Biomaterials 24 (2003) 4833–4841

ARTICLE IN PRESS

*Correspondin

87951948.

E-mail addres

0142-9612/03/$ -

doi:10.1016/S014

Collagen/chitosan porous scaffolds with improvedbiostability for skin tissue engineering

Lie Maa, Changyou Gaoa,*, Zhengwei Maoa, Jie Zhoua, Jiacong Shena,Xueqing Hub, Chunmao Hanb

aDepartment of Polymer Science and Engineering, Zhejiang University, Hangzhou 310027, ChinabFaculty of Burn, Second Affiliated Hospital of Zhejiang University, Hangzhou 310027, China

Received 1 November 2002; accepted 13 May 2003

Abstract

Porous scaffolds for skin tissue engineering were fabricated by freeze-drying the mixture of collagen and chitosan solutions.

Glutaraldehyde (GA) was used to treat the scaffolds to improve their biostability. Confocal laser scanning microscopy observation

confirmed the even distribution of these two constituent materials in the scaffold. The GA concentrations have a slight effect on the

cross-section morphology and the swelling ratios of the cross-linked scaffolds. The collagenase digestion test proved that the

presence of chitosan can obviously improve the biostability of the collagen/chitosan scaffold under the GA treatment, where

chitosan might function as a cross-linking bridge. A detail investigation found that a steady increase of the biostability of the

collagen/chitosan scaffold was achieved when GA concentration was lower than 0.1%, then was less influenced at a still higher GA

concentration up to 0.25%. In vitro culture of human dermal fibroblasts proved that the GA-treated scaffold could retain the

original good cytocompatibility of collagen to effectively accelerate cell infiltration and proliferation. In vivo animal tests further

revealed that the scaffold could sufficiently support and accelerate the fibroblasts infiltration from the surrounding tissue.

Immunohistochemistry analysis of the scaffold embedded for 28 days indicated that the biodegradation of the 0.25% GA-treated

scaffold is a long-term process. All these results suggest that collagen/chitosan scaffold cross-linked by GA is a potential candidate

for dermal equivalent with enhanced biostability and good biocompatibility.

r 2003 Elsevier Ltd. All rights reserved.

Keywords: Collagen; Chitosan; Biostability; Cross-link; Tissue engineering

1. Introduction

The skin loss is one of the oldest and still not totallyresolved problems in surgical field. Due to the sponta-neous healing of the dermal defects would not occur, thescar formation for the full thickness skin loss would beinevitable unless some skin substitutes are used. In thepast decades, many skin substitutes such as xenografts,allografts and autografts have been employed for woundhealing. However, because of the antigenicity or thelimitation of donor sites, the skin substitutes mentionedabove cannot accomplish the purpose of the skinrecovery and yet not be used widely [1–5]. Therefore,many studies are turning toward the tissue engineering

g author. Tel.: +86-571-87951108; fax: +86-571-

s: cygao@mail.hz.zj.cn (C. Gao).

see front matter r 2003 Elsevier Ltd. All rights reserved.

2-9612(03)00374-0

approach, which utilizes both engineering and lifescience discipline to promote organ or tissue regenera-tion and to sustain, recover their functions [6–9]. Onecrucial factor in skin tissue engineering is the construc-tion of a scaffold. A three-dimensional scaffold providesan extra cellular matrix analog which functions as anecessary template for host infiltration and a physicalsupport to guide the differentiation and proliferation ofcells into the targeted functional tissue or organ [10,11].An ideal scaffold used for skin tissue engineering shouldpossess the characteristics of excellent biocompatibility,suitable microstructure such as 100–200 mm mean poresize and porosity above 90%, controllable biodegrad-ability and suitable mechanical property [12–15].Collagen is known to be the most promising materialsand have been found diverse applications in tissueengineering for their excellent biocompatibility andbiodegradability. However, the fast biodegrading rate

