composite fibrous membranes of plga and chitosan prepared by coelectrospinning and coaxial...

Composite fibrous membranes of PLGA and chitosanprepared by coelectrospinning and coaxial electrospinning

Lili Wu,1 Hua Li,1 Shuo Li,1 Xiaoran Li,1 Xiaoyan Yuan,1 Xiulan Li,2 Yang Zhang21School of Materials Science and Engineering, and Tianjin Key Laboratory of Composite and Functional Materials,Tianjin University, Tianjin 300072, China2Institute of Orthopedics, Tianjin Hospital, Tianjin 300211, China

Received 4 October 2007; revised 10 September 2008; accepted 17 November 2008Published online 20 February 2009 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jbm.a.32393

Abstract: Membranes made of hybrid poly(lactide-co-gly-colide)/chitosan fibers (H-PLGA/chitosan) and core/shellstructured PLGA/chitosan fibers (C-PLGA/chitosan) wereproduced by coelectrospinning and coaxial electrospinning,respectively. The morphology, mechanical properties, wateruptakes of the electrospun fibrous membranes were charac-terized, and the cytocompatibility of human embryo skinfibroblasts (hESFs) was investigated in comparison witheach other as well as with the electrospun PLGA and chito-san membranes. Results of transmission electron micro-scope and X-ray photoelectron spectroscopy confirmed thecore/shell structure of the C-PLGA/chitosan fiber. Becauseof the introduction of chitosan, both H-PLGA/chitosan andC-PLGA/chitosan membranes showed significantly higherwater uptakes than that of PLGA but there was no signifi-cant difference between those of C-PLGA/chitosan and chi-

tosan membranes. In dry state, the C-PLGA/chitosan mem-branes exhibited extremely higher Young’s moduli (178.7 650.4 MPa) and strength (2.73 6 0.30 MPa) than those of H-PLGA/chitosan membranes (40.48 6 4.07, 1.44 6 0.12 MPa),respectively, but the values in wet state went down sharplybecause of the large amount (about 91%) of chitosan as theshell. Both H-PLGA/chitosan and C-PLGA/chitosan mem-branes showed better cytocompatibility than the PLGAmembrane in adhesion, viability assays as well as morphol-ogy observation. The obtained composite H- or C-PLGA/chitosan membranes would be potentially applied in wounddressings or skin tissue engineering. � 2009 Wiley Periodi-cals, Inc. J Biomed Mater Res 92A: 563–574, 2010

Key words: coelectrospinning; coaxial electrospinning;fibrous membrane; PLGA; chitosan

INTRODUCTION

Driven by a high-voltage electrostatic field, electro-spinning is a special fiber spinning technique to pro-duce ultrafine fibers as non-woven membranes frompolymeric solutions.1–3 Owing to their high specificsurface area and high porosity, the electrospunfibrous membranes could be potentially used aswound dressings to perform as a barrier to bacteria.4

Moreover, by mixing with the polymer solution forelectrospinning, some biomedical additives, such asdrugs and growth factors, could been introducedinto the ultrafine fibers.5,6 So far, both synthetic poly-mers including polylactide (PLA),7,8 poly(lactide-co-glycolide) (PLGA),9,10 poly(e-caprolactone) (PCL),11,12

and natural ones such as chitosan,13,14 collagen15,and gelatin16 have been electrospun into ultrafinefibrous membranes.

Electrospinning of multicomponent polymer solu-tions into ultrafine fibrous membranes has attractedmuch attention recently. To produce electrospuncomposite membranes, coelectrospinning and coaxialelectrospinning were mostly used. Two or moresyringes and power supplies were applied so as toelectrospin different polymer solutions simultane-ously to obtain excellent hybrid fibrous membranescomposed of two or more components.17,18 Coaxialelectrospinning is a technique using a spinneret com-posed of two coaxial capillaries, by which two differ-ent polymer solutions can be electrospun simultane-ously into core/shell structured ultrafine fibers.19 Byusing the compound spinneret, two componentscould be fed through the inner or outer coaxial capil-lary channels and integrated into a core/shell struc-tured composite fiber to fulfill a special applicationpurposes, for instance, to load drugs or growth fac-tors in the core component of the fibers.20,21

Correspondence to: X. Yuan; e-mail: [email protected] [email protected] grant sponsor: Natural Science Foundation of

China; contract grant numbers: 50573055, 50273027

� 2009 Wiley Periodicals, Inc.

PLGA and chitosan were chosen in this study asthe raw materials of electrospinning because of theirexcellent properties. PLGA is an easily processablepolymer, which microstructures could be achievedcontrollably and its mechanical properties and deg-radation period could be tailored by altering lactide/glycolide ratio and other morphological variables.22

Chitosan, a derivative of chitin, which is the mostabundant natural polymer after cellulose, hasbecome a widely used biomaterial, especially in thefield of skin regeneration because of its good bio-compatibility, biodegradability, hemostatic activity,anti-infectional activity, and properties to acceleratewound-healing.23

In this article, fibrous membranes of hybridPLGA/chitosan (H-PLGA/chitosan) and core/shellPLGA/chitosan (C-PLGA/chitosan) were producedby coelectrospinning and coaxial electrospinning,respectively. The morphology, mechanical proper-ties, water uptake in phosphate buffered solution(PBS) of the electrospun membranes were character-ized, and the cytocompatibility of fibroblasts wasinvestigated.

MATERIALS AND METHODS

Materials

PLGA (LA/GA 5 80:20, Mw 5 2.52 3 105) was kindlydonated by the Changchun Institute of Applied Chemistry,Chinese Academy of Sciences, China. Chitosan (Mw 5 6 3105, degree of deacetylation 5 85%), provided by QingdaoPharmaceutical Institute (Qingdao, China), was degradedto a lower molecular weight chitosan (Mw 5 2 3 105)using Co60-irradiation method according to references.24,25

All other reagents were analytical grade.

Electrospinning

As described in our previous report,17 to obtain electro-spun PLGA membranes, the polymer was dissolved at aconcentration of 0.06 g/mL in an organic solvent mixturecomposed of chloroform and N,N-dimethylformamide(DMF) (80:20, v/v) and electrospun at 15-kV positive volt-age, 12-cm working distance, and 0.2-mL/h solution flowrate. Chitosan solution (0.1 g/mL) was obtained by dis-solving chitosan in the admixture of trifluoroacetic acidand dichloromethane (80:20, v/v) and electrospun at 16-kVpositive voltage, 12-cm working distance, and 0.2-mL/hsolution flow rate to produce electrospun chitosanmembranes.

Coelectrospinning with dual sources and dual powerswas described in our previous article.18 PLGA and chito-san solutions were electrospun simultaneously on oppositesides of the rotating drum lying in the middle of the twosets of syringe pumps and power supplies to fabricate the

hybrid fibrous membranes of PLGA and chitosan (H-PLGA/chitosan).

For coaxial electrospinning, which setup was shown inour previous report,26 the PLGA solution was delivered tothe inner coaxial needle at the flow rate of 0.05 mL/h witha syringe pump, whereas the chitosan solution was pushedto the outer needle at the flow rate of 0.1 mL/h withanother syringe pump. An 18-kV positive voltage and a12-cm working distance were adopted. The core/shellstructured fibers of PLGA/chitosan (C-PLGA/chitosan)were collected on a grounded rotating drum wrappedwith aluminum foil. The rotating speed of the groundeddrum was 4 m/min.

The collected membranes containing chitosan includ-ing H-PLGA/chitosan and C-PLGA/chitosan werefurther crosslinked in glutaraldehyde vapor of a 25% glu-taraldehyde aqueous solution at 378C for 8 h. After cross-linking, the electrospun membranes were treated with a0.1 M glycine aqueous solution to block unreacted alde-hyde groups. All the electrospun membranes were vac-uumed and dried at room temperature for removing theremaining solvents and stored in a desiccator for furthercharacterizations.

Characterization

The morphology of the electrospun membranes of H-PLGA/chitosan and C-PLGA/chitosan as well as PLGAand chitosan were observed under a scanning electronmicroscope (SEM, Philips XL-30) after gold coating. Verifi-cation of the core/shell structure of C-PLGA/chitosanfibers was conducted by a JEOL JEM-100CX II transmis-sion electron microscope (TEM) operated at 100 kV. Thesamples for the TEM observation were prepared bydirectly depositing the as-electrospun fibers onto thecopper mesh coated with carbonic film.

Determination of the component proportion in both H-PLGA/chitosan and C-PLGA/chitosan membranes wascarried out by a weighing method after immersing themembranes before crosslinking in 2% acetic acid or chloro-form to eliminate one of PLGA and chitosan. Surface anal-ysis of the electrospun membranes was also carried out byX-ray photoelectron spectroscopy (XPS) in a Perkin ElmerPHI-1600 spectrometer using Mg Ka radiation (1253.6 eV,250 W).

Samples of electrospun PLGA, H-PLGA/chitosan, C-PLGA/chitosan, and chitosan membranes, cut into asquare shape of 20 3 20 mm2, were accurately weighed inan electronic balance and placed into glass bottles each ofwhich held 20 mL of PBS. The samples were incubated inPBS at (37.0 6 0.1)8C for 24 h and then weighed again im-mediately after removing PBS and absorbing surface waterwith a filter paper. The water uptake of electrospun mem-branes was calculated using the following equation:

Water uptake ¼ ðm1 �m0Þ=m03100%

where m0 and m1 were the masses of the membranesbefore and after immersion in PBS, respectively. All resultsrepresented the average results of three tests. Statistical

564 WU ET AL.

Journal of Biomedical Materials Research Part A

analysis was performed using Student’s t-test method anda value of p < 0.05 was considered to be statistically signif-icant. The static water-contact angle of the electrospunmembranes was measured at room temperature using acontact angle goniometer (JC2000C Contact Angle Meter,Powereach, Shanghai, China).

Mechanical properties of electrospun membranes of H-PLGA/chitosan, C-PLGA/chitosan as well as PLGA andchitosan were tested in a universal testing machine (Test-metric M350-20KN, UK) equipped with a 100 N load-cell.Samples were cut into strips in 60 3 10 mm2 and thegauge length was 40 mm. Each tensile test was operatedunder a crosshead speed of 5 mm/min at room tempera-ture. All the reported Young’s modulus, tensile strength,and elongation represented the average results of 5–7 tests.Statistical analysis was performed using Student’s t-testmethod. A value of p < 0.05 was considered to be statisti-cally significant.

Cytocompatibility evaluation

Fibroblast culture

Human embryo skin fibroblasts (hESFs), supplied bycell culture centre of institute of basic medical science,Chinese Academy of Medical Sciences, were cultured inDulbecco’s modified Eagle’s medium (DMEM, Gibcol,USA) containing 10% fetal bovine serum, 50 U/mL penicil-lin, and 50 U/mL streptomycin. The medium was replacedevery 3 days and cultures were maintained in a humidi-fied incubator at 378C with 5% CO2. After reaching about80% confluence, the cells were detached by 0.05% trypsin/0.02% ethylenediaminetetraacetic acid (EDTA) and pas-saged. The extraction was considered to be completed ifthe cells were prone to be rounded when observed undera phase contrast microscope. The number of cells wascounted with a hemocytometer.

Cell adhesion

The crosslinked chitosan, H-PLGA/chitosan, C-PLGA/chitosan, and PLGA membranes were cut into 0.9 3 0.9cm2 pieces and placed in 6-well cell culture plates forsterilizing and prewetting by ethanol for 1 day, thenwashed by Hank’s balanced salt solution 3 times andafter that were moved in 24-well cell culture plates. ThehESFs were seeded onto samples and tissue culture poly-styrene (TCPS) as well by dipping a cell suspension of 23 105 cells/100 lL. The cells were allowed to attach ontothe electrospun membranes and TCPS undisturbed in theincubator for 1, 3, 5, and 7 h. At each time point, the cel-lular constructs were rinsed with Hank’s solution andthe number of attached hESFs was determined using ahematocytometer after detaching the cells from the fab-rics using 0.1% trypsin. Data of the attachment measure-ment were collected from triplicate samples andexpressed as mean 6 standard deviation (SD). Statisticalanalysis was performed using one-way ANOVA methodwith a Scheffe test.

Cell viability

For cell viability study, hESFs were seeded on a 24-wellculture plates fixed with the four kinds of electrospunmembranes and TCPS by dipping a cell suspension of 2 3105 cells/100 lL and 4 h later another 1 mL of culture me-dium was filled into each culture well in the incubator.Medium was changed every 2 days. The proliferation ofhESFs on the substrates was quantified after 1, 3, and 5days by MTT assay. Briefly, the electrospun membranesand TCPS attached with cells were incubated for 4 h in100 lL MTT solution (5 mg/mL) at 378C and 5% CO2. Theintense red colored formazan derivatives formed were dis-solved with 500 lL dimethyl sulfoxide and stirred for 10min. Each time 100 lL suspension was taken out and theabsorbance was measured at 570 nm with a microplatereader. Data of the cell viability were also collected fromtriplicate samples and expressed as mean 6 SD. Statisticalanalysis was also performed using one-way ANOVA witha Scheffe test.

Cell morphology

For cell morphology observation, hESFs were seededonto the mentioned four kinds of electrospun membranesby dipping a cell suspension of 1 3 105 cells/100 lL. Andthen another 1 mL of culture medium was filled to the cul-ture wells after 4 h in the incubator. Medium was changedevery 2 days. After 3 and 5 days of culture, cellular con-structs were harvested, rinsed with Hanks’ solution toremove non-adherent cells. For SEM observation, the cellu-lar constructs were fixed with 2.5% glutaraldehyde in PBSat 48C for 4 h, subsequently dehydrated through a seriesof graded alcohol solutions and air-dried overnight. Afterdrying, the samples were sputtered with gold andobserved under SEM.

RESULTS AND DISCUSSION

Preparation of electrospun membranes

At the optimized electrospinning conditionsdescribed in the section of materials and methods,the ultrafine fibers of PLGA, H-PLGA/chitosan, C-PLGA/chitosan, and chitosan ultrafine fibers wereprepared and their SEM micrographs are shown inFigure 1. It could be seen in Figure 1(a) that severalPLGA fibers connected to each other, possiblybecause DMF in the mixed solvent volatilized not soquickly and the electrospun fibers were still wetwhen collected onto the grounded drum. The aver-age diameter of PLGA fibers was 232 6 76 nm witha distribution in the range of 150–300 nm. Althoughthere were some granules on the surface, the averagediameter of the crosslinked chitosan fibers was 2596 74 nm with a distribution of 200–350 nm [Fig.1(d)]. It could also be seen that the crosslinked chito-

ELECTROSPUN PLGA/CHITOSAN 565


san fibers were still in good conditions. The H-PLGA/chitosan fibers exhibited an average diameterof 272 6 62 nm in the range of 200–500 nm. A smallamount of surface granules in Figure 1(b) indicatedthat chitosan fibers and PLGA fibers lay separatelyon the collector. Compared with the H-PLGA/chito-san membrane, the C-PLGA/chitosan fibers exhib-ited an average diameter of 226 6 75 nm in 150–300nm, and some granules were also found on thefibers [Fig. 1(c)]. In general, the diameter differencesamong the four kinds of electrospun fibers were stat-istically insignificant. The core/shell structure of C-PLGA/chitosan composite fiber was clearly verifiedin its TEM micrograph as shown in Figure 2, inwhich the core held an average diameter of 83 nm.

Chitosan can dissolve in dilute acetic acid solutionbut PLGA cannot, whereas chloroform is a good sol-vent for PLGA but not for chitosan. Owing to thesolubility differences between PLGA and chitosan,the analysis of the individual PLGA or chitosan com-ponent proportion in both H-PLGA/chitosan and C-PLGA/chitosan membranes was carried out by incu-

bating the composite membranes in an aqueous 2%acetic acid solution or chloroform, and subsequentlyweighing the membrane remnants after removedthem from the solvent and dried. The obtained valueby using the acetic acid solution as the solvent wassupposed to be the PLGA amount in the composite

Figure 1. SEM micrographs (310k) of electrospun fibers of PLGA (a), H-PLGA/chitosan (b), C-PLGA/chitosan (c), andchitosan after crosslinking (d).

Figure 2. TEM micrograph of the C-PLGA/chitosan fiber.

566 WU ET AL.


membranes and the value of chitosan was deter-mined by the mass differences between the obtainedPLGA amount and the original membrane. The alter-native analysis by dissolving the composite mem-branes in chloroform was also performed. Detailedresults are shown in Table I. The PLGA amounts inboth H-PLGA/chitosan and C-PLGA/chitosan mem-branes measured after incubation in chloroformwere about 37% and 7%, respectively, lower thanthose (about 44% and 10%) measured by dissolving

the membranes in 2% acetic acid. The deviation wasprobably caused by the experimental method. Whentaken out from chloroform, the remnant, which wasassumed as chitosan in the membrane could prob-ably attach a little amount of PLGA, causing themeasured chitosan amount was higher than the truevalue. It might be the same reason why the PLGAamount was higher when measured by incubatingthe membrane in 2% acetic acid. It could be thoughtthat the H-PLGA/chitosan membrane containedaround 41% of PLGA and about 59% of chitosandetermined by the average results of the two meth-ods, whereas the C-PLGA/chitosan membrane held8.6% of PLGA and 91% of chitosan, as shown inTable I.

Figure 3 shows SEM micrographs (10 k3) of H-PLGA/chitosan and C-PLGA/chitosan membranesafter incubation of them in dilute acetic acid or chlo-roform, respectively. It could be seen that the H-PLGA/chitosan membrane kept its general fiberstructure after incubation in acetic acid [Fig. 3(a)] orchloroform [Fig. 3(b)] to eliminate chitosan or PLGA,

TABLE IComponent Percentage of Electrospun Composite

Membranes Determined by Weighing

Samples Solvent PLGA (wt %) Chitosan (wt %)

H-PLGA/CS Acetic acid 44.2 6 0.6 55.8 6 0.6Chloroform 37.3 6 3.1 62.7 6 3.1Average 40.8 59.2

C-PLGA/CS Acetic acid 10.2 6 0.8 89.8 6 0.8Chloroform 6.9 6 0.8 93.1 6 0.8Average 8.6 91.4

Figure 3. SEM micrographs (310k) of electrospun membranes of H-PLGA/chitosan (a, b) and C-PLGA/chitosan (c, d)after incubation in aqueous 2% acetic acid solution (a, c) or chloroform (b, d), respectively.



respectively, because there were no obvious interac-tions between PLGA and chitosan components inthe membrane. However, it could be found that theremained PLGA component of the C-PLGA/chitosanmembrane [Fig. 3(c)] after incubation in acetic acidcould not keep its previous integral fibrous struc-tures, and the survived chitosan component of theC-PLGA/chitosan membrane [Fig. 3(d)] after dis-solved in chloroform swelled significantly, althoughpresented as fibers. It was the core/shell fiber struc-ture in C-PLGA/chitosan membrane, in whichPLGA formed the core and chitosan formed theshell, and the higher amount of chitosan that gaverise to the above phenomena.

XPS is an analysis technique to measure the sur-face element components. In this study, XPS wasused to analyze the element components in the elec-trospun PLGA, H-PLGA/chitosan, C-PLGA/chito-san, and chitosan membranes. The atom percentagesof the samples are shown in Table II. Results verifiedthat the H-PLGA/chitosan membrane contained allthe elements of carbon, oxygen, and nitrogen, andthe nitrogen amount in the H-PLGA/chitosan mem-brane was 2.0% between those in electrospun PLGA(0%) and chitosan (6.1%) membranes, indicating thatthe H-PLGA/chitosan membrane was composed byboth PLGA and chitosan fibers. In respect that theshells of C-PLGA/chitosan fibers were much thickerthan 10 nm, the detectable resolution of XPS, onlythe surface components of C-PLGA/chitosan couldbe detected. Table II shows that the element compo-nents (C1s 5 61.8%, O1s 5 32.8%, N1s 5 5.4%) ofC-PLGA/chitosan membrane and those of chitosanmembrane (C1s 5 61.1%, O1s 5 32.8%, N1s 5 6.1%)were very similar, which suggested that the surfacecontent of C-PLGA/chitosan fibers was mainly chito-san and most of PLGA was encapsulated in the coreof the fibers. This result also implied the core/shellstructure of the electrospun C-PLGA/chitosan fibersas well.

Water uptake and static water-contact angle

The differences of the water uptakes among theelectrospun PLGA, H-PLGA/chitosan, C-PLGA/chi-

tosan, and chitosan membranes were investigated toevaluate the hydrophilicity of them. As shown in Ta-ble II, the water uptake of the electrospun PLGAmembrane exhibited the lowest value of 75.8% 66.3% because PLGA contained hydrophobic methyland methylene groups, and the electrospun PLGAmembrane mainly depended on its large porosityand specific surface area to absorb water. Onaccount of hydrophilic hydroxyl and amino groupsin chitosan, all the electrospun membranes contain-ing chitosan exhibited significantly higher wateruptakes than the electrospun PLGA membrane (p <0.01). The water uptake of the electrospun chitosanmembrane held the largest value and reached to476.2% 6 55.8%. The water uptake of the C-PLGA/chