The three-dimensional nanofiber scaffold culture conditionimproves viability and function of islets

Ming Zhao,1 Chun Song,1 Weihui Zhang,1 Yan Hou,2 Renping Huang,1 Yimin Song,3

Wanjun Xie,4 Yubo Shi,1 Chunfang Song11The Key Laboratory of Cell Transplantation of Ministry of Health and Department of General Surgery,The First Affiliated Hospital of Harbin Medical University, Harbin 150001, China2Department of Biostatistics, Harbin Medical University, Harbin 150086, China3Department of General Surgery, Peking Union Medical College Hospital, Beijing 100730, China4Department of Endocrine, The First Affiliated Hospital of Harbin Medical University, Harbin 150001, China

Received 20 January 2009; revised 21 April 2009; accepted 30 June 2009Published online 24 March 2010 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jbm.a.32624

Abstract: Significant problems existing in the islet trans-plantation include a poor survival ability of the islet cellscultured under static conditions in vitro, decreased secre-tion function, and limited transplantation efficiency. In thisstudy, we cocultured the three-dimensional (3D) self-assembling peptide nanofiber hydrogel scaffold with theislets from adult Wistar rats. The nanofiber scaffold con-structed a 3D environment for the islets culture. Theresults of DTZ staining showed that the purity of the isletsin the scaffold was >80%. The result of the fluorescentstaining with AO-PI demonstrated that the viability of theislets in the 3D culture environment (within scaffold) wasgreater than those in the two-dimensional (2D) cultureenvironment (without scaffold). The islets encapsulated inthe 3D peptide nanofiber scaffold exhibited better secretion

function. The insulin releasing index in the 3D group wasremarkably higher than that in the 2D group. By scanningelectron microscopy, it was observed that the 3D self-assembling peptide nanofiber hydrogel scaffold formed anano scale fiber with a geometric form and the islets wereencapsulated in this scaffold. Our research demonstratedthat this nanofiber scaffold provided a favorable 3D envi-ronment for the islets to be cultured in vitro and thenimprove the secretion function and prolong the survivaltime of the islet in vitro. � 2010 Wiley Periodicals, Inc.J Biomed Mater Res 94A: 667–672, 2010

Key words: islet transplantation; diabetes; three-dimen-sional self-assembling peptide nanofiber hydrogel scaffold;scanning electron microscope

INTRODUCTION

The islet transplantation is an ideal therapy for thetreatment of type 1 and 2 diabetes.1,2 The islets can becultured in vitro for a period of time and theirimmunogenicity is decreased.3 However, the viabilityof islets cultured under two-dimensional (2D) staticconditions in vitro is poor, which affects the secretionfunction of islets. Therefore, it is necessary to improvethe microenvironment in vitro for islets culture.

During the last few years, progress has been madein the area of tissue engineering.4–9 Some biomateri-als involving microencapsulation.10,11 fibrin glue,12

polymeric biomaterials, such as polyglycolide

(PGA),13 polylactide (PLA),14 polylactide-co-glycolide(PLGA),15,16 poly-L-lactide-co-caprolactone (L-PLCL)and polydioxanone (PDO), have already beenapplied to the field of tissue engineering.17–19

Furthermore, recent reports in the tissue engineer-ing literature underline the importance of controllingthe geometry of the biomaterials. To make the bio-materials suitable for cell growth and differentiation,modeled extracellular matrix (ECM) that has a three-dimensional (3D) geometric structure and biophysi-cal properties somewhere between the cell-matrixand normal cell–cell interactions are desired. The 3Dself-assembling peptide nanofiber hydrogel scaffoldcan spontaneously assemble into the defined geome-tries of a nanostructure at the molecular level. Thesepeptides can form two b-sheet structures of distinctsurfaces in aqueous solution,20,21one hydrophilic andthe other hydrophobic.22,23 The presence of thehydrophobic surface facilitates the self-assembly inwater, similar to the condition of protein folding.Because the water content of these nanofibers is larger

Correspondence to: C. Song; e-mail: songchun56@yahoo.com.cnContract grant sponsor: National Natural Science Foun-

dation of China; contract grant number: 30570931

� 2010 Wiley Periodicals, Inc.

than 99% (peptide content 1–10 mg/mL),24,25 they arethree orders of magnitude thinner than other biopoly-mers, and thus can be fabricated into different geomet-rical shapes that can allow some nutritional substancesand macromolecules to pass through as well as allow-ing metabolism to occur.26 The 3D peptide nanofiberscaffold can improve the growth and secretion functionof the cells by creating a microenvironment that canbetter simulate the condition of the living body in vitroand provide the essential structural pattern to facilitatethe correct composition of their ECMs.27

So far, no research has been conducted about thisscaffold effect on islet culture, therefore, in ourresearch, we cocultured 3D self-assembling peptidenanofiber hydrogel scaffold and islets of adult Wis-tar rats. The results demonstrated that this 3D nano-fiber scaffold culture condition improved the viabil-ity and secretion function of the islets and prolongedtheir survival time by providing a favorable micro-environment in vitro.

