion functions of the control group (without Hb-C)and Hb-C group with different islet density are shown in Fig 5A and 5B Thebasal and glucose insulin secretion tests were performed on day 3 and week 4after encapsulation The tests performed on day 3 after encapsulation did not showa significant dependence on number of islets encapsulated Therefore the dataobtained with 1000 isletsdevice were employed as control

In the case of the experimental control group (without Hb-C) a dramaticdecrease of basal and glucose-stimulated insulin secretion was observed with 2000isletsdevice group (Fig 5A) Compared with control (1st week data) only 81and 150 insulin was secreted by basal and high glucose stimulation respectivelyWhen decreasing islet density to 1000 and 500 isletsdevice the insulin-secretionfunctions were enhanced The 1000 isletsdevice group showed 635 and 630of relative insulin secretion by basal and high glucose stimulation respectivelyAlthough slightly decreased insulin secretion was observed with 500 isletsdevicegroup compared to the 1000 isletsdevice group they did not show statisticaldifferences between the two groups

Introduction of the Hb-C into the macrocapsule device significantly enhancedthe insulin-secretion function measured at 4 weeks after static culture The threeHb-C groups showed a level of insulin secretion superior to the experimentalcontrol groups (without Hb-C) In case of the 2000 isletsdevice group 650and 730 relative (in comparison with 1st week control) insulin secretion wasobserved by basal and high glucose stimulation respectively The situations wereenhanced by decreasing the islet density to 500 and 1000 isletsdevice More than80 of initial insulin secretion was measured in the 1000 and 500 isletsdevicegroup independent of islet density and glucose concentration Reduced cellviability with rising islet density and enhanced cell viability by incorporation ofHb-C into islet macrocapsules data were obtained from cell viability tests usingMTS reagents (Fig 6) In the case of Hb-C-containing groups the relative cellviabilities (compared to week 1 data of 1000 isletspouch) were maintained at morethan 80 in the 1000 and 500 isletsdevice groups after 4 weeks static cultureSlightly reduced cell viability (relative viability of 655) was observed in the2000 isletsdevice group In the case of control groups (without Hb-C) remarkabledifferences in cell viability were observed The relative cell viabilities of around50 for the 1000 and 500 isletsdevice groups dropped to 225 by increasing theislet density to 2000 isletsdevice

Cell viability staining using AOPI showed similar trends As shown in Fig 7 theretrieved islets from the Hb-C-containing groups showed a green colored viablecell state A negligible difference of cell staining was observed between 1000and 500 isletsdevice Increasing the islet density to 2000 isletsdevice showed anegative effect on cell viability because of a slight development of central necrosis

Enhanced function of islet macrocapsule 1531

Figure 5 Static insulin secretion of macroencapsulated islets cultured in medium with basal andhigh glucose levels (A) Without Hb-C (B) with 025 mM Hb-C Macrocapsules with different isletdensity (2000 1000 and 500 isletsdevice) were incubated with 6 ml RPMI-1640 medium for 4 h andsecreted insulin contents were measured by insulin radioimmunoassay The unit of G is mgdl Thedata are expressed as mean plusmn SD (n = 4) Statistical differences between 1st week and 4th week datawere calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

1532 Y Y Kim et al

Figure 6 Cell viability of macroencapsulated islets with different islet density after 4 weeks staticculture Data are shown as meanplusmnSD (n = 4) Statistical differences between 1st week and 4th weekdata were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

as shown in Fig 7 However the results for the control groups (without Hb-C)showed a remarkable difference between the 2000 and 1000 isletsdevice groupsThe islets from the 2000 isletsdevice group showed a typical red colored deadcell staining with destruction of integrated islet structure due to severe cell necrosisandor apoptosis Compared to the 2000 isletsdevice group slightly enhanced cellviability but with large necrotic core and viable rim of islets was observed in the1000 isletsdevice group as shown in Fig 7

DISCUSSION

In this study we investigated the optimum conditions of Hb-C concentration andislet density for islet macrocapsules Islets were immobilized by a thermore-versible polymeric ECM Among several devices proposed for islet immunoisola-tion macrocapsule devices possess advantages such as ease of implantation and ex-plantation with minimal surgical risk variety of implantation site and reduced trans-plantation volume by increasing islet density The thermosensitive polymeric ECMwhich showed a unique sol-to-gel transition may allow one to replace degeneratedislets after graft failure in the macroencapsulation devices by fresh islets by slightlydecreasing the temperature without surgical operations [6 19ndash21] Moreover the