ARTICLE IN PRESSL. Ma et al. / Biomaterials 24 (2003) 4833–48414834

and the low mechanical strength of the untreatedcollagen scaffold are the crucial problems that limitthe further use of this material. Cross-linking of thecollagen-based scaffolds is an effective method tomodify the biodegrading rate and to optimize themechanical property.For this reason, the cross-linking treatment to

collagen has become one of the most important issuesfor the collagen-based scaffolds. Currently, there aretwo different kinds of cross-linking methods employedin improving the properties of the collagen-basedscaffolds: chemical methods and physical methods.The latter include the use of photooxidation, dehy-drothermal treatments (DHT) and ultraviolet irradia-tion, which could avoid introducing potential cytotoxicchemical residuals and sustain the excellent biocompat-ibility of the collagen materials [16]. However, most ofthe physical treatments cannot yield high enough cross-linking degree to satisfy the demand of skin tissueengineering. Therefore, the treatments by chemicalmethods are still necessary in almost all cases. Thereagents used in the cross-linking treatment recentlyinvolve traditional glutaraldehyde (GA), 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide (EDAC), polygly-cidyl ether and polyepoxidic resins, etc. [17–21]. GA is akind of bifunctional cross-linking reagents that canbridge amino groups between two adjacent polypeptidechains and has become the predominant choice in skintissue engineering because of its water solubility, highcross-linking efficiency and low cost [22].Chitosan is another biomaterials used in a variety of

biomedical fields such as drug delivery carriers, surgicalthread, and wound healing materials [23]. Due to itsmany advantages for wound healing such as hemostasis,accelerating the tissue regeneration and the fibroblastsynthesis of collagen, many applications of chitosan inskin tissue engineering have been reported [24–27]. Inaddition, chitosan can function as a bridge to increasethe cross-linking efficiency of GA in the collagen-based

COOH

NH2 NH2

(Collagen)

GA

Chitosan

NH2

H2N

NH2

H2NNH2 NH2

COOH

Fig. 1. Schematic presentation of collagen cross-linked

scaffolds owing to the large number of amino groups inits molecular chain (Fig. 1). Hence, one can expect thatless GA could be used in the presence of chitosan andthe potential cytotoxicity of GA might be decreased.Herein we describe the fabrication of collagen porous

scaffold in the presence of 10wt% chitosan, whichfunctions as a cross-linking bridge in the furthertreatment of GA cross-linkage. The microstructure,the swelling capacity, as well as the degradability bothin vivo and in vitro of the collagen/chitosan scaffoldwere investigated. In vitro culture of human dermalfibroblasts and in vivo animal tests demonstrated thatthe scaffolds showed good cytocompatibility andcould effectively guide the infiltration and growth offibroblasts.

2. Materials and methods

2.1. Materials

Chitosan (viscosity average molecular weight MZ:1.0� 105–1.7� 105, 75–85% deacetylation degree), col-lagenase I (278U/mg), rhodamine B isothiocyanate,fluorescein isothiocyanate (FITC) and fluorescein dia-cetate (FDA) were purchased from Sigma. Trypsin(250U/mg) was a commercial product from Amresco.Glutaraldehyde (GA), 25% water solution, was pur-chased from Shanghai Pharm. Co. (China). All otherreagents and solvents are of analytical grade and used asreceived.Collagen type I was isolated from fresh bovine tendon

by trypsin digestion and acetic acid dissolution method.Briefly, after removed the fat and muscle impuritysubstances the bovine tendon was cut into pieces as thinas possible and digested in trypsin solution (0.25%) at37�C for 24 h. The tendon pieces were then incubated in0.5m acetic acid (HAc) at 4�C for 48 h. A tissuetriturator was employed to agitate the swollen tendon

N

N

N

COOH

COOH

N

N

R

CH

CH

N

CH

HC

R

CH

HC

R

Chitosan

NH2

H2N

Chitosan

NH2

H2N

with glutaraldehyde in the presence of chitosan.

ARTICLE IN PRESSL. Ma et al. / Biomaterials 24 (2003) 4833–4841 4835

pieces violently so that the collagen fibers could be welldispersed. The collagen solution was then centrifuged toget rid of insoluble impurities. The supernatant wasprecipitated by 5wt% NaCl solution. The precipitatewas re-dissolved in 0.5m HAc to repeat the same processfor purification. Finally the collagen extraction wasdialyzed with double distilled water for 72 h, changingthe water every 12 h, and then was lyophilized. Thecomposition and purity of the collagen type I wascharacterized and confirmed by UV spectroscopy, IRspectroscopy and amino acid analysis.Rhodamine labeled collagen (Rd-Col) and FITC

labeled chitosan (FITC-Chi) were prepared by mixing0.2mg/ml rhodamine B isothiocyanate or FITC into0.5% (w/v) biomacromolecule solutions at 4�C for 48 h,respectively. The free dyes were dialyzed off in 0.05macetic acid solution for 4 weeks.