MATERIALS AND METHODS

Islet cell isolation and purification

Adult Wistar rats, either female or male, weighing from250 to 300 g, fasted for 12 h, were anaesthetized by 0.5%pentobarbital sodium (Shanghai Chemical Reagent, China,30 mg/kg) in the peritoneal cavity. The skin preparation ofthe chest and abdomen was sterilized using an iodine com-plex and the skin was cut to the peritoneum step by step(the pancreas tube and common bile duct were observed).First, one number silk thread was used to ligate the tail ofthe common bile duct, near the duodenum department andthen conducting a function test by pulling straight the com-mon bile duct at the point where it nears the common bileduct at the hepatic portal. Number four trocar was used topuncture the common bile duct and inject a prepared cold0.25% type V collagenase (Sigma) solution 8–10 mL. This so-lution was injected into the pancreas via the common bileduct, the pancreatic tube. When the pancreas was com-pletely expanded, it was immediately collected and placedin an aseptic centrifugal tube and incubated for digestionfor 10–15 min in a 388C water bath concussing vessel. Whenthe pancreatic tissue was reduced to fine particles, 48C30 mL RPMI-1640 culture solution (Hyclon) containing 20%fetal bovine serum was added to stop digestion; the solutionwas placed at room temperature for 5–10 min and then wasfiltrated with a 500 lm stainless steel mesh. The islets werepurified by the Ficoll method (GE).

Detection of islets purity

The DTZ (Sigma) and islets were mixed and cultured inthe incubator for 5 to 10 min, and then the purity and sizeof the islet was detected. The purity of islets is stained/total (stained and unstained) cell clusters 3 100%.

Islets culture

The islets were divided into the control group and theexperimental group for culture and we added 600 islets toeach flask. In the control group, islets were cultured inRPMI-1640 culture medium containing 20% fetal bovine se-rum, 1% penicillin–streptomycin–amphotericinB (Amresco)and 10 mmol/L Hepes (Hyclon). Islets and the substancescontained in the culture medium were mixed sufficiently.In the experimental group, islets and the 3D self-assem-bling peptide nanofiber hydrogel scaffold (BD) were cocul-tured, and then were put into the same culture medium asthat of the control group and cultured further in the incu-bator. The islets of the two groups were both cultured inthe incubator (MCO-15AC. Sanyo Inc) with the volumefraction 95% air, 5% CO2, 378C constant temperature.

3D self-sssembling peptide nanofiber hydrogelscaffold encapsule islets

We first placed 5 mL of peptide nanofibe scaffold at pH3 into ultrasonic processors for 30 min to redissolve thepeptide scaffold, which makes it easier to be absorbed.The cell sediments were resuspended in 10% sucrose, thenmixed immediately with an equal volume peptide scaffoldwhich had been ultrasonically treated. Next, the islets/peptide scaffold mixture was pipetted with a sterile capil-lary and then immediately pipetted into the requiredRPMI-1640 culture dish (containing 20% fetal bovine se-rum, 1% penicillin–streptomycin–amphotericinB, and 10mmol/L Hepes). The result was a jelly-like pellet with avolume of almost 35 lL containing the islets. The mediumwas changed two to three times within the next 30 minand the pH value of the medium was maintained at 7. Themedium dish containing the islets/peptide scaffold pelletswas placed in the aforementioned incubator.

Assessment of viability in vitro

The viability of the islets was assessed using AcridineOrange-Propidium lodide fluorescent staining (AO-PI,Sigma). For staining, first 0.01 mL of AO and 1 mL of PIwere mixed, then were diluted 10 times in Hanks solution,and then filtered and sterilized with a 0.22 lm filter de-vice. Next, the AO-PI solution was mixed with the isletsand then incubated for 5 to 10 min. Green (AO) and red(PI) fluorescence were both observed using fluorescent mi-croscopy with a 490 nm excitation filter and a 510 nm gra-ting filter to indicate the viability of the islets. Living isletsare shown in green and dead islets are shown in red. Thesurvival rate of islets is green fluorescence/total number ofislets (red and green) 3 100%.