Enhanced function of islet macrocapsule 1533

Figure 7 Fluorescence images of islets in the macrocapsules after 4 weeks of culture (A) 2000isletspouch with 025 mM Hb-C (B) 2000 isletspouch without Hb-C (C) 1000 isletspouch with025 mM Hb-C and (D) 1000 isletspouch without Hb-C

introduction of a bioactive ingredient (Hb-C) may enhance the cell functionality andviability as shown in our previous islet microencapsulation system [18 25]

As shown in Figs 3 and 4 insulin-secreting functions and cell viability weregreatly affected by the incorporated Hb-C in a concentration-dependent fashionThe optimum concentration of Hb-C as obtained from the experiments was025 mM (based on only Hb) The enhancing effect of the incorporated Hb-C oncell functionality and viability could be explained by the oxygen supply and otherreported bioactivities such as radical scavenging buffering and cell-stabilizingeffect [15 18 25] Most importantly increased oxygen supply may play acritical role in the enhanced cell functionality and viability Many studies havefocused on the limited oxygen supply on isolated and encapsulated islets Insulinsecretion especially glucose-stimulated insulin secretion was greatly affected byoxygen partial pressure [9] and the malfunction of islets or insulinoma cellswas overcome by increasing oxygen supply using hemoglobin Hb-C and waterelectrohydrolysis [15 16 18] However increasing the Hb-C concentration overthe optimum concentration resulted in decreased functionality and viability ofencapsulated islets as shown in Figs 3 and 4 The phenomena may be attributed todeteriorated mass transport by increased viscosity andor density of the polymericECM Although the polymeric ECM itself did not act as a diffusional barrier forsmall molecules such as insulin nutrients and wastes [29] a high concentrationof Hb-C in polymeric ECM may act as a mass transport barrier which resulted inreduced cell functionality and viability

Size or volume issue of a BAP is another important factor for its successfulapplication In the case of islet macroencapsulation the volume issue could be

1534 Y Y Kim et al

overcome by increasing islet density However cell viability is significantly affectedby cell density Figures 6 and 7 clearly show the decreased cell viability byincreasing islet density The enhanced cell viability by incorporation of Hb-Cwas caused by Hb-C groups however a high islet density (2000 isletsdevice)reduced viability by developing central necrosis By increasing islet density strongcompetition between encapsulated islets for nutrients and oxygen and seriousaccumulation of waste may occur in the device which results in reduced cellviability by increasing cell necrosis andor apoptosis On the other hand decreasingthe islet density increases the device volume which may evoke the volume issue

CONCLUSIONS

In this report we investigated the effect of Hb-C on the functionality and viabilityof an islet macroencapsulation device with a thermally reversible polymeric ECMBy introduction of 025 mM Hb-C into the islet macrocapsule more than 100enhanced insulin secretion and cell viability were observed after 4 weeks cultureFurthermore increase of islet density up to 1000 isletsdevice (154 microl volume)without any negative effect on insulin secretion and viability was possible byintroduction of 025 mM Hb into islet macrocapsules Although the practical in vivostudy remains as further study the incorporation of Hb-C into islet macrocapsulemight prolong the life span of encapsulated islets and reduce the implant volume byincreasing the functionality of islets and islet density

Acknowledgements

This work was supported by NIH DK56884 and BK 21 program (South Korea)

REFERENCES

1 G Reach Diabet Med 10 105 (1993)2 R P Lanza and W L Chick Ann N Y Acad Sci 831 323 (1997)3 R Calafiore Diabet Metab Rev 14 315 (1998)4 M K Lee and Y H Bae Adv Drug Deliv Rev 42 103 (2000)5 C K Colton and E S Avgoustiniators J Biomech Eng 113 152 (1992)6 Q P Hou and Y H Bae Adv Drug Deliv Rev 35 271 (1999)7 D Brandhorst H Brandhorst A Zwolinski F Nahidi and R G Bretzel Transplantation 71

179 (2001)8 J C Hutton and W J Malaisse Diabetologia 18 395 (1980)9 K E Dionne C K Colton and M L Yarmush Diabetes 42 12 (1993)

10 J Schrezenmeir L Gero M Solhdju J Kirchgessner C Laue J Beyer H Stier and W Muller-Klieser Transplant Proc 26 809 (1994)