2.2. Preparation of collagen/chitosan scaffold

Collagen or chitosan was dissolved in 0.5m HAcsolution to prepare a 0.5% (w/v) solution, respectively.The chitosan solution was slowly dropped into collagensuspension in the ratio of 9:1 (collagen:chitosan) andhomogenized to obtain collagen/chitosan blend. Afterdeaerated under vacuum to remove entrapped air-bubbles, the collagen/chitosan blend was injected intoa home-made mould (diameter: 16mm, depth: 2mm),frozen in 70% ethanol bath at �20�C for 1 h and thenlyophilized for 24 h to obtain a porous collagen/chitosanscaffold.

2.3. Cross-linking treatment

To improve the biostability, the collagen/chitosanscaffolds were treated with GA. All scaffolds wererehydrated in 0.05m HAc solutions for 15min firstly,and then were cross-linked in the GA solutions (double-distilled water, pH 5.6) with different concentrations(0.05–0.25%) at 4�C for 24 h [21,28]. After washed withdouble-distilled water (10min� 5 times), the scaffoldswere freeze-dried again to obtain the GA treatedcollagen/chitosan scaffolds.

2.4. Microstructure observation

The microstructure of the scaffolds was observedunder scanning electron microscopy (SEM, Cambridgestereoscan 260) and confocal laser scanning microscopy(CLSM, Biorad 2100). Rd-Col and FITC-Chi were usedfor CLSM detection with double channels’ mode.

2.5. Swelling test

The collagen/chitosan scaffolds were placed intodistilled water at room temperature and the wet weight

(w) of the scaffold was determined after incubated for24 h. The swelling ratio of the scaffolds was defined asthe ratio of weight increase (w�w0) to the initial weight(w0). Each value was averaged from three parallelmeasurements.

2.6. In vitro collagenase degradation

In vitro biodegradation test of the collagen/chitosanscaffolds cross-linked by GA with different concentrations(0–0.25%) was performed by collagenase digestion. Eachkind of scaffolds was immersed in phosphate bufferedsaline (PBS, pH 7.4) containing 100mg/ml (28 units)collagenase (type I, Sigma) at 37�C for 4, 16, 24 and 48h.The degradation was discontinued at the desired timeinterval by incubating the assay mixture in an ice bathimmediately. Following centrifugation at 1500 rpm for10min, the clear supernatant was hydrolyzed with 6m HClat 120�C for 12h. The content of hydroxyproline releasedfrom the scaffold was measured with ultraviolet spectro-scopy [29]. The biodegradation degree is defined as thepercentage of the released hydroxyproline from thescaffolds at different time to the completely degradedone with same composition and same weight.

2.7. Cell culture

Fibroblasts used in this study were isolated fromhuman dermis by collagenase digestion. Briefly, theepidermis and subcutaneous tissue of human skin wereremoved by the scalpel. The residual dermis was dicedinto 0.5–1mm3 sized tissues, and washed with phosphatebuffer saline (PBS, pH 7.4) supplemented with penicillin(100U/ml) and streptomycin (100U/ml) 3 times. Thenthese dermis pieces were placed in a spinner flaskcontaining 10ml of 1mg/ml collagenase (type I, Sigma)in Dulbecco’s modified Eagle medium (DMEM) sup-plemented with penicillin (100U/ml) and streptomycin(100U/ml). After digested in an incubator (37�C, 5%CO2) for 5 h, the digestion was discontinued by addingDMEM supplemented with penicillin (100U/ml), strep-tomycin (100U/ml) and 10% FBS (complete medium).The digesting solution was filtered through a coppermesh (cell strainer, 200 meshes) and then was centri-fuged at 1000 rpm for 10min. The cell suspension werecultured at 37�C, 5% CO2, and 95% humidity in thecomplete medium. The culture medium was changedevery 3 days. Cells were passaged at confluence and the4–8th passage fibroblasts were used for the seeding.The 0.25% GA treated collagen/chitosan scaffold

(both rhodamine-labeled) was immersed in 75% ethanolfor 12 h for sterilization, followed with solvent exchangeby PBS for 6 times. The scaffold was then placed on a24-well polystyrene plate and seeded with 200 ml humandermal fibroblast suspension at a density of 5� 106 cells/ml. After incubation for 4 h, 1ml complete medium was

ARTICLE IN PRESSL. Ma et al. / Biomaterials 24 (2003) 4833–48414836

added and cultured in a 5% CO2 incubator at 37�C for 3

days. After washed with PBS for 2 times, the fibroblastswere stained with 5 mg/ml FDA solution in the incubatorfor 15min. Following with removal of the unreactedFDA with double washing in PBS, 1ml completemedium was then added. The live fibroblasts canmetabolize FDA to form a fluorescence product. Hence,the fibroblasts existed in the scaffolds are distinct fromthe rhodamine labeled scaffolds (red color) by thegeneration of green color under CLSM.