In vitro detection of secretion function

The test to determine if glucose can simulate the releaseof insulin was done as follows: wash the islets three timesin Hanks solution, and then add them to RPMI-1640 cul-ture medium without serum at a low concentration of glu-

668 ZHAO ET AL.

Journal of Biomedical Materials Research Part A

cose (5.6 mmol/L). These islets were cultured for 4 h at378C. A volume of 1 mL of culture solution was collectedand used to detect the insulin content. A similar methodwas performed under high concentrations of glucose (16.7mmol/L) and then the same process was repeated. The in-sulin releasing index is equal to insulin content in the cul-ture medium at a high concentration of glucose/insulincontent in the culture medium at a low concentration ofglucose. A content of insulin <0.5 ng/mL indicates that alower amount of insulin is secreted.

Islets/peptide scaffold under scanning electronmicroscopy

The islets/peptide scaffold mixture was cocultured inRPMI-1640 culture medium containing 20% fetal bovine se-rum, 1% penicillin–streptomycin–amphotericinB, and 10mmol/L Hepes, where the islets/peptide scaffold mix cul-ture in the shape of jelly-like pellets, and the islets/peptidescaffold were enwrapped in the peptide scaffold. Theislets/peptide scaffold mixture was quickly pipetted andadded to 3% glutaraldehyde for fixing. The sample waswashed after 2 h and then fixed with 1% osmic acid foranother 2 h and then dehydrated with increasing concen-trations of 50, 70, and 90% ethanol for 15 min each time.Finally, samples were dehydrated in 100% ethanol for 10min. Sections were frozen, dried, and coated. The resultwas observed under scanning electron microscope (SEM)(S-3400N, Hitachi, Japan).

Statistical analysis

Because the survival rate is not normally distributed, wetransformed it into a normal distribution using a mathemat-

ical method. The description of the islet releasing index andthe transformed islet survival rate are expressed as meanvalues 6 standard deviation (X 6 SD). Differences in the is-let releasing index and transformed islet survival rate at thefive different time points under the two conditions were an-alyzed by multivariate analysis using SAS 9.1. A p-value <0.05 was considered to be statistically significant.

RESULTS

Islets in peptide scaffold specific-stain

Only the islets are stained red by DTZ and the exo-crine gland and other cells are not stained. Under theinverted microscope, round, ellipse, or irregularshapes of islets are visible and clearly outlined. Theislets are 50 to 350 lm in diameter. The purity of theislets in the 3D peptide nanofibe scaffold is �80% (Fig.1), whereas the islets cultured in the 2D environmentbegin to die after the prolonged time and are not sig-nificantly dyed by the DTZ stain in this group (Fig. 2).

Viability and survival rate of islets

The islets stained by AO-PI were observed underthe inversed fluorescent microscope. The living cellsare shown in green and the dead cells are shown inred. The viability of the experimental group (3Dculture) was better in comparison with that of thecontrol group (2D culture) (Figs. 3 and 4). The result

Figure 1. The islets in 3D cultures. The islets in the pep-tide hydrogel scaffold are shown significant when stainedby DTZ and are in a tight junction, forming cell clusters inround or irregular shapes with a clear outline. The purityof the islets are >80% (inverted microscope 3320).

Figure 2. The islets in 2D cultures. The islets in the controlgroup begin to die with prolonged culture time. The isletsshow little significance when stained by DTZ (invertedmicroscope 3320).

3D NANOFIBER SCAFFOLD CULTURE IMPROVES VIABILITY AND FUNCTION OF ISLETS 669

Journal of Biomedical Materials Research Part A

of survival rate of islets demonstrated a statisticallysignificant difference (p < 0.05) between the meanvectors of the viability at five different time pointsin the experimental group and the control group(Table I).

Secretion function of islets in the 3Dpeptide scaffold

The islets encapsulated in the 3D peptide nano-fiber scaffold exhibit better secretion function. Theinsulin releasing index of the experimental groupwas remarkably higher than that of the controlgroup (Table II). The result of the statistical analysisalso showed that the mean vectors of the insulinreleasing index calculated at five different timepoints in the experimental group and the controlgroup were statistically significant (p < 0.05) and thedifferences between the two groups at the 11th and15th days were most obvious.