11 M Ohta D Nelson M D Meglasson and M Erecinska J Biol Chem 265 17525 (1990)12 J S Whang S Y Chae M K Lee and Y H Bae Biomaterials 19 1198 (1998)13 K H Park S W Kim and Y H Bae J Biomed Mater Res 55 72 (2001)14 H Gappa M Baudys J J Koh S W Kim and Y H Bae Tissue Eng 7 35 (2001)

Enhanced function of islet macrocapsule 1531

Figure 5 Static insulin secretion of macroencapsulated islets cultured in medium with basal andhigh glucose levels (A) Without Hb-C (B) with 025 mM Hb-C Macrocapsules with different isletdensity (2000 1000 and 500 isletsdevice) were incubated with 6 ml RPMI-1640 medium for 4 h andsecreted insulin contents were measured by insulin radioimmunoassay The unit of G is mgdl Thedata are expressed as mean plusmn SD (n = 4) Statistical differences between 1st week and 4th week datawere calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

1532 Y Y Kim et al

Figure 6 Cell viability of macroencapsulated islets with different islet density after 4 weeks staticculture Data are shown as meanplusmnSD (n = 4) Statistical differences between 1st week and 4th weekdata were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

as shown in Fig 7 However the results for the control groups (without Hb-C)showed a remarkable difference between the 2000 and 1000 isletsdevice groupsThe islets from the 2000 isletsdevice group showed a typical red colored deadcell staining with destruction of integrated islet structure due to severe cell necrosisandor apoptosis Compared to the 2000 isletsdevice group slightly enhanced cellviability but with large necrotic core and viable rim of islets was observed in the1000 isletsdevice group as shown in Fig 7

DISCUSSION

In this study we investigated the optimum conditions of Hb-C concentration andislet density for islet macrocapsules Islets were immobilized by a thermore-versible polymeric ECM Among several devices proposed for islet immunoisola-tion macrocapsule devices possess advantages such as ease of implantation and ex-plantation with minimal surgical risk variety of implantation site and reduced trans-plantation volume by increasing islet density The thermosensitive polymeric ECMwhich showed a unique sol-to-gel transition may allow one to replace degeneratedislets after graft failure in the macroencapsulation devices by fresh islets by slightlydecreasing the temperature without surgical operations [6 19ndash21] Moreover the

Enhanced function of islet macrocapsule 1533

Figure 7 Fluorescence images of islets in the macrocapsules after 4 weeks of culture (A) 2000isletspouch with 025 mM Hb-C (B) 2000 isletspouch without Hb-C (C) 1000 isletspouch with025 mM Hb-C and (D) 1000 isletspouch without Hb-C

introduction of a bioactive ingredient (Hb-C) may enhance the cell functionality andviability as shown in our previous islet microencapsulation system [18 25]

As shown in Figs 3 and 4 insulin-secreting functions and cell viability weregreatly affected by the incorporated Hb-C in a concentration-dependent fashionThe optimum concentration of Hb-C as obtained from the experiments was025 mM (based on only Hb) The enhancing effect of the incorporated Hb-C oncell functionality and viability could be explained by the oxygen supply and otherreported bioactivities such as radical scavenging buffering and cell-stabilizingeffect [15 18 25] Most importantly increased oxygen supply may play acritical role in the enhanced cell functionality and viability Many studies havefocused on the limited oxygen supply on isolated and encapsulated islets Insulinsecretion especially glucose-stimulated insulin secretion was greatly affected byoxygen partial pressure [9] and the malfunction of islets or insulinoma cellswas overcome by increasing oxygen supply using hemoglobin Hb-C and waterelectrohydrolysis [15 16 18] However increasing the Hb-C concentration overthe optimum concentration resulted in decreased functionality and viability ofencapsulated islets as shown in Figs 3 and 4 The phenomena may be attributed todeteriorated mass transport by increased viscosity andor density of the polymericECM Although the polymeric ECM itself did not act as a diffusional barrier forsmall molecules such as insulin nutrients and wastes [29] a high concentrationof Hb-C in polymeric ECM may act as a mass transport barrier which resulted inreduced cell functionality and viability

Size or volume issue of a BAP is another important factor for its successfulapplication In the case of islet macroencapsulation the volume issue could be