2.8. In vivo animal evaluation

Twelve health rabbits weighing about 2 kg wereobtained from the animal laboratory and were dividedinto four groups randomly. The 0.25% GA treatedcollagen/chitosan (10wt%) scaffolds were sterilized byimmerged into 75% (v/v) ethanol for 30min andwashing with PBS (pH 7.4) (5 times� 5min). Beforeimplantation, the dorsal surface hairs of the rabbit earswere shaved. Then all rabbits were anesthetized byintravenous administration of 20mg/kg ketamine-HCl.The ears of rabbits were sterilized with 5% PVP-I, onwhich subcutaneous pockets were made [30,31]. In everygroup, four scaffolds (0.5� 1 cm2) were embeddedsubcutaneously on the dorsal surface of rabbit ear.Harvests were performed randomly in selected group at3 days, and 1, 2, 4 weeks after implantation. At harvest,the implantation sites were cut in a full thickness manner(including both sides of the ear skin and cartilage).Paraffin sections were stained with hematoxylin-eosin(HE) reagent for histological observations.

2.9. Immunohistochemistry

Sample of 0.25% GA treated collagen/chitosanscaffold after embedded for 28 days was fabricated intoparaffin section. After dewaxed and blocked with 3%(w/v) bovine serum albumin in PBS (pH 7.4) (BSA/PBS)for 20min at 20�C, sections of the scaffold were

50µm 50µm

(a) (b)

Fig. 2. CLSM images of the distribution of chitosan (a) and collagen (b) in th

(b). � 400.

incubated for 12 h at 4�C with mouse anti-bovine typeI collagen IgG (diluted 1:100) and washed with PBS (pH7.4) (3 times, each for 5min). Subsequently, the sectionswere incubated for 30min at 37�C with biotinylatedgoat anti-mouse IgG (diluted 1:300) and washed withPBS (pH 7.4). The slides were then reacted with avidin-conjugated peroxidase (diluted 1:30) at 37�C for 30min.Finally, the sections were displayed with DAB andembedded by paraffin to yield a positive stain. Sectionswere observed under light-microscope.

3. Results and discussion

3.1. Distribution of collagen and chitosan

One of the important purposes adding chitosan isproviding additional amino groups which function asbinding cites to increase the GA cross-linking efficiency.Therefore, the interpenetration of collagen and chitosanin the scaffold is crucial. Exploiting the sequentialscanning mode of CLSM, the distribution of FITC-Chi(Fig. 2a) and Rd-Col (Fig. 2b) in their complex scaffoldwas separately measured at wet state. A merged image isshown in Fig. 2c. The CLSM observations indicate thatthe scaffold was indeed composed with chitosan andcollagen which were evenly dispersed through thescaffold. In acidic solution, both collagen and chitosanare positively charged, either forming a real solution (forchitosan) or suspension (for collagen) [32]. Theirmixture in solution is stable and does not precipitateas that for collagen/chondroitin sulfate blend, wherechondroitin sulfate is negatively charged [4,7]. There-fore, sufficient mixing of these two hydrophilic bioma-cromolecules in sub-molecular level can be achieved.

3.2. Morphology

It is known that the microstructure such as pore sizeand its distribution, porosity as well as pore shape has

50µm

(c)

e Rd-Col/FITC-Chi porous scaffold; (c) is the merged image of (a) and

ARTICLE IN PRESSL. Ma et al. / Biomaterials 24 (2003) 4833–4841 4837

prominent influence on cell intrusion, proliferation andfunction in tissue engineering. The cross-sectionmorphologies of the collagen/chitosan scaffolds beforeand after GA treatment are shown in Fig. 3. Theinterconnected 3D porous structure of the scaffolds wasretained after GA treatment; however, some othersignificant changes occurred with respect to pore sizeand morphology. The mean pore size increased from�100 mm, the uncross-linked (Fig. 3a), to >200 mm, thecross-linked scaffolds (Fig. 3b–e). Accompanying withreduction of the fibers in between pores, more sheet-likestructure appeared together with condensed walls. Nobig difference between the cross-linked scaffolds wasobserved, except for which treated with highest GAconcentration (Fig. 3e), where elongated pores wereexisted.The results indicate that the morphology difference is

mainly caused by rehydration and relyophilizationprocess in the GA cross-linking treatment. This addi-tional refreeze-drying can induce the collagen fibers tobe combined again to form sheets, leading to the fusionof some smaller pores to generate larger ones. It has tobe noted that the slight collapse of the scaffold duringthis process should have an opposite effect to the porefusion; i.e., reducing the pore size. Hence, one candeduce from the above results that the fusion effect ismore prominent than the collapse. As a result, the poresare enlarged. On the other hand, this collapse, if not

(a) (b)

(d)

Fig. 3. The cross-section SEM images of collagen/chitosan scaffolds treated w

(c): 0.1% GA; (d): 0.2% GA; (e): 0.25% GA.

occurs homogeneously in 3D, will inevitably produceelongated pores as shown in Fig. 3e.

3.3. Swelling test

The ability of a scaffold to preserve water is animportant aspect to evaluate its property for skin tissueengineering. The swelling ratios of various scaffoldsare shown in Fig. 4. The swelling property of theuncross-linked scaffold was doubled than the GAtreated scaffolds. However, the cross-linked scaffoldsdid not show obvious difference regardless of the GAconcentration.The water-binding ability of the collagen/chitosan

scaffold could be attributed to both of their hydro-philicity and the maintenance of their three-dimensionalstructure. In general, the swelling ratio is decreased asthe cross-linking degree is increased because of thedecrease of the hydrophilic groups [33]. The results inFig. 4 indicate that the primary factor affected theswelling property is the procedure of the GA treatmentother than the GA concentration (hence, the cross-linking degree). As mentioned above, the collapseduring the refreeze-drying procedure will cause thereduction of the porosity, hence, the volume for waterstorage, leading to the decrease of the swelling capacity.However, the absolute value is still over 80 times of its

(c)

(e)

ith different concentration of GA, � 100. (a): control; (b): 0.05% GA;

ARTICLE IN PRESS

0 10 20 30 40 50

0

20

40

60

80

100T

he

bio

deg

rad

atio

n d

egre

e (%

)

Biodegrading time (h)

control0.05%0.1%0.2%

0.25%

Fig. 6. The effect of GA concentrations on the biodegradability of the

collagen/chitosan scaffolds. (a): control; (b): 0.05% GA; (c): 0.1% GA;

(d): 0.2% GA; (e): 0.25% GA.

0 0.05 0.1 0.2 0.250

2

4

6

8

10

12

14

16

18

20

Th

e sw

ellin

g r

atio

s (1

000%

)

GA concentrations (%)

Fig. 4. The effect of GA concentrations on the swelling ratios of the

collagen/chitosan scaffolds. Values are mean 7S.D. (n=3).

col col-GA col/chi col/chi-GA0

20

40

60

80

100

Th

e b

iod

egra

dat

ion

deg

ree

(%)

Fig. 5. The biodegradation degree of the pure collagen scaffolds and

the collagen/chitosan scaffolds (uncross-linked or GA treated) after

incubated in 100mg/ml (30 units) collagenase for 12 h. Values are mean7S.D. (n=3).

L. Ma et al. / Biomaterials 24 (2003) 4833–48414838

initial weight after GA treatment, which is high enoughfor skin tissue engineering.

3.4. In vitro biodegradability

The Fig. 5 compares the biodegradation degree of thepure collagen scaffold and the collagen/chitosan scaffoldbefore and after GA treatment. After incubated incollagenase solution for 12 h, the pure collagen scaffold(col) had been thoroughly biodegraded. The addition ofchitosan (col/chi) can somewhat increase the biostabil-ity, where slight lower biodegradation degree, 92.1%,was found. After cross-linked with 0.25% GA, thebiostability of the pure collagen scaffold (col-GA) wasgreatly enhanced, where only 12.8% was degraded in12 h. Owing to the expected larger cross-linking degree(Fig. 1), the ability to resist collagenase degradation wasfurther enhanced for the chitosan-combined scaffold.These results reveal that both the addition of chitosanand GA cross-linking are indispensable for improvingthe scaffold biostability and the presence of chitosan canobviously improve the biostability of the collagen/chitosan scaffold under the GA treatment, wherechitosan might function as a cross-linking bridge.The dynamic degradation of the collagen/chitosan

scaffolds cross-linked by different concentrations of GAis illustrated in Fig. 6. The uncross-linked collagen/chitosan scaffold was biodegraded so fast that itsbiodegradation degree had achieved to 41.5% justtreated by the collagenase solution for 2 h. Afterbiodegradation for 16 h, the uncross-linked scaffoldhad been dissolved in the collagenase solution thor-oughly. Fig. 6 shows that the biostability of the GAtreated scaffolds were better than the uncross-linkedone. For example, even treated with the lowest GAconcentration (0.05%), the biodegradation degreeof the scaffold was only 6.3% in 4 h. When the GA

concentration was up to 0.1%, the biodegradationdegree increased very slowly with the degrading. Thehighest biodegradation degree was just 26.1% after 48 h.Fig. 6 shows also that with the GA concentrationincrease, the effect of GA concentration on theimprovement of the biostability was slowed down.

3.5. Cell culture

Cell infiltration and proliferation are crucial for ascaffold to support and guide tissue regeneration. Fig. 7represents the CLSM images of the human fibroblastscultured for 3 days in the collagen/chitosan scaffoldtreated by 0.25% GA. Exploiting the sequential scan-ning mode, a great number of fibroblasts (Fig. 7a) wereeasily distinguished from the scaffold (Fig. 7b). Themerged image (Fig. 7c) reveals that the fibroblasts wereadhered on the walls of the scaffold tightly with typical

ARTICLE IN PRESS

50µm50µm 50µm

(a) (c)(b)

Fig. 7. CLSM images of human dermal fibroblasts (a) cultured over the collagen/chitosan scaffold (b, rhodamine-labeled) for 3 days; (c) is the

merged image of (a) and (b). � 400.

(a) (b)

(d)(c)

Fig. 8. The histological response to the collagen/chitosan scaffolds treated with 0.25% GA, after embedded in rabbit ear for different time, � 100.

(a): 3 days; (b): 7 days; (c): 14 days; (d): 28 days. Bar indicates 200mm. M: the implanted collagen/chitosan scaffold. T: the subcutaneous connective

tissue. Arrowhead: the infiltrated fibroblasts.

L. Ma et al. / Biomaterials 24 (2003) 4833–4841 4839

shuttle-like morphology. This result proves that thechitosan-combined and GA-treated scaffold preservesthe original good cytocompatibility of collagen. Poten-tial cytotoxicity of GA residue was not evidenced. Thisensures the further study of the tissue response to thescaffolds in vivo.

3.6. Histological examination

The histological results of the 0.25% GA-treatedscaffold embedded in the rabbit ear for different timeare shown in Fig. 8. The pure collagen scaffold was easyto lose its contour structure and biodegraded quickly in3 days because of its low stability. On the contrary, the

structure of the 0.25% GA-treated scaffold was retainedentirely and a few of fibroblasts and inflammatory cellscould be observed in the scaffold after implanted for3 days (Fig. 8a). After implanted for 7 days, morefibroblasts were grown into the scaffold and theinflammatory cells were still in existence (Fig. 8b).When the test had processed for 14 days, a large numberof fibroblasts were infiltrated into the scaffold. Themorphology of the scaffold was similar to the surround-ing dermal tissue and its structure could not beobviously observed (Fig. 8c). After 28 days’ implanta-tion, the scaffold had almost disappeared and the bloodvessels could be observed (Fig. 8d). These resultsdemonstrate that the collagen/chitosan scaffolds can

ARTICLE IN PRESS

(a) (b)

Fig. 9. The light microscopic (a) and immunostaining images (b) of the 0.25% GA-treated collagen/chitosan scaffold which embedded for 28 days.

� 100. Bar indicates 200mm. Arrowhead: the un-degraded collagen scaffold.

L. Ma et al. / Biomaterials 24 (2003) 4833–48414840

effectively sufficiently support and accelerate the fibro-blasts infiltration from the surrounding tissue. All thein vitro and in vivo results have shown that the collagen/chitosan scaffold treated with GA has a good biocom-patibility.

3.7. Immunohistochemistry

To study the biodegradation behavior of the collagen/chitosan scaffold in vivo, the image of the paraffinsection of the 0.25% GA cross-linked scaffold afterimmunohistochemistry treatment is shown in Fig. 9. TheGA cross-linked collagen/chitosan scaffold could not bedistinguished from the new-formed collagen fiber underroutine paraffin section with light microscope after 28days’ implantation (Fig. 9a). However, the immunohis-tochemistric assay shows that the bovine type I collagenhad been partially preserved though the scaffold hadmacroscopically disappeared (Fig. 9b). This resultindicates that the biodegrading behavior of the GA-treated collagen/chitosan scaffold is a long-term process.The long-term biodegradation of this kind scaffoldin vivo should be studied further.

4. Conclusion

Herein we have described the fabrication of porouscollagen/chitosan scaffold by freeze-drying their mixtureand the further cross-linking with GA. Collagen andchitosan were evenly distributed in the scaffold. The GAtreatment had an influence on the morphology and theswelling property of the scaffold, while no significantdifferences were observed among the scaffolds treatedwith different concentration GA. After addition ofchitosan, the ability to resist the collagenase degradationwas augmented obviously and can be controlled with thechange of the GA concentration. The cell culture andanimal test prove that the GA-treated scaffold retainedthe original good biocompatibility and could induce the

fibroblasts infiltration from the surrounding tissuesuccessfully. Immunohistochemistric assay indicatesthat the biodegrading behavior of the 0.25% GA-treated collagen/chitosan scaffold is a long-term period.In conclusion, the GA-treated collagen/chitosan scaf-fold is a potential candidate for dermal equivalent withenhanced biostability and good biocompatibility.

Acknowledgements

The authors thank Prof. Yiyong Chen for his valuablediscussion. This work was supported by the NaturalScience Foundation of China (50173024) and the MajorState Basic Research Program of China (G1999054305).

References

[1] Boyce ST. Design principles of composition and performance of

cultured skin substitutes. Burns 2001;27:523–33.

[2] Schul III JT, Tompkins RG, Burks JF. Artificial skin. Annu Rev

Med 2000;51:231–44.

[3] Yannas IV. Regeneration templates. In: Bronzino JD, editor. The

biomedical engineering handbook. Boca Raton, FL: CRC Press;

1995. p. 1619.

[4] Yanas IV, Burke JF. Design of an artificial skin I. Basic design

principles. J Biomed Mater Res 1980;14(4):65–81.

[5] Bell E, Ehrlich HP, Battle DJ, Nakatsuji T. Living tissue formed

in vitro and accepted as skin-equivalent tissue of full thickness.

Science 1981;211:1052–4.

[6] Langer R, Vacanti, Joseph P. Tissue engineering. Science

1993;260:920–6.

[7] Yannas IV, Lee E, Orgill DP, Skrabut EM, Murphy GF.

Synthesis and characterization of a model extracellular matrix

that induces partial regeneration of adult mammalian skin. Proc

Nat Acad Sci USA 1989;86:933–7.

[8] Wintermantel E, Mayer J, Blum J, Eckert KL, Luscher P, Mathey

M. Tissue engineering scaffolds using superstructures. Biomater-

ials 1996;17:83–91.

[9] Widmer MS, Mikos AG. Fabrication of biodegradable polymer

scaffolds. In: Patrick Jr CW, Mikos AG, Mclntire LV, editors.

Frontiers in tissue engineering. UK: Rdvood Books Ltd; 1998.

p. 107.

ARTICLE IN PRESSL. Ma et al. / Biomaterials 24 (2003) 4833–4841 4841

[10] Hutmacher DW, Goh JCH, Teoh SH. An introduction to

biodegradable materials for tissue engineering applications. Ann

Acad Med Singapore 2001;30:183–91.

[11] Si-Nae Park, Jong-Chul Park, Hea Ok Kim, Min Jung Song,

Hwal Suh. Characterization of porous collagen/hyaluronic acid

scaffold modified by 1-ethyl-3-(3-dimethylaminopropyl) carbodii-

mide cross-linking. Biomaterials 2002;23:1205–12.

[12] Dagalakis N, Flink J, Stasikelis P, Burke JF, Yannas IV. Design

of an artificial skin III. Control of Pore Structure 1980;14(4):

511–28.

[13] Freyman TM, Yannas IV, Gibson LJ. Cellular materials as

porous scaffolds for tissue engineering. Progr Mater Sci 2001;46:

273–82.

[14] Radhika M, Mary B, Sehgal PK. Cellular proliferation on

desamidated collagen matrices. Comparative Biochem Physiol

Part C 1999;124:131–9.

[15] Chen GP, Ushida Y, Tateishi T. Scaffold design for tissue

engineering. Macromolecular Biosci 2002;2:67–77.

[16] Lee JE, Park JC, Hwang YS, Kin JK, Kim JG, Suh H.

Characterization of UV-irradiated dense/porous collagen mem-

branes: morphology, enzymatic degradation and mechanical

properties. Yonsei Med J 2001;42:172–9.

[17] Khor E. Methods for the treatment of collagenous tissues for

bioprostheses. Biomaterials 1997;18:95–105.

[18] Osborne CS, Reid WH, Grant MH. Investigation into the

biological stability of collagen/chondroitin-6-sulphate gels and

their contraction by fibroblasts and keratinocytes: the effect of

crosslinking agents and diamines. Biomaterials 1999;20:283–90.

[19] Courtman DW, Errett BF, Wilson GJ. The role of crosslinking

in modification of the immune response elicited against xenogenic

vascular acellular matrices. J Biomed Mater Res 2001;55:576–86.

[20] Raymond Z, Pieter JD, van Wachem Pauline B, et al. Successive

epoxy and carbodiimide cross-linking of dermal sheep collagen.

Biomaterials 1999;20:921–31.

[21] Sung WW, Hsu HL, Shih CC, Lin DS. Cross-linking character-

istics of biological tissues fixed with monofunctional or mutifunc-

tional epoxy compounds. Biomaterials 1996;17:1405–10.

[22] Jorge-Herrero E, Fernandez P, Turnay J, et al. Influence of

different chemical cross-linking treatments on the properties

of bovine pericardium and collagen. Biomaterials 1999;20:

539–45.

[23] Shanmugasundaram N, Ravichandran P, Neelakanta Reddy P,

Ramamurty N, Subrata Pal, Panduranga Rao K. Collagen-

chitosan polymeric scaffolds for the in vitro culture of human

epidermoid carcinoma cells. Biomaterials 2001;22(14):1943–1951.

[24] Yong-Woo Cho, Yong-Nam Cho, Sang-Hun Chung, Gyeol Yoo,

Sohk-Won Ko. Water-soluble chitin as a wound healing accel-

erator. Biomaterials 1999;20:2139–45.

[25] Jianbiao Ma, Hongjun Wang, Binglin He, Jiatong Chen.

A preliminary in vitro study on the fabrication and tissue

engineering applications of a novel chitosan bilayer material as

a scaffold of human fetal dermal fibroblasts. Biomaterials 2001;

22:331–6.

[26] Taravel MN, Domard A. Collagen and its interaction with

chitosan II. Influence of the physicochemical characteristics of

collagen. Biomaterials 1995;16:865–71.

[27] Taravel MN, Domard A. Collagen and its interaction with

chitosan III. Some biological and mechanical properties. Bioma-

terials 1996;17:451–5.

[28] Ming-Thau Sheu, Ju-Chun Huang, Geng-Chang Yeh, Hsiu-O

Ho. Characterization of collagen gel solutions and collagen

matrices for cell culture. Biomaterials 2001;22:1713–9.

[29] Pieper JS, Oosterhof A, Dijkstra PJ, Veerkamp JH, van

Kuppevelt TH. Preparation and characterization of porous

crosslinked collagenous matrices containing bioavailable chon-

droitin sulphate. Biomaterials 1999;20:847–58.

[30] Eppley BL. Experimental assessment of the revascularization of

acellular human dermis for soft-tissue augumentation. Plast

Reconstr Surg 2001;107(3):757–62.

[31] Tsuboi DR, Shi CM, Oshita Y, Ogawa H. Endothelin-1 promotes

contraction and healing of wounds. J Dermatol Sci 1995;10(1):82.

[32] Martins VCA, Plepis AMG, Machado AAS. Thermal and

rheological behavior of collagen: chitosan blends. J Thermal

Anal Calorim 2002;67:491–8.

[33] Rehakova M, Bakos D, Vizarova K, Soldan M, Jurickova M.

Properties of collagen and hyaluronic acid composite materials

and their modification by chemical cross-linking. J Biomed Mater

Res 1995;29:1373–9.