Scanning electron microscopy

The islets encapsulated by the 3D peptide nano-fiber scaffold are in well shapes (Figs. 5 and 6).Thepeptide nanofiber scaffold separates the islets, whichshow that the islets grow in a 3D environment.Because the islets cultured in the 2D environmentbegin to die after 7 days in vitro, few living islets arevisible under the SEM.

DISCUSSION

An ideal therapy for the diabetes lacking insulin isto be able to transplant islets.28,29 Islets for diabetestreatment cultured in vitro for a period of time canhelp to reduce the immunogenicity and decrease theimmunological rejection of the receptors.3 However,islets cultured in the 2D environment have shownpoor survival ability that affected the secretion func-

Figure 3. The viability of the islets in 3D cultures. Fluo-rescent staining with AO-PI for islets in the 3D environ-ment (experimental group). Living cells shown green arehigher in number than the dead cells shown in red. It indi-cates that the viability of the islets is good (inverted fluo-rescent microscope 3150). [Color figure can be viewed inthe online issue, which is available at www.interscience.wiley.com.]

Figure 4. The viability of the islets in 2D cultures. Fluo-rescent staining with AO-PI for islets in the 2D environ-ment (control group). Living cells shown green are less innumber than dead cells shown in red. It indicates that theviability of the islets is poor (inverted fluorescent micro-scope 3150). [Color figure can be viewed in the onlineissue, which is available at www.interscience.wiley.com.]

TABLE IIslet Survival Rate Transformed from Five Different Time Points Under Two Conditions (X 6 SD)

Group n

Time (Day)

0 3 7 11 15

Control 8 1.32 6 0.01 1.29 6 0.01 1.13 6 0.09 0.87 6 0.13 0.70 6 0.11Experimental group* 8 1.32 6 0.01 1.29 6 0.01 1.20 6 0.04 1.14 6 0.05 1.07 6 0.02

The mean vectors of transformed survival rates of islets between two groups are statistically different and significant bymultivariate analysis. The Wilks’ Lambda test result demonstrated that the mean vector of control group was differentfrom that of experimental group (p < 0.05).*p < 0.05 versus control using Wilks’ Lambda test.

670 ZHAO ET AL.

Journal of Biomedical Materials Research Part A

tion of the islets. Our research applies the 3D self-assembling peptide nanofiber hydrogel scaffold andcocultures it with the islets from adult Wistar rats. Thisscaffold can structure a 3D culture system, providing abetter simulated microenvironment for improving theviability and the secretion function of the islets.

Our research utilizing a peptide nanofiber scaffoldproviding a 3D environment for cell growth demon-strated that these islets exhibit better viability andthe secretion function. The result of these experi-ments demonstrate that the islets in the scaffold canstill be stained red by DTZ, and the tight junctionsare visible between cells. The islets form round orirregular cell clusters, which are 50 to 350 lm in di-ameter and whose boundaries are clear. The islets inthe 3D peptide scaffold have a purity of >80%(Fig. 1). Contrary to this, the islets cultured in the 2Denvironment begin to die after the 7th day and thusare not significantly dyed by the DTZ stain (Fig. 2).The result of the fluorescent staining with AO-PIdemonstrates that the viability of the islets in the 3Denvironment is much better compared with those inthe 2D environment (Figs. 3 and 4).

Our results indicate that the survival rates of isletsin the 3D environment are much greater in compari-son with those in the 2D environment. The statisticalanalysis demonstrates that there is a statistically sig-nificant (p < 0.05) difference between the mean vec-tors of the viability at five different time points in theexperimental group and the control group (Table I).The insulin releasing index of the islets in the 3Denvironment is much higher than that in the 2D envi-ronment (Table II). The statistical analysis also provesthat the mean vectors of the insulin releasing index atfive different time points in the experimental groupand the control group are statistically significant (p <0.05) and the differences between the two groupswere most obvious at the 11th and 15th days.

From the observations of the SEM, this kind ofscaffold separates the islets and stimulates the ECM,which supplies sufficient nutrition for islets growthand also indicates better grow of the islets in a 3Denvironment. The islets, which were cocultured for 7days with the peptide nanofiber scaffold exhibit awell shape (Figs. 3 and 4). Whereas the islets in the2D environment are lacking nutrition, they begin to

TABLE IIIslet Releasing Index from Five Different Time Points Under Two Conditions (X 6 SD)

Group n

Time (Day)

0 3 7 11 15

Control 8 1.22 6 0.26 1.38 6 0.18 1.33 6 0.40 1.27 6 0.40 0.37 6 0.60Experimental group* 8 1.18 6 0.14 1.13 6 0.22 1.22 6 0.26 1.83 6 0.53 1.47 6 0.25

The mean vectors of islet releasing index between two groups are statistically different and significant by multivariateanalysis. The Wilks’ Lambda test result demonstrated that the mean vector of control group was different from that of ex-perimental group (p < 0.05).*p < 0.05 versus control using Wilks’ Lambda test.