1534 Y Y Kim et al

overcome by increasing islet density However cell viability is significantly affectedby cell density Figures 6 and 7 clearly show the decreased cell viability byincreasing islet density The enhanced cell viability by incorporation of Hb-Cwas caused by Hb-C groups however a high islet density (2000 isletsdevice)reduced viability by developing central necrosis By increasing islet density strongcompetition between encapsulated islets for nutrients and oxygen and seriousaccumulation of waste may occur in the device which results in reduced cellviability by increasing cell necrosis andor apoptosis On the other hand decreasingthe islet density increases the device volume which may evoke the volume issue

CONCLUSIONS

In this report we investigated the effect of Hb-C on the functionality and viabilityof an islet macroencapsulation device with a thermally reversible polymeric ECMBy introduction of 025 mM Hb-C into the islet macrocapsule more than 100enhanced insulin secretion and cell viability were observed after 4 weeks cultureFurthermore increase of islet density up to 1000 isletsdevice (154 microl volume)without any negative effect on insulin secretion and viability was possible byintroduction of 025 mM Hb into islet macrocapsules Although the practical in vivostudy remains as further study the incorporation of Hb-C into islet macrocapsulemight prolong the life span of encapsulated islets and reduce the implant volume byincreasing the functionality of islets and islet density

Acknowledgements

This work was supported by NIH DK56884 and BK 21 program (South Korea)

REFERENCES

1 G Reach Diabet Med 10 105 (1993)2 R P Lanza and W L Chick Ann N Y Acad Sci 831 323 (1997)3 R Calafiore Diabet Metab Rev 14 315 (1998)4 M K Lee and Y H Bae Adv Drug Deliv Rev 42 103 (2000)5 C K Colton and E S Avgoustiniators J Biomech Eng 113 152 (1992)6 Q P Hou and Y H Bae Adv Drug Deliv Rev 35 271 (1999)7 D Brandhorst H Brandhorst A Zwolinski F Nahidi and R G Bretzel Transplantation 71

179 (2001)8 J C Hutton and W J Malaisse Diabetologia 18 395 (1980)9 K E Dionne C K Colton and M L Yarmush Diabetes 42 12 (1993)

10 J Schrezenmeir L Gero M Solhdju J Kirchgessner C Laue J Beyer H Stier and W Muller-Klieser Transplant Proc 26 809 (1994)

11 M Ohta D Nelson M D Meglasson and M Erecinska J Biol Chem 265 17525 (1990)12 J S Whang S Y Chae M K Lee and Y H Bae Biomaterials 19 1198 (1998)13 K H Park S W Kim and Y H Bae J Biomed Mater Res 55 72 (2001)14 H Gappa M Baudys J J Koh S W Kim and Y H Bae Tissue Eng 7 35 (2001)

1532 Y Y Kim et al

Figure 6 Cell viability of macroencapsulated islets with different islet density after 4 weeks staticculture Data are shown as meanplusmnSD (n = 4) Statistical differences between 1st week and 4th weekdata were calculated using unpaired Studentrsquos t-tests (lowastP lt 005 lowastlowastP lt 001)

as shown in Fig 7 However the results for the control groups (without Hb-C)showed a remarkable difference between the 2000 and 1000 isletsdevice groupsThe islets from the 2000 isletsdevice group showed a typical red colored deadcell staining with destruction of integrated islet structure due to severe cell necrosisandor apoptosis Compared to the 2000 isletsdevice group slightly enhanced cellviability but with large necrotic core and viable rim of islets was observed in the1000 isletsdevice group as shown in Fig 7

DISCUSSION

In this study we investigated the optimum conditions of Hb-C concentration andislet density for islet macrocapsules Islets were immobilized by a thermore-versible polymeric ECM Among several devices proposed for islet immunoisola-tion macrocapsule devices possess advantages such as ease of implantation and ex-plantation with minimal surgical risk variety of implantation site and reduced trans-plantation volume by increasing islet density The thermosensitive polymeric ECMwhich showed a unique sol-to-gel transition may allow one to replace degeneratedislets after graft failure in the macroencapsulation devices by fresh islets by slightlydecreasing the temperature without surgical operations [6 19ndash21] Moreover the

Enhanced function of islet macrocapsule 1533

Figure 7 Fluorescence images of islets in the macrocapsules after 4 weeks of culture (A) 2000isletspouch with 025 mM Hb-C (B) 2000 isletspouch without Hb-C (C) 1000 isletspouch with025 mM Hb-C and (D) 1000 isletspouch without Hb-C

introduction of a bioactive ingredient (Hb-C) may enhance the cell functionality andviability as shown in our previous islet microencapsulation system [18 25]

