Biomaterials 25 (2004) 3211–3222

ARTICLE IN PRESS

*Correspondin

8490.

E-mail addres

0142-9612/$ - see

doi:10.1016/j.bio

Chondrogenic differentiation of adipose-derived adult stem cells inagarose, alginate, and gelatin scaffolds

Hani A. Awada, M. Quinn Wickhama, Holly A. Leddya, Jeffrey M. Gimbleb,Farshid Guilaka,*

aDepartment of Surgery, Division of Orthopaedic Surgery, Duke University Medical Center, Durham, NC 27710, USAbPennington Biomedical Research Center, Louisiana State University, Baton Rouge, LA, USA

Received 9 September 2003; accepted 29 September 2003

Abstract

The differentiation and growth of adult stem cells within engineered tissue constructs are hypothesized to be influenced by cell-

biomaterial interactions. In this study, we compared the chondrogenic differentiation of human adipose-derived adult stem

(hADAS) cells seeded in alginate and agarose hydrogels, and porous gelatin scaffolds (Surgifoam), as well as the functional

properties of tissue engineered cartilage constructs. Chondrogenic media containing transforming growth factor beta 1 significantly

increased the rates of protein and proteoglycan synthesis as well as the content of DNA, sulfated glycosaminoglycans, and

hydroxyproline of engineered constructs as compared to control conditions. Furthermore, chondrogenic culture conditions resulted

in 86%, and 160% increases (po0:05) in the equilibrium compressive and shear moduli of the gelatin scaffolds, although they didnot affect the mechanical properties of the hydrogels over 28 days in culture. Cells encapsulated in the hydrogels exhibited a

spherical cellular morphology, while cells in the gelatin scaffolds showed a more polygonal shape; however, this difference did not

appear to hinder the chondrogenic differentiation of the cells. Furthermore, the equilibrium compressive and shear moduli of the

gelatin scaffolds were comparable to agarose by day 28. Our results also indicated that increases in the shear moduli were

significantly associated with increases in S-GAG content (R2 ¼ 0:36; po0:05) and with the interaction between S-GAG and

hydroxyproline (R2 ¼ 0:34; po0:05). The findings of this study suggest that various biomaterials support the chondrogenicdifferentiation of hADAS cells, and that manipulating the composition of these tissue engineered constructs may have significant

effects on their mechanical properties.

r 2003 Elsevier Ltd. All rights reserved.

Keywords: Alginate; Agarose; Hydrogel; Collagen; Gelatin; Cartilage tissue engineering; Stem cell; Differentiation

1. Introduction

Tissue engineering is a promising therapeutic ap-proach that combines cells, biomaterials, and environ-mental factors to induce differentiation signals intosurgically transplantable formats and promote tissuerepair and/or functional restoration [1–5]. Despite manyadvances, tissue engineers still face significant challengesin repairing or replacing tissues that serve predomi-nantly biomechanical functions such as articular carti-lage. One obstacle has been the development of acompetent cartilage scaffold that: (a) has mechanical

g author. Tel.: +1-919-684-2521; fax: +1-919-681-

s: guilak@duke.edu (F. Guilak).

front matter r 2003 Elsevier Ltd. All rights reserved.

materials.2003.10.045

properties and capability to withstand the large contactstresses and strains of an articulating joint, (b) allowsfunctional tissue growth, and (c) provides for appro-priate cell–matrix interactions to stimulate tissue growth[6,7]. An evolving discipline termed ‘‘functional tissueengineering’’ (FTE) seeks to address the functionalchallenges of tissues such as cartilage by attempting todefine sets of criteria that must be satisfied in order toovercome challenges associated with developing asuccessful tissue-engineered graft [6].Challenges related to the cellular component of an

engineered tissue include cell sourcing, expansion, anddifferentiation as well as regulatory and productionissues, such as sterility, safety, storage, shipping, qualitycontrol, and scale-up [8]. The use of human adiposederived adult stem (hADAS) cells represents a feasible

ARTICLE IN PRESSH.A. Awad et al. / Biomaterials 25 (2004) 3211–32223212

approach to many of these issues [9]. Adipose tissue isroutinely available in liter quantities from liposuctionsurgeries and yields an average of 400,000 cells per ml oftissue after expansion, providing the numbers of cellsnecessary for many tissue engineering applications [9].The hADAS cells are pluripotent, expressing thebiochemical profile of adipocytes, chondrocytes, hema-topoietic supporting cells, myocytes, neurons, andosteoblasts under appropriate culture conditionsin vitro [9].Other challenges are associated with the biomaterial

scaffolds designed to deliver the cells and guide tissuegrowth and differentiation. These biomaterials mustmeet several criteria to maximize the chances of asuccessful repair, including biodegradability and/orbiocompatibility, facilitating functional tissue growth,and appropriate biomechanical properties [10–13].Biomaterials used for cartilage tissue engineering canhave the form of either hydrogels or porous scaffolds.Among the hydrogel biomaterials, the seaweed-derivedalginate and agarose are typically thought to be inertbecause they lack native ligands that could allowinteraction with mammalian cells [14]. However, thesehydrogels provide a number of advantages for tissueengineering including the possibility of minimallyinvasive injection of hydrogel/cell constructs [15,16].Various researchers have investigated the ability ofalginate and agarose hydrogels to act as a scaffoldmaterial for chondrocytes to regenerate cartilage tissue[14,17–25]. While many of these studies demonstratedthat alginate and agarose promote maintenance ofthe chondrocytic phenotype in vitro, the successof tissue engineering applications using these hydrogelsmay be hindered by their poor biomechanical propertiesand handling characteristics. Furthermore, the ability ofthese hydrogels to support chondrogenic differentiationof adult stem cells has been less studied [26,27].Gelatin, a porous denatured collagen scaffold, has

been recently used as a scaffolding structure for cartilagetissue engineering [28,29]. The biological origin ofcollagen-derived gelatin makes this material an attrac-tive choice for tissue engineering [10]. However, there issome concern that type I collagen scaffolds may notpreserve the chondrocyte phenotype of cells as well astype II collagen scaffolds [30]. Furthermore, very little isknown about the functional (mechanical) properties ofthis biomaterial although its ability to support chon-drogenic differentiation of adult stem cells has beendemonstrated [28].The goal of this study was to assess the functional

(biologic, biochemical, and biomechanical) properties ofalginate hydrogel, agarose hydrogel, and gelatin (Surgi-foams, denatured porcine collagen type I) poroussponge as scaffolding biomaterials for cartilage tissueengineering using hADAS cells that have been shown topossess a chondrogenic potential under defined culture

conditions [26,31]. We hypothesized that the functionalproperties of tissue engineered cartilage will depend onthe choice of biomaterial scaffold, and that the constructmechanical properties would be directly correlated tothe synthesis and accumulation of extracellular matrixcomponents.

2. Materials and methods

2.1. Isolation of hADAS cells

Human adipose derived adult stem (hADAS) cellswere isolated from subcutaneous adipose tissue (n ¼ 3female donors, 29.777.2 years old (mean7standarddeviation) with a body mass index of 28.672.6 kg/m2)as previously described [26,32,33]. Briefly, liposuctionwaste tissue was digested with 0.25mg collagenase type I(200 units/mg) per ml of Krebs-Ringer-Bicarbonatesolution (Sigma, St Louis, MO) for 40min at 37�C withintermittent shaking. The floating adipocytes wereseparated from the precipitating stromal fraction bycentrifugation. The stromal cells were then plated intissue culture flasks at approximately 3500 cell/cm2 instromal media (DMEM/F-12 with 10% fetal bovineserum (FBS), 100 units/ml penicillin and 100mg/mlstreptomycin). The primary cells (P0) were cultured for4 to 5 days, after which they were harvested bytreatment with trypsin (0.05%)/EDTA, counted, andthen frozen in liquid nitrogen in cryopreservationmedium (80% FBS, 10% dimethylsulfoxide, 10%Dulbecco’s Modified Eagle Medium (DMEM)) untilthey were used in the following experiments.

2.2. Preparation of the biomaterial scaffolds

Cryopreserved cells were thawed and plated instromal media for 5 to 7 days until the cultures becameconfluent. Cells were harvested using trypsin/EDTA,counted, and then loaded onto alginate, agarose, andgelatin scaffolds as described. For the alginate scaffolds,cells were suspended in 2% (w/v) low viscosity alginate(Sigma) in 0.9% NaCl at a concentration of 107 cells/ml.The cell suspensions were cast in custom molds (25mmdiameter and 2mm thickness). The alginate molds wereplaced into a bath of 102mm CaCl2 and allowed to gelfor 10min. The CaCl2 was removed and the moldswere washed three times in PBS. Similarly, cells weresuspended in 2% (w/v) low-melting point agarose(Type VII, Sigma) at a concentration of 107 cells/ml.The agarose molds were allowed to gel at 4�C for10min. Smaller alginate and agarose disks were then cutout using a 6mm diameter biopsy punch and placed inthe appropriate culture conditions.Porous, absorbable gelatin (Surgifoams, Ethicon,

Inc., Somerville, NJ) disks (8mm diameter) were

ARTICLE IN PRESSH.A. Awad et al. / Biomaterials 25 (2004) 3211–3222 3213

pre-wetted in culture medium in flat bottom tubes.Aliquots (200 ml) of the cell suspension (107 cells/ml)were pipetted on top of each scaffold. The disk and cellsuspension were then centrifuged at 50� g for 30 s toseed the cells into the scaffold. The tubes containing theseeded scaffolds were then incubated on an orbitalshaker at 5% CO2 and 37

�C for 2 h to enhance cellinfiltration into the scaffold. The disks were thenincubated undisturbed overnight to allow for cellattachment. The following day, all disks were removedfrom the tubes and grown in the appropriate culturemedia. Acellular blank scaffolds were also prepared andincubated in identical conditions.

2.3. Culture conditions

The scaffolds were grown in control or chondrogenicculture media in a humidified environment at 5% CO2and 37�C for up to 28 days, with culture mediareplenished every 3 days. Control culture mediacomprised Dulbecco’s Modified Eagle Media–highglucose (DMEM-hg), 10% fetal bovine serum,100 units/ml penicillin, and 100mg/ml streptomycin.The chondrogenic culture media comprised the controlmedia, 1� insulin-transferrin-selenium supplement(ITS+, Collaborative Biomedical, Becton Dickinson,Bedford, MA), 0.15mm ascorbate 2-phosphate (Sigma),100 nm dexamethasone (Sigma), and 10 ng/ml rh-TGF-b1 (R & D Systems, Minneapolis, MN) [26,34,35].A broad array of biological, biochemical, and

mechanical analyses were used to assess the functionalproperties of the scaffolds (Table 1).

2.4. Biological properties

Cell viability and cellular morphology were examinedin situ on days 1, 7, 14, and 28 using a confocal laserscanning microscope (LSM 510, Carl Zeiss Microima-ging, Inc., Thornwood, NY ) and the fluorescent Live-

Table 1

Summary of the techniques used to assess the functional properties of the sc

Functional assay

Cellular viability and

morphology

Fluorescent calcein-AM and ethidium

Imaging using confocal laser scanning

Protein and proteoglycan

biosynthesis rates

Radioactive [3H]-proline incorporation

Radioactive [35S]-sulfate incorporation

Biochemical composition DNA content (PicoGreen assay)

Sulfated glycosaminoglycan content (D

Collagen content (hydroxyproline assa

Immunohistochemistry Type II collagen

Chondroitin sulfate

Biomechanical properties Equilibrium compressive modulus (ste

unconfined compression)

Equilibrium shear modulus (step-wise

Rheological properties (frequency swe

aSample size per culture condition, per time point. Scaffolds were prepare

Dead probes (Calcein AM and Ethidium homodimer,Molecular Probes, Eugene, OR). To quantify theprotein and proteoglycan biosynthetic activity, con-structs were dual-labeled with 10 mCi/ml [3H]-prolineand 5 mCi/ml [35S]-sulfate for 24 h on days 1, 7, 14 and28. Afterwards, the scaffolds were washed 4 times toremove unincorporated free label and then digested in1ml of a 50 mg/ml papain solution in glass scintillationcounting tubes at 65�C overnight. Aliquots (100 ml) ofthe scaffold digests were sampled from each vial, dilutedto 1ml, and stored at –80�C for later DNA quantifica-tion as described below. To the remaining 900 ml of theconstruct digests, 4.5ml of Hionic-Fluor ScintillationFluid was added (Packard Instrument Company,Meriden, CT) and the [3H]-proline and [35S]-sulfatedisintegrations per minute were measured on a Model1900TR Liquid Scintillation Analyzer (Packard).

2.5. Biochemical composition

The DNA content in the scaffold digests wasdetermined using the fluorescent picoGreen dsDNAquantification assay (Molecular Probes, Eugene, OR).Sulfated glycosaminoglycan (S-GAG) content of scaf-fold digests was measured using a modified dimethyl-methylene blue (DMB) assay in 96 well plates. Alginatedisk digests were analyzed using a pH 1.5 dye, aspreviously described [36], while agarose and gelatindigests were analyzed using a pH 3.0 dye.Hydroxyproline (OHP) content was measured using

the Ehrlich’s reaction assay previously described [37,38].Aliquots (50 ml) of scaffold digests, after proper dilu-tions, were hydrolyzed in 6n HCl (Pierce) at 110�C for18 h and then lyophilized. The samples were thenreconstituted in 200 ml of the assay buffer (5 g/l citricacid (monohydrate), 12 g/l sodium acetate (trihydrate),3.4 g/l sodium hydroxide, and 1.2ml/l glacial acetic acidin distilled water, pH 6.0). The reconstituted samplesolutions were subsequently filtered through activated

affold materials

Days in culture Sample sizea

labeling, and

microscopy (CLSM)

1, 7, 14, 28 days n ¼ 3

1, 7, 14, 28 days n ¼ 9

1, 7, 14, 28 days n ¼ 9MB assay)

y)

28 days n ¼ 9

p-wise stress-relaxation in 0, 14, 28 days n ¼ 9

shear stress relaxation)

ep in dynamic shear)

d from cells of three different donors.

ARTICLE IN PRESSH.A. Awad et al. / Biomaterials 25 (2004) 3211–32223214

charcoal. Added to 50 ml of each filtered sample in a 96well plate was 50 ml of 62mm chloramine-T (Sigma). Themixture was incubated at room temperature for 15minto allow oxidation reaction. Oxidized samples were thenmixed with 50 ml of 0.94m dimethylaminobenzaldehyde(p-DMBA) colorimetric solution and incubated at 37�Cfor 30min. The optical densities of the assayed sampleswere measured using a plate reader at 540 nm, andthe OHP content of the samples was computed relativeto a standard curve of trans-4-hydroxy-L-proline(0–300 mg/ml).For immunohistochemical analysis, scaffolds were

fixed in 10% buffered formalin for 24 h at roomtemperature. The fixed scaffolds were dehydrated in agradient of alcohols and then embedded in paraffinblocks. Sections of 7 mm thickness were obtained fromeach scaffold and mounted on microscope slides.Following deparaffinization, rehydration, and endogen-ous peroxidase activity quenching, the sections were pre-digested for 60min at room temperature with 0.25 units/ml chondroitinase ABC, or for 10min at 37�C withpepsin to allow for the retrieval of the chondroitinsulfate (CS) or collagen type II (Coll-II) antigens,respectively. Sections were then incubated with the2B6 monoclonal antibody specific to CS (kind gift fromDr. Virginia Kraus, Duke University Medical Center)overnight at 4�C or with the II-II6B3 antibody specificto Col II (Developmental Studies Hybridoma Bank,Iowa City, IA) for 1 h at 37�C, respectively. Immunos-taining was detected using Histostain-Plus Kit for AEC(Zymed Laboratories Inc., San Francisco, CA).

2.6. Biomechanical properties

The elastic compressive modulus was determinedfrom equilibrium stress–strain curves generated fromstepwise stress relaxation tests in unconfined compres-sion at strains of 4%, 8%, 12%, and 16%. Similarly, theelastic shear modulus was determined from equilibriumshear stress–strain curves from stepwise shear stress-relaxation (pure torsion) experiments at shear strainsof (0.03, 0.04, and 0.05 rad). Following the stress-relaxation tests, the rheological properties of theconstructs were determined by subjecting the samplesto oscillatory shear strain gðtÞ ¼ go:sinðotÞ of afixed amplitude (go ¼ 0:05 rad) and varying frequency(1–100 rad/s). The resultant oscillatory shear stresssðtÞ ¼ so:sinðot þ dÞ was recorded and the rheologicalproperties such as the complex shear modulusjG�ðoÞj2 ¼ ½G0ðoÞ2 þ ½G00ðoÞ2 and the loss angle d weredetermined for each of the applied frequencies; where G0

is the storage modulus [G0ðoÞ ¼ so:cosðdðoÞÞ=go] and G00

is the loss modulus [G00ðoÞ ¼ so:sinðdðoÞÞ=go]. Biome-chanical tests were performed in a bath of DMEM atroom temperature using an ARES Rheometrics System(Rheometric Scientific, Piscataway, NJ).

2.7. Statistical analysis

Analysis of variance with Student–Newman–Keuls(SNK) multiple ranges tests were used to compare thedifferent biomaterials and culture conditions (a ¼ 0:05).Data was further examined in a multiple linearregression context to determine which of the biochem-ical parameters (DNA, OHP, S-GAG) or combinationof parameters correlated with the biomechanical proper-ties (E; G; d; |G�|). Statistical analyses were performedusing Statistical Analysis Software (SASs, Cary, NC)and S-Pluss (Insightful Corp. Seattle, WA).

3. Results

3.1. Biological properties

Cell viability and morphology in the different scaffoldmaterials were visualized using confocal laser scanningmicroscopy and the Live-Dead fluorescent probes. Allscaffolds showed relatively uniform distributions of cellswith viability greater than 95% at all time points. Cellsin agarose (Fig. 1a) and alginate (Fig. 1b) scaffoldsdisplayed a spherical morphology that persistedthroughout the culture period, regardless of cultureconditions. In contrast, the cells in the gelatin scaffolds(Fig. 1c) displayed a distinct ‘‘fibroblastic’’ morphology.By day 28, the cells in the gelatin scaffolds proliferatedand became confluent with notable cell-to-cell contactthat was associated with the significant cell-mediatedcontraction of the gelatin disks, with reduction of up to70% and 87% their initial diameters under chondro-genic and control culture conditions, respectively(Fig. 1d). Alginate and agarose disks containing cellsand acellular gelatin disks did not exhibit any contrac-tion. Furthermore, culture conditions did not appear toaffect cell morphology, but cells were sparse in controlconditions compared to chondrogenic conditions.Protein and proteoglycan biosynthesis rates were

quantified by [3H]-proline (Fig. 2a) and [35S]-sulfate(Fig. 2b) incorporation, respectively. Biosynthesis ratesin the hydrogel (agarose and alginate) scaffolds weresignificantly greater in chondrogenic conditions com-pared to control conditions (1.2–20 fold greater; datanot shown). However, for scaffolds grown in chondro-genic conditions, protein and proteoglycan biosynthesisrates were significantly lower for the agarose scaffoldsthan those of the alginate and gelatin scaffoldsthroughout most of the culture (po0:05). Whennormalized by the DNA content (Figs. 2c and d),protein and proteoglycan biosynthesis rates in thegelatin scaffolds were significantly greater than agarose(31%) and alginate (68%), respectively, on day 1(po0:05). However, the differences between biosynth-esis rates in the different scaffolds diminished by days 14

ARTICLE IN PRESS

Fig. 1. Cell viability and morphology in the different scaffold materials visualized using confocal laser scanning microscopy and the Live–Dead

fluorescent probes. Cells in agarose (a) and alginate (b) scaffolds displayed a spherical morphology that persisted throughout the culture period,

regardless of culture conditions. By contrast, the cells in the gelatin scaffolds (c) displayed a distinct ‘‘fibroblastic’’ morphology at day 7. By day 28,

the cells in the gelatin scaffolds proliferated and became confluent with significant cell–cell contact as they exerted considerable contraction of the

scaffolds (d).

H.A. Awad et al. / Biomaterials 25 (2004) 3211–3222 3215

and 28. Regardless of culture conditions or scaffoldbiomaterial, incorporation rates decreased significantlywith time (po0:05).

3.2. Biochemical composition

Biochemical analysis of the scaffolds was performedto quantify the DNA, S-GAG, and OHP content ofthe scaffolds at different times in culture. The DNA,S-GAG, and OHP content of scaffolds grown inchondrogenic conditions were significantly higher thantheir control conditions counterparts on days 7, 14,and 28. For scaffolds grown in chondrogenic conditions,the DNA content of gelatin scaffolds was 37–51%greater than agarose and alginate scaffolds on days 14and 28 (Fig. 3a, po0:05). DNA content increased,reaching peak values of nearly 4.3 mg/scaffold on day 7for the agarose and alginate and 6.5 mg/scaffold on day14 for the gelatin with insignificant declines afterwards.For scaffolds grown in chondrogenic conditions, therewere no significant differences in S-GAG content amongthe scaffold materials (Fig. 3b), whereas the OHPcontent in gelatin was 28–47% greater than agaroseand alginate on days 14 and 28 (Fig. 3c, po0:05).

Furthermore, the S-GAG and OHP content increasedsignificantly by 2.5–9 fold between days 1 and 28 for allscaffold materials grown in chondrogenic conditions.When normalized by DNA content, S-GAG (Fig. 4a)and OHP (Fig. 4b) content increased significantlybetween days 1 and 28 for all scaffold materials grownin chondrogenic conditions (po0:05). However, ingeneral, there were no significant differences betweenthe different scaffold materials.The accumulation of cartilage matrix macromolecules

was evident in the agarose and alginate hydrogel in diskscultured under chondrogenic conditions, as demon-strated by the positive immunohistochemical stainingagainst the 2B6 epitope of chondroitin sulfate and typeII collagen (Fig. 5). The staining was most intense in thepericellular matrix, characteristics associated with cellsfound in native cartilage. Positive staining against thesame antigens was also observed in the gelatin scaffolds.In areas of sparse cell density where little contractionhad occurred, the staining was confined to regions ofneomatrix within the folds of the scaffold (Fig. 5c). Bycontrast, in regions where significant contraction hasoccurred, there was intense staining of both antigens in ahyper cellular matrix (Fig. 5f).

ARTICLE IN PRESS

Fig. 2. Protein and proteoglycan synthesis rates in chondrogenic culture conditions, quantified by [3H]-proline (a) and [35S]-sulfate (b) incorporation,

respectively, were significantly higher for the gelatin scaffolds compared to agarose and alginate cultured in chondrogenic conditions early in culture.

When normalized by the DNA content (c and d), the differences between biosynthesis rates in the different scaffolds diminished especially at later

times in culture. Data presented are mean7standard deviation. � po0:05; �� po0:01:

H.A. Awad et al. / Biomaterials 25 (2004) 3211–32223216

3.3. Biomechanical properties

The stress–strain behavior of all scaffold materialswas linear in both compression and shear stress-relaxation experiments over the range of strains used.There were no significant effects of culture conditions onthe mechanical properties of the hydrogel materials(agarose and alginate), however, after 28 days gelatinscaffolds grown in chondrogenic conditions had equili-brium compressive and shear moduli that were 86%,and 160%, respectively, greater than gelatin scaffoldsgrown in control conditions (po0:05; data not shown).The equilibrium compressive modulus (Fig. 6a) showeda 48% and 53% softening between days 0 and 14 for theagarose and alginate disks, respectively (po0:05). Bycontrast, the equilibrium compressive modulus ofgelatin scaffolds increased by 3.9 and 4.6 fold of dayzero values by days 14 and 28, respectively (po0:05).The equilibrium shear modulus (Fig. 6b) increased

significantly over time in chondrogenic culture by 2.6,1.8 and 6 folds for the agarose, alginate, and gelatinscaffolds, respectively. Likewise, the complex shearmodulus at o ¼ 10 rad/s and go ¼ 0:05 (Fig. 6c)increased over time by 2.5, 1.8 and 8.3 folds for theagarose, alginate, and gelatin scaffolds, respectively. Bythe end of 28 days in chondrogenic culture, the

equilibrium shear modulus of alginate and gelatin was22% and 67% of agarose (po0:05). Similarly, thedynamic shear modulus at o ¼ 10 rad/s (|G�(10 rad/s)|)of alginate and gelatin was 22% and 79% of agarose,respectively (po0:05) (Fig. 7a). Values for the complexshear modulus jG�ðoÞj showed linear trends with thelogarithm of frequency. The loss angle (d) showed nospecific trends with frequency (Fig. 7b). The loss anglefor all scaffold materials was less than 15�, indicatingthat all scaffolds tested behave like viscoelastic solids.Multiple linear regression analysis suggested that

increases in G; and jG�j; were significantly associatedwith increases in SGAG content (Table 2, po0:05).Increases in E and d were associated with increases inOHP content, though not significantly (Table 2,p ¼ 0:09).

4. Discussion

Cellular based tissue engineering approaches haveincreasingly used adult stem cells from different sourcesincluding bone marrow [39–42], trabecular bone [43],muscle [44], and adipose tissue [26,31–33,45–48]. Despitethe many advantages of these abundant and accessiblecells, progress in their utility in tissue engineering has

ARTICLE IN PRESS

Fig. 3. Biochemical analysis of the scaffolds at different times in chondrogenic culture conditions. The DNA content (a) of gelatin scaffolds was

significantly higher than agarose and alginate on days 14 and 28. DNA content in all scaffolds increased initially reaching peak values on day 7 for

the agarose and alginate and on day 14 for the gelatin with insignificant declines afterwards. There were no significant differences in s-GAG content

among the scaffold materials (b). OHP content in gelatin was significantly higher than agarose and alginate on days 14 and 28 (c). Data presented are

mean7standard deviation. � po0:05; �� po0:01:

H.A. Awad et al. / Biomaterials 25 (2004) 3211–3222 3217

been limited by our ability to exercise precise controlover the cells’ differentiation potential. While much ofthe stem cell based-tissue engineering research has beenfocused on controlling differentiation using solublechemical factors (such as growth factors) and/ormanipulating the mechanical signals to which the cellsare exposed, less attention has been paid to theimportance of the biomaterial scaffold in regulatingdifferentiation and tissue growth [10,11]. In this study,we corroborated our previous findings that hADAS cellscan differentiate into a chondrocytic phenotype underdefined culture conditions [26,49]. We further demon-strated that the functional properties of tissue engi-neered-cartilage vary with culture conditions, culturetime, and the choice of the scaffolding biomaterial.

Culturing the hADAS cell-laden constructs in chon-drogenic media containing TGF-b1 significantly in-creased DNA, S-GAG, and OHP content over controlconditions by nearly 3 fold in the hydrogel materials and1.5 fold in the gelatin disks. In general, there were nosignificant differences in the S-GAG content of allscaffolds cultured in chondrogenic conditions, althoughthe DNA and OHP contents in the gelatin scaffolds wasgreater than those in the hydrogels late in culture.Furthermore, the biosynthesis rates of proteins andproteoglycans were significantly higher for gelatin diskscompared to the agarose and alginate hydrogels. Whilethese results are similar to the previous observation thatTGF-b1 stimulates chondrogenic differentiation ofadult stem cells [49], they also imply that the cells’

ARTICLE IN PRESS

Fig. 4. S-GAG and hydroxyproline content normalized by DNA

content. The normalized s-GAG (a) and hydroxyproline (b) content

increased significantly between days 1 and 28 for all scaffold materials.

However, in general, there were no significant differences between the

different scaffold materials. Data presented are mean7standarddeviation. � po0:05:

Fig. 5. Immunohistochemical sections of the agarose (a,d), alginate (b,e), a

Sections were stained for antibodies against the 2B6 epitope of chondroitin

Scale bar 100m).

H.A. Awad et al. / Biomaterials 25 (2004) 3211–32223218

response to chondrogenic mediators may depend on thephysical and biological properties of the biomaterialscaffold. These properties may include the diffusionrates that regulate nutrient and metabolic waste trans-port, the ability to regulate the cellular morphology thatis thought to affect differentiation, and the presence ofbioactive ligands that can provide anchorage sites forcell attachment.We have recently shown that molecular diffusion

coefficients in agarose, alginate, and gelatin constructsseeded with hADAS cells, measured by fluorescencerecovery after photobleaching (FRAP) of dextranmolecules of equal or greater size than culture nutrientsand growth factors (70 kDa), are at least twice those innative cartilage even after 28 days in culture with nosignificant differences among the constructs [50]. Theimplications of these findings are two fold: (1) thedifferences in molecular diffusion kinetics do not fullyexplain the differences in cell and tissue growth in theseconstructs and (2) the transport of nutrients andmetabolites to cells within the constructs is not hinderedin the early stages of tissue generation.The ability of the scaffolds to biologically interact

with the cells, however, may explain some of thedifferences in tissue growth within the constructs. Theentrapment of the cells in the hydrogels imposed aspherical cellular morphology, whereas the cells seededin gelatin displayed various morphologies but werepredominantly of fibroblastic morphology. These find-ings, together with the biochemical and biologicalproperties we measured, suggest that the beneficialeffects of spherical cellular morphology in chondrogen-esis may be hindered in the absence of bioactive cellattachment ligands within the matrix. In addition, theimportance of providing a natural substrate for cellattachment is manifested by the cell-mediated contrac-tion of the gelatin scaffolds and the concomitant

nd (c, f) scaffolds cultured up to 28 days in chondrogenic conditions.

sulfate (CS; 20� , Scale bar 50m) and Collagen type II (Coll-II; 40� ,

ARTICLE IN PRESS

Table 2

Multiple linear regression correlations between the biomechanical

properties and the biochemical composition of the tissue-engineered

cartilage constructs

Biomechanical Biochemical Slope Intercept R2 p

E OHP 30.56 8.67 0.28 p ¼ 0:08SGAG 29.2 7.69 0.22 p ¼ 0:12SGAG�OHP 109.11 9.51 0.19 p ¼ 0:16

G OHP 10.83 1.56 0.2 p ¼ 0:15SGAG 14.89 0.87 0.32 po0:05�

SGAG�OHP 56.55 1.67 0.28 p ¼ 0:07d OHP 21.25 5.44 0.26 p ¼ 0:09

SGAG 13.96 5.42 0.10 p ¼ 0:32SGAG�OHP 94.88 5.82 0.28 p ¼ 0:08

|G�| OHP 14.98 1.83 0.24 p ¼ 0:10SGAG 19.62 0.81 0.36 po0:05�

SGAG�OHP 76.72 1.99 0.34 po0:05�

Fig. 6. Biomechanical properties of the scaffold materials at different times in

and alginate showed no improvements between days 0 and 28 following a sign

of gelatin scaffolds increased significantly with time reaching values compa

dynamic shear modulus (c), at o ¼ 10 rad/s and go ¼ 0:05; increased significanagarose. Data points represent the mean7SEM. �� po0:01:

H.A. Awad et al. / Biomaterials 25 (2004) 3211–3222 3219

increases in cell proliferation and collagen synthesis,which supports previous findings in the literature[51,52]. Moreover, recent studies have demonstratedthat overcoming the inert nature of the alginatehydrogel by inserting RGD-containing peptide se-quences promotes cell multiplication and cellular andstructural organization [53].The growth of the tissue engineered cartilage con-

structs appears to have progressed in two stages thatdepend upon the biomaterial scaffold. The first stage, acell growth phase, was characterized by increased cellproliferation and lasted up to 7 days in the hydrogelmaterials and continued up to 14 days in the gelatinscaffolds. The second stage can be described as a celldifferentiation and tissue growth stage, which wascharacterized by decreased proliferation and increased

chondrogenic culture. The elastic compressive modulus (a) of agarose

ificant decrease on day 14. By contrast, the elastic compressive modulus

rable with agarose on day 28. The elastic shear modulus (b) and the

tly with time for all scaffold materials, and was significantly highest for

ARTICLE IN PRESS

Fig. 7. Typical dynamic data for the frequency sweep (shear) response for the different scaffold materials at go ¼ 0:05 and o ¼ 12100 rad/s.Scaffolds were cultured 28 days in chondrogenic conditions. Data points represent the mean7SEM.

H.A. Awad et al. / Biomaterials 25 (2004) 3211–32223220

collagen and proteoglycan deposition. An improvedunderstanding of the interactions between the cell andthe biomaterial scaffold (cell shape, cell–cell contact,cytoskeletal changes, and cell–matrix interactions) mayimprove the ability to control the switch betweencellular proliferation and differentiation [51].The compressive Young’s modulus of the hydrogel

scaffolds (alginate and agarose) decreased in valuebetween days 0 and 14, suggesting that the hydrogelshave experienced softening, possibly due to a loss of thecross-linking Ca2+ cations [54] in alginate and a loss ingel stability due to thermal factors in agarose. However,the subsequent increase in the compressive moduli inagarose and alginate between days 14 and 28 is likelydue to increases in new matrix synthesis. The compres-sive Young’s modulus of the gelatin scaffolds increasedsignificantly over time, reaching values comparable toagarose by day 28. The mechanism behind the stiffeningof the gelatin scaffolds likely involves the packing of thescaffold material as a result of the cell-mediatedcontraction and the deposition of matrix macromole-cules within the scaffold. Similarly, the equilibrium anddynamic shear moduli increased significantly over timefor all scaffold materials, with agarose and gelatinhaving shear moduli almost 3 times greater thanalginate. However, it should be noted that the compres-sive and shear moduli of these scaffolds are on the orderof 5% or less of those of native cartilage [55,56]. Theseresults are not different from observations reported byothers [17,54,57], although it was recently suggested thatthe mechanical properties of agarose disks can beimproved by increasing cell seeding density and theapplication of a compressive loading regimen tostimulate neotissue synthesis [17,58]. Even with the useof large numbers of chondrocytes and prolonged cultureperiods in these studies, the application of mechanicalloading improved the functional properties of thechondrocyte-seeded agarose constructs to no more than20% of native cartilage [58]. These studies, whiledemonstrating the importance of in vitro mechanical

conditioning and cell seeding density, indirectly under-score the importance of the inherent mechanical proper-ties of the biomaterial scaffold and suggest that even theuse of large numbers of cells and prolonged cultureperiods might not be sufficient to overcome themechanical deficiencies of the hydrogels.Our results also indicate that increases in G and jG�j

for all scaffolds are significantly correlated withincreases in S-GAG content and with the interactionbetween S-GAG and OHP (Table 2), but not with OHPalone. This intriguing finding is similar to previousstudies using chondrocytes in agarose disks [58] andgives more credence to the presence of a structure–function relationship in these tissues. Although ouranalysis did not examine the structure of the developingcollagen and proteoglycan networks, our data indicatethat manipulating the composition and structure ofthese tissue-engineered constructs may have importantimplications on the construct’s ability to assume theirmechanical functions.In conclusion, it is quite apparent that the biomaterial

scaffold of choice influences the growth and differentia-tion of adult stem cells. Biologically active biopolymers,such as gelatin, have distinct advantages stemming fromthe fact that they modulate cell functions in mannersthat can be exploited to create biologically functionaltissue-engineered grafts. While neither of the biomater-ials studied approached native cartilage mechanicalproperties they demonstrated significant composition-function relationships that could provide importantclues for engineering functional tissues.

Acknowledgements

This study was supported in part by Artecel Sciences,Inc., the North Carolina Biotechnology Center, theKenan Institute, and NIH grant AR49294. We wouldlike to thank Dr. Lori Setton and Charlene Flahiff for

ARTICLE IN PRESSH.A. Awad et al. / Biomaterials 25 (2004) 3211–3222 3221

help with the biomechanical testing; and Julie Fuller andSteve Johnson for help with the histology.

References

[1] Mooney DJ, Mikos AG. Growing new organs. Sci Am

1999;280:60–5.

[2] Chaikof EL, Matthew H, Kohn J, Mikos AG, Prestwich GD, Yip

CM. Biomaterials and scaffolds in reparative medicine. Ann NY

Acad Sci 2002;961:96–105.

[3] Naughton GK. From lab bench to market: critical issues in tissue

engineering. Ann NY Acad Sci 2002;961:372–85.

[4] Langer R. Biomaterials in drug delivery and tissue engineering:

one laboratory’s experience. Acc Chem Res 2000;33:94–101.

[5] Vacanti CA, Vacanti JP. Bone and cartilage reconstruction with

tissue engineering approaches. Otolaryngol Clin North Am

1994;27:263–76.

[6] Guilak F, Butler DL, Goldstein SA. Functional tissue engineer-

ing: the role of biomechanics in articular cartilage repair. Clin

Orthop 2001;391(Suppl:S):295–305.

[7] Butler DL, Goldstein SA, Guilak F. Functional tissue engineer-

ing: the role of biomechanics. J Biomech Eng 2000;122:570–5.

[8] Naughton G. An industry imperiled by regulatory bottlenecks.

Nat Biotechnol 2001;19:709–10.

[9] Gimble J, Guilak F. Adipose-derived adult stem cells: isolation,

characterization, and differentiation potential. Cytotherapy

2003;5:362–9.

[10] Awad H, Erickson G, Guilak F. Biomatrials for cartilage tissue

engineering. In: Lewandrowski K-U, Wise D, Trantolo D,

Gresser J, Yaszemski M, Altobelli D, editors. Tissue engineering

and biodegradable equivalents: Scientific and clinical applications.

New York: Marcel Dekker Inc; 2002. p. 267–99.

[11] Hunziker EB. Articular cartilage repair: basic science and clinical

progress. A review of the current status and prospects. Osteoarthr

Cartilage 2002;10:432–63.

[12] Freed LE, Marquis JC, Nohria A, Emmanual J, Mikos AG,

Langer R. Neocartilage formation in vitro and in vivo using cells

cultured on synthetic biodegradable polymers. J Biomed Mater

Res 1993;27:11–23.

[13] Freed LE, Vunjak-Novakovic G, Biron RJ, Eagles DB, Lesnoy

DC, Barlow SK, Langer R. Biodegradable polymer scaffolds for

tissue engineering. Biotechnology 1994;12:689–93.

[14] Rowley JA, Madlambayan G, Mooney DJ. Alginate hydrogels

as synthetic extracellular matrix materials. Biomaterials 1999;20:

45–53.

[15] Paige KT, Cima LG, Yaremchuk MJ, Vacanti JP, Vacanti CA.

Injectable cartilage. Plast Reconstr Surg 1995; 96:1390–8; discus-

sion 1399–400.

[16] Woolverton C, Fulton J, Lopina S, Landis W. Mimicking the

natural tissue environment. In: Lewandrowski K-U, Wise D,

Trantolo D, Gresser J, Yaszemski M, Altobelli D, editors. Tissue

engineering and biodegradable equivalents: scientific and clinical

applications. New York: Marcel Dekker, Inc; 2002. p. 43–75.

[17] Mauck RL, Soltz MA, Wang CC, Wong DD, Chao PH, Valhmu

WB, Hung CT, Ateshian GA. Functional tissue engineering of

articular cartilage through dynamic loading of chondrocyte-

seeded agarose gels. J Biomech Eng 2000;122:252–60.

[18] Buschmann MD, Gluzband YA, Grodzinsky AJ. Hunziker EB:

mechanical compression modulates matrix biosynthesis in chon-

drocyte/agarose culture. J Cell Sci 1995;108:1497–508.

[19] Paige KT, Cima LG, Yaremchuk MJ, Schloo BL, Vacanti JP,

Vacanti CA. De novo cartilage generation using calcium alginate-

chondrocyte constructs. Plast Reconstr Surg 1996;97:168–80.

[20] van Susante JL, Buma P, van Osch GJ, Versleyen D, van der

Kraan PM, van der Berg WB, Homminga GN. Culture of

chondrocytes in alginate and collagen carrier gels. Acta Orthop

Scand 1995;66:549–56.

[21] Lee DA, Noguchi T, Knight MM, O’Donnell L, Bentley G, Bader

DL. Response of chondrocyte subpopulations cultured within

unloaded and loaded agarose. J Orthop Res 1998;16:726–33.

[22] Buschmann MD, Gluzband YA, Grodzinsky AJ, Kimura JH,

Hunziker EB. Chondrocytes in agarose culture synthesize a

mechanically functional extracellular matrix. J Orthop Res

1992;10:745–58.

[23] Fragonas E, Valente M, Pozzi-Mucelli M, Toffanin R, Rizzo R,

Silvestri F, Vittur F. Articular cartilage repair in rabbits by using

suspensions of allogenic chondrocytes in alginate. Biomaterials

2000;21:795–801.

[24] Chubinskaya S, Huch K, Schulze M, Otten L, Aydelotte MB, Cole

AA. Gene expression by human articular chondrocytes cultured in

alginate beads. J Histochem Cytochem 2001;49:1211–20.

[25] Chang SC, Rowley JA, Tobias G, Genes NG, Roy AK, Mooney

DJ, Vacanti CA, Bonassar LJ. Injection molding of chondrocyte/

alginate constructs in the shape of facial implants. J Biomed

Mater Res 2001;55:503–11.

[26] Erickson GR, Gimble JM, Franklin DM, Rice HE, Awad H,

Guilak F. Chondrogenic potential of adipose tissue-derived

stromal cells in vitro and in vivo. Biochem Biophys Res Commun

2002;290:763–9.

[27] Diduch DR, Jordan LC, Mierisch CM, Balian G. Marrow

stromal cells embedded in alginate for repair of osteochondral

defects. Arthroscopy 2000;16:571–7.

[28] Ponticiello MS, Schinagl RM, Kadiyala S, Barry FP. Gelatin-

based resorbable sponge as a carrier matrix for human

mesenchymal stem cells in cartilage regeneration therapy.

J Biomed Mater Res 2000;52:246–55.

[29] Pesakova V, Stol M, Adam M. Comparison of the influence of

gelatine and collagen substrates on growth of chondrocytes. Folia

Biol (Praha) 1990;36:264–70.

[30] Nehrer S, Breinan HA, Ramappa A, Shortkroff S, Young G,

Minas T, Sledge CB, Yannas IV, Spector M. Canine chondrocytes

seeded in type I and type II collagen implants investigated in vitro.

J Biomed Mater Res 1997;38:95–104.

[31] Wickham MQ, Erickson GR, Gimble JM, Vail TP, Guilak F.

Multipotent stromal cells derived from the infrapatellar fat pad of

the knee. Clin Orthop 2003; 196–212.

[32] Halvorsen YD, Bond A, Sen A, Franklin DM, Lea-Currie YR,

Sujkowski D, Ellis PN, Wilkison WO, Gimble JM. Thiazolidine-

diones and glucocorticoids synergistically induce differentiation

of human adipose tissue stromal cells: biochemical, cellular, and

molecular analysis. Metabolism 2001;50:407–13.

[33] Halvorsen YD, Franklin D, Bond AL, Hitt DC, Auchter C,

Boskey AL, Paschalis EP, Wilkison WO, Gimble JM. Extra-

cellular matrix mineralization and osteoblast gene expression by

human adipose tissue-derived stromal cells. Tissue Eng 2001;7:

729–41.

[34] Johnstone B, Hering TM, Caplan AI, Goldberg VM, Yoo JU. In

vitro chondrogenesis of bone marrow-derived mesenchymal

progenitor cells. Exp Cell Res 1998;238:265–72.

[35] Yoo JU, Barthel TS, Nishimura K, Solchaga L, Caplan AI,

Goldberg VM, Johnstone B. The chondrogenic potential of

human bone-marrow-derived mesenchymal progenitor cells.

J Bone Joint Surg Am 1998;80:1745–57.

[36] Enobakhare BO, Bader DL, Lee DA. Quantification of sulfated

glycosaminoglycans in chondrocyte/alginate cultures, by use of

1,9-dimethylmethylene blue. Anal Biochem 1996;243:189–91.

[37] Stegemann H, Stalder K. Determination of hydroxyproline. Clin

Chim Acta 1967;18:267–73.

[38] Neidert MR, Lee ES, Oegema TR, Tranquillo RT. Enhanced

fibrin remodeling in vitro with TGF-beta1, insulin and plasmin

for improved tissue-equivalents. Biomaterials 2002;23:3717–31.

ARTICLE IN PRESSH.A. Awad et al. / Biomaterials 25 (2004) 3211–32223222

[39] Caplan AI. Mesenchymal stem cells. J Orthop Res 1991;9:641–50.

[40] Majumdar MK, Banks V, Peluso DP, Morris EA. Isolation,

characterization, and chondrogenic potential of human bone

marrow-derived multipotential stromal cells. J Cell Physiol

2000;185:98–106.

[41] Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R,

Mosca JD, MoormanMA, Simonetti DW, Craig S, Marshak DR.

Multilineage potential of adult human mesenchymal stem cells.

Science 1999;284:143–7.

[42] Caterson EJ, Nesti LJ, Danielson KG, Tuan RS. Human marrow-

derived mesenchymal progenitor cells: isolation, culture expan-

sion, and analysis of differentiation. Mol Biotechnol 2002;20:

245–56.

[43] Noth U, Osyczka AM, Tuli R, Hickok NJ, Danielson KG, Tuan

RS. Multilineage mesenchymal differentiation potential of human

trabecular bone-derived cells. J Orthop Res 2002;20:1060–9.

[44] Peng H, Huard J. Stem cells in the treatment of muscle and

connective tissue diseases. Curr Opin Pharmacol 2003;3:329–33.

[45] Safford KM, Hicok KC, Safford SD, Halvorsen YD, Wilkison

WO, Gimble JM, Rice HE. Neurogenic differentiation of murine

and human adipose-derived stromal cells. Biochem Biophys Res

Commun 2002;294:371–9.

[46] Zuk PA, Zhu M, Mizuno H, Huang J, Futrell JW, Katz AJ,

Benhaim P, Lorenz HP, Hedrick MH. Multilineage cells from

human adipose tissue: implications for cell-based therapies. Tissue

Eng 2001;7:211–28.

[47] Zuk PA, Zhu M, Ashjian P, De Ugarte DA, Huang JI, Mizuno H,

Alfonso ZC, Fraser JK, Benhaim P, Hedrick MH. Human

adipose tissue is a source of multipotent stem cells. Mol Biol Cell

2002;13:4279–95.

[48] De Ugarte DA, Morizono K, Elbarbary A, Alfonso Z, Zuk PA,

Zhu M, Dragoo JL, Ashjian P, Thomas B, Benhaim P, Chen I,

Fraser J, Hedrick MH. Comparison of multi-lineage cells from

human adipose tissue and bone marrow. Cells Tissues Organs

2003;174:101–9.

[49] Awad HA, Halvorsen Y-D, Gimble JM, Guilak F. Effects of

dexamethasone and TGF-beta1 on the chondrogenic differentia-

tion of adipose derived stromal cells in vitro. Tissue Eng 2003;9:

1301–12.

[50] Leddy HA, Awad HA, Wickham MQ, Guilak F. Molecular

diffusion in tissue-engineered cartilage constructs: Effects of time

and culture conditions. In: Summer Bioengineering Conference,

Key Biscayne, FL, ASME-BED, 2003. p. 933–4.

[51] Ben-Ze’ev A, Farmer SR, Penman S. Protein synthesis requires

cell-surface contact while nuclear events respond to cell shape in

anchorage-dependent fibroblasts. Cell 1980;21:365–72.

[52] Macieira-Coelho A, Azzarone B. Correlation between contrac-

tility and proliferation in human fibroblasts. J Cell Physiol 1990;

142:610–4.

[53] Alsberg E, Anderson KW, Albeiruti A, Rowley JA, Mooney DJ.

Engineering growing tissues. Proc Natl Acad Sci USA

2002;99:12025–30.

[54] LeRoux MA, Guilak F, Setton LA. Compressive and shear

properties of alginate gel: effects of sodium ions and alginate

concentration. J Biomed Mater Res 1999;47:46–53.

[55] Leroux MA, Cheung HS, Bau JL, Wang JY, Howell DS, Setton

LA. Altered mechanics and histomorphometry of canine tibial

cartilage following joint immobilization. Osteoarthr Cartilage

2001;9:633–40.

[56] Setton LA, Elliott DM, Mow VC. Altered mechanics of cartilage

with osteoarthritis: human osteoarthritis and an experimental

model of joint degeneration. Osteoarthr Cartilage 1999;7:2–14.

[57] Wong M, Siegrist M, Wang X, Hunziker E. Development of

mechanically stable alginate/chondrocyte constructs: effects of

guluronic acid content and matrix synthesis. J Orthop Res 2001;

19:493–9.

[58] Mauck RL, Seyhan SL, Ateshian GA, Hung CT. Influence of

seeding density and dynamic deformational loading on the

developing structure/function relationships of chondrocyte-seeded

agarose hydrogels. Ann Biomed Eng 2002;30:1046–56.