Figure 5. The shape of islets. The 3D peptide nanofiberscaffold self-assemble into a porous configuration. Thehole wall is a nano order of magnitude in thickness. Theislets, cocultured with the peptide nanofiber scaffold andencapsulated by the scaffold, exhibit the well shape (scan-ning electron microscopy). Scale bar 5 20 lm.

Figure 6. 3D scaffold stimulates the ECM. The peptidenanofiber scaffold stimulates the ECM by forming a layerof a nano order of magnitude which separates the isletsfrom each other (scanning electron microscopy). Scale bar5 40 lm.

3D NANOFIBER SCAFFOLD CULTURE IMPROVES VIABILITY AND FUNCTION OF ISLETS 671

Journal of Biomedical Materials Research Part A

die after being cultured for 7 days, and thus, underthe SEM, few living cells are visible.

CONCLUSIONS

The results of our studies demonstrate that 3Dself-assembling nanofiber peptide hydrogel scaffoldprovides a favorable 3D environment for culturingthe islets in vitro and also improves the viability andthe secretion function of the islets, thus prolongingthe survival time of the islets in vitro.

All authors thank Dr. Kang Li for helpful commentsand suggestions on the statistical analysis and Li Zhangand Yakuen Zhang for the vigorous supports and theirparticipation in this research.

References

1. Gao F, Wu DQ, Hu YH, Jin GX. Extracellular matrix gel isnecessary for in vitro cultivation of insulin producing cellsfrom human umbilical cord blood derived mesenchymalstem cells. Chin Med J 2008;121:811–818.

2. Teng Y, Xue WJ, Ding XM, Feng XS, Xiang HL, Jiang YZ,Tian PX. Isolation and culture of adult sertoli cells and theireffects on the function of co-cultured allogeneic islets in vitro.Chin Med J 2005;118:1857–1862.

3. Kuttler B, Hartmann A, Wanka H. Long-term culture of isletsabrogates cytokineinduced or lymphocyte-induced increaseof antigen expression on ß cells. Transplantation 2002;74:440–445.

4. Venugopal J, Low S, Choon AT, Kumar AB, Ramakrishna S.Electrospun-modified nanofibrous scaffolds for the minerali-zation of osteoblast cells. J Biomed Mater Res A 2008;85:408–417.

5. Li WJ, Danielson KG, Alexander PG, Tuan RS. Biologicalresponse of chondrocytes cultured in three-dimensional nano-fibrous poly (epsilon-caprolactone) scaffolds. J Biomed MaterRes A 2003;67:1105–1114.

6. He W, Ma Z, Teo WE, Dong YX, Robless PA, Lim TC, Ram-akrishna S. Tubular nanofiber scaffolds for tissue engineeredsmall-diameter vascular grafts. J Biomed Mater Res A 2009;90:205–216.

7. Carampin P, Conconi MT, Lora S, Menti AM, Baiguera S, Bel-lini S, Grandi C, Parnigotto PP. Electrospun polyphosphazenenanofibers for in vitro rat endothelial cells proliferation.J Biomed Mater Res A 2007;80:661–668.

8. Gugala Z, Gogolewski S. In vitro growth and activity of pri-mary chondrocytes on a resorbable polylactide three-dimen-sional scaffold. J Biomed Mater Res 2000;49:183–191.

9. Meretoja VV, Malin M, Seppala JV, Narhi TO. Osteoblastresponse to continuous phase macroporous scaffolds understatic and dynamic culture conditions. J Biomed Mater Res A2009;89:317–325.

10. Calafiore R, Basta G, Luca G, Lemmi A, Racanicchi L, Man-cuso F, Montanucci MP, Brunetti P. Standard technical proce-dures for microencapsulation of human islets for graft intononimmunosuppressed patients with type 1 diabetes melli-tus. Transplant Proc 2006;38:1156–1157.

11. Tashiro H, Iwata H, Warnock GL, Tanigawa M, Ototani Y,Tsuji T. Functional assessment of microencapsulated porcine

islets with agarose polystyrene sulfonic acid in vitro and inxenotransplantation. Transplant Proc 2005;37:3512–3513.

12. Andrades P, Asiedu C, Rodriguez C, Goodwin KJ, Mccarn J,Thomas JM. Subcutaneous pancreatic islet transplantationusing fibrin glue as a carrier. Transplant Proc 2006;39:191–192.

13. Wang YC, Lin MC, Wang DM, Hsieh HJ. Fabrication of anovel porous PGA-chitosan hybrid matrix for tissue engineer-ing. Biomaterials 2003;24:1047–1057.

14. Howard D, Partridge K, Yang X, Clarke NM, Okubo Y, Bes-sho K, Howdle SM, Shakesheff KM, Oreffo RO. Immunose-lection and adenoviral genetic modulation of human osteo-progenitors: In vivo bone formation on PLA scaffold. Bio-chem Biophys Res Commun 2002;299:208–215.

15. Chen G, Liu D, Tadokoro M, Hirochika R, Ohgushi H,Tanaka J, Tateishi T. Chondrogenic differentiation of humanmesenchymal stem cells cultured in a cobweb-like biodegrad-able scaffold. Biophys Res Commun 2004;322:50–55.

16. Partridge K, Yang X, Clarke NM, Okubo Y, Bessho K, SebaldW, Howdle SM, Shakesheff KM, Oreffo RO. AdenoviralBMP-2 gene transfer in mesenchymal stem cells: In vitro andin vivo bone formation on biodegradable polymer scaffolds.Biophys Res Commun 2002;292:144–152.

17. Ameer GA, Mahmood TA, Langer R. A biodegradable com-posite scaffold for cell transplantation. J Orthop Res 2002;20:16–19.

18. Ishaug-Riley SL, Okun LE, Prado G, Applegate MA, RatcliffeA. Human articular chondrocyte adhesion and proliferationon synthetic biodegradable polymer films. Biomaterials 1999;20:2245–2256.

19. Lu HH, Cooper JA Jr, Manuel S, Freeman JW, Attawia MA,Ko FK, Laurencin CT. Anterior cruciate ligament regenerationusing braided biodegradable scaffolds: In vitro optimizationstudies. Biomaterials 2005;26:4805–4816.

20. Holmes TC. Novel peptide-based biomaterial scaffolds for tis-sue engineering. Trends Biotechnol 2002;20:16–21.

21. Kisiday J, Jin M, Kurz B, Hung H, Semino C, Zhang S, Grod-zinsky AJ. Self-assembling peptide hydrogels fosters chon-drocyte extracellular matrix production and cell division:Implications for cartilage tissue repair. Proc Natl Acad SciUSA 2002;99:9996–10001.

22. Zhang S. Fabrication of novel biomaterial through molecularself-assembly. Nat Biotechnol 2003;21:1171–1178.

23. Zhang S, Holmes TC, Dipersio CM, Hynes RO, Su X, Rich A.Self-complementary oligopeptide matrices support mamma-lian cell attachment. Biomaterials 1995;16:1385–1393.

24. Holmes TC, Lacalle SD, Su X, Liu G, Rich A, Zhang S. Exten-sive neurite outgrowth and active synapse formation on self-assembling peptide scaffolds. Proc Natl Acad Sci USA2000;97:6728–6733.

25. Genove E, Shen C, Zhang S, Semino CE. The effect of func-tionalized self-assembling peptide scaffolds on human aorticendothelial cell function. Biomaterials 2005;26:3341–3351.

26. Semino CE, Merok JR, Crane GG, Panagiotakos G, Zhang S.Functional differentiation of hepatocyte-like spheroid struc-tures from putative liver progenitor cells in three-dimen-sional peptide scaffolds. Differentiation 2003;71:262–270.

27. Garreta E, Genove E, Borros S, Semino CE. Osteogenic differ-entiation of mouse embryonic stem cells and mouse embry-onic fibroblasts in a three-dimensional self-assembling pep-tide scaffold. Tissue Eng 2006;12:2215–2227.

28. Smith RM, Gale EAM. Survival of the fittest? Natural selectionin islet transplantation. Transplantation 2005;79:1301–1303.

29. Swanson CJ, Olack BJ, Goodnight D, Zhang L, MohanakumarT. Improved methods for the isolation and purification ofporcine islets. Hum Immunol 2001;62:739–749.

672 ZHAO ET AL.

Journal of Biomedical Materials Research Part A