As shown in Figs 3 and 4 insulin-secreting functions and cell viability weregreatly affected by the incorporated Hb-C in a concentration-dependent fashionThe optimum concentration of Hb-C as obtained from the experiments was025 mM (based on only Hb) The enhancing effect of the incorporated Hb-C oncell functionality and viability could be explained by the oxygen supply and otherreported bioactivities such as radical scavenging buffering and cell-stabilizingeffect [15 18 25] Most importantly increased oxygen supply may play acritical role in the enhanced cell functionality and viability Many studies havefocused on the limited oxygen supply on isolated and encapsulated islets Insulinsecretion especially glucose-stimulated insulin secretion was greatly affected byoxygen partial pressure [9] and the malfunction of islets or insulinoma cellswas overcome by increasing oxygen supply using hemoglobin Hb-C and waterelectrohydrolysis [15 16 18] However increasing the Hb-C concentration overthe optimum concentration resulted in decreased functionality and viability ofencapsulated islets as shown in Figs 3 and 4 The phenomena may be attributed todeteriorated mass transport by increased viscosity andor density of the polymericECM Although the polymeric ECM itself did not act as a diffusional barrier forsmall molecules such as insulin nutrients and wastes [29] a high concentrationof Hb-C in polymeric ECM may act as a mass transport barrier which resulted inreduced cell functionality and viability

Size or volume issue of a BAP is another important factor for its successfulapplication In the case of islet macroencapsulation the volume issue could be

1534 Y Y Kim et al

overcome by increasing islet density However cell viability is significantly affectedby cell density Figures 6 and 7 clearly show the decreased cell viability byincreasing islet density The enhanced cell viability by incorporation of Hb-Cwas caused by Hb-C groups however a high islet density (2000 isletsdevice)reduced viability by developing central necrosis By increasing islet density strongcompetition between encapsulated islets for nutrients and oxygen and seriousaccumulation of waste may occur in the device which results in reduced cellviability by increasing cell necrosis andor apoptosis On the other hand decreasingthe islet density increases the device volume which may evoke the volume issue

CONCLUSIONS

In this report we investigated the effect of Hb-C on the functionality and viabilityof an islet macroencapsulation device with a thermally reversible polymeric ECMBy introduction of 025 mM Hb-C into the islet macrocapsule more than 100enhanced insulin secretion and cell viability were observed after 4 weeks cultureFurthermore increase of islet density up to 1000 isletsdevice (154 microl volume)without any negative effect on insulin secretion and viability was possible byintroduction of 025 mM Hb into islet macrocapsules Although the practical in vivostudy remains as further study the incorporation of Hb-C into islet macrocapsulemight prolong the life span of encapsulated islets and reduce the implant volume byincreasing the functionality of islets and islet density

Acknowledgements

This work was supported by NIH DK56884 and BK 21 program (South Korea)

REFERENCES

1 G Reach Diabet Med 10 105 (1993)2 R P Lanza and W L Chick Ann N Y Acad Sci 831 323 (1997)3 R Calafiore Diabet Metab Rev 14 315 (1998)4 M K Lee and Y H Bae Adv Drug Deliv Rev 42 103 (2000)5 C K Colton and E S Avgoustiniators J Biomech Eng 113 152 (1992)6 Q P Hou and Y H Bae Adv Drug Deliv Rev 35 271 (1999)7 D Brandhorst H Brandhorst A Zwolinski F Nahidi and R G Bretzel Transplantation 71

179 (2001)8 J C Hutton and W J Malaisse Diabetologia 18 395 (1980)9 K E Dionne C K Colton and M L Yarmush Diabetes 42 12 (1993)

10 J Schrezenmeir L Gero M Solhdju J Kirchgessner C Laue J Beyer H Stier and W Muller-Klieser Transplant Proc 26 809 (1994)

11 M Ohta D Nelson M D Meglasson and M Erecinska J Biol Chem 265 17525 (1990)12 J S Whang S Y Chae M K Lee and Y H Bae Biomaterials 19 1198 (1998)13 K H Park S W Kim and Y H Bae J Biomed Mater Res 55 72 (2001)14 H Gappa M Baudys J J Koh S W Kim and Y H Bae Tissue Eng 7 35 (2001)

Enhanced function of islet macrocapsule 1533

Figure 7 Fluorescence images of islets in the macrocapsules after 4 weeks of culture (A) 2000isletspouch with 025 mM Hb-C (B) 2000 isletspouch without Hb-C (C) 1000 isletspouch with025 mM Hb-C and (D) 1000 isletspouch without Hb-C

introduction of a bioactive ingredient (Hb-C) may enhance the cell functionality andviability as shown in our previous islet microencapsulation system [18 25]

As shown in Figs 3 and 4 insulin-secreting functions and cell viability weregreatly affected by the incorporated Hb-C in a concentration-dependent fashionThe optimum concentration of Hb-C as obtained from the experiments was025 mM (based on only Hb) The enhancing effect of the incorporated Hb-C oncell functionality and viability could be explained by the oxygen supply and otherreported bioactivities such as radical scavenging buffering and cell-stabilizingeffect [15 18 25] Most importantly increased oxygen supply may play acritical role in the enhanced cell functionality and viability Many studies havefocused on the limited oxygen supply on isolated and encapsulated islets Insulinsecretion especially glucose-stimulated insulin secretion was greatly affected byoxygen partial pressure [9] and the malfunction of islets or insulinoma cellswas overcome by increasing oxygen supply using hemoglobin Hb-C and waterelectrohydrolysis [15 16 18] However increasing the Hb-C concentration overthe optimum concentration resulted in decreased functionality and viability ofencapsulated islets as shown in Figs 3 and 4 The phenomena may be attributed todeteriorated mass transport by increased viscosity andor density of the polymericECM Although the polymeric ECM itself did not act as a diffusional barrier forsmall molecules such as insulin nutrients and wastes [29] a high concentrationof Hb-C in polymeric ECM may act as a mass transport barrier which resulted inreduced cell functionality and viability

Size or volume issue of a BAP is another important factor for its successfulapplication In the case of islet macroencapsulation the volume issue could be

1534 Y Y Kim et al

overcome by increasing islet density However cell viability is significantly affectedby cell density Figures 6 and 7 clearly show the decreased cell viability byincreasing islet density The enhanced cell viability by incorporation of Hb-Cwas caused by Hb-C groups however a high islet density (2000 isletsdevice)reduced viability by developing central necrosis By increasing islet density strongcompetition between encapsulated islets for nutrients and oxygen and seriousaccumulation of waste may occur in the device which results in reduced cellviability by increasing cell necrosis andor apoptosis On the other hand decreasingthe islet density increases the device volume which may evoke the volume issue

CONCLUSIONS

In this report we investigated the effect of Hb-C on the functionality and viabilityof an islet macroencapsulation device with a thermally reversible polymeric ECMBy introduction of 025 mM Hb-C into the islet macrocapsule more than 100enhanced insulin secretion and cell viability were observed after 4 weeks cultureFurthermore increase of islet density up to 1000 isletsdevice (154 microl volume)without any negative effect on insulin secretion and viability was possible byintroduction of 025 mM Hb into islet macrocapsules Although the practical in vivostudy remains as further study the incorporation of Hb-C into islet macrocapsulemight prolong the life span of encapsulated islets and reduce the implant volume byincreasing the functionality of islets and islet density

Acknowledgements

This work was supported by NIH DK56884 and BK 21 program (South Korea)

REFERENCES

1 G Reach Diabet Med 10 105 (1993)2 R P Lanza and W L Chick Ann N Y Acad Sci 831 323 (1997)3 R Calafiore Diabet Metab Rev 14 315 (1998)4 M K Lee and Y H Bae Adv Drug Deliv Rev 42 103 (2000)5 C K Colton and E S Avgoustiniators J Biomech Eng 113 152 (1992)6 Q P Hou and Y H Bae Adv Drug Deliv Rev 35 271 (1999)7 D Brandhorst H Brandhorst A Zwolinski F Nahidi and R G Bretzel Transplantation 71

179 (2001)8 J C Hutton and W J Malaisse Diabetologia 18 395 (1980)9 K E Dionne C K Colton and M L Yarmush Diabetes 42 12 (1993)

10 J Schrezenmeir L Gero M Solhdju J Kirchgessner C Laue J Beyer H Stier and W Muller-Klieser Transplant Proc 26 809 (1994)

11 M Ohta D Nelson M D Meglasson and M Erecinska J Biol Chem 265 17525 (1990)12 J S Whang S Y Chae M K Lee and Y H Bae Biomaterials 19 1198 (1998)13 K H Park S W Kim and Y H Bae J Biomed Mater Res 55 72 (2001)14 H Gappa M Baudys J J Koh S W Kim and Y H Bae Tissue Eng 7 35 (2001)

1534 Y Y Kim et al

overcome by increasing islet density However cell viability is significantly affectedby cell density Figures 6 and 7 clearly show the decreased cell viability byincreasing islet density The enhanced cell viability by incorporation of Hb-Cwas caused by Hb-C groups however a high islet density (2000 isletsdevice)reduced viability by developing central necrosis By increasing islet density strongcompetition between encapsulated islets for nutrients and oxygen and seriousaccumulation of waste may occur in the device which results in reduced cellviability by increasing cell necrosis andor apoptosis On the other hand decreasingthe islet density increases the device volume which may evoke the volume issue

CONCLUSIONS

In this report we investigated the effect of Hb-C on the functionality and viabilityof an islet macroencapsulation device with a thermally reversible polymeric ECMBy introduction of 025 mM Hb-C into the islet macrocapsule more than 100enhanced insulin secretion and cell viability were observed after 4 weeks cultureFurthermore increase of islet density up to 1000 isletsdevice (154 microl volume)without any negative effect on insulin secretion and viability was possible byintroduction of 025 mM Hb into islet macrocapsules Although the practical in vivostudy remains as further study the incorporation of Hb-C into islet macrocapsulemight prolong the life span of encapsulated islets and reduce the implant volume byincreasing the functionality of islets and islet density

Acknowledgements

This work was supported by NIH DK56884 and BK 21 program (South Korea)

REFERENCES

1 G Reach Diabet Med 10 105 (1993)2 R P Lanza and W L Chick Ann N Y Acad Sci 831 323 (1997)3 R Calafiore Diabet Metab Rev 14 315 (1998)4 M K Lee and Y H Bae Adv Drug Deliv Rev 42 103 (2000)5 C K Colton and E S Avgoustiniators J Biomech Eng 113 152 (1992)6 Q P Hou and Y H Bae Adv Drug Deliv Rev 35 271 (1999)7 D Brandhorst H Brandhorst A Zwolinski F Nahidi and R G Bretzel Transplantation 71

179 (2001)8 J C Hutton and W J Malaisse Diabetologia 18 395 (1980)9 K E Dionne C K Colton and M L Yarmush Diabetes 42 12 (1993)

10 J Schrezenmeir L Gero M Solhdju J Kirchgessner C Laue J Beyer H Stier and W Muller-Klieser Transplant Proc 26 809 (1994)

11 M Ohta D Nelson M D Meglasson and M Erecinska J Biol Chem 265 17525 (1990)12 J S Whang S Y Chae M K Lee and Y H Bae Biomaterials 19 1198 (1998)13 K H Park S W Kim and Y H Bae J Biomed Mater Res 55 72 (2001)14 H Gappa M Baudys J J Koh S W Kim and Y H Bae Tissue Eng 7 35 (2001)

Enhanced function of islet macrocapsule 1535

15 J Schrezenmeir F Velten and J Beyer Transplant Proc 26 792 (1994)16 H Wu E S Avgoustiniatos L Swette S Bonner-Weir G C Weir and C K Colton Ann N

Y Acad Sci 875 105 (1999)17 U Noth P Grohn A Jork U Zimmermann A Haase and J Lutz Magn Reson Med 42 1039

(1999)18 S Y Chae S W Kim and Y H Bae Tissue Eng 8 379 (2002)19 C K Han and Y H Bae Polymer 39 2809 (1998)20 Y H Bae B Vernon C K Han and S W Kim J Control Rel 53 249 (1998)21 B Vernon S W Kim and Y H Bae J Biomed Mater Res 51 69 (2000)22 A Kikuchi and T Okano J Control Rel 102 361(2005)23 W Leobandung H Ichikawa Y Fukumori and N A Peppas J Control Rel 80 357 (2002)24 H von Recum T Okano and S W Kim J Control Rel 55 121 (1998)25 S Y Chae Y Y Kim S W Kim and Y H Bae Transplantation 78 392 (2004)26 S Y Chae M Lee S W Kim and Y H Bae Biomaterials 25 843 (2004)27 J W Park MSc thesis Kwangju Institute of Science and Technology Kwangju (2001)28 H L Bank Diabetologia 30 812 (1987)29 B Vernon S W Kim and Y H Bae J Biomater Sci Polymer Edn 10 183 (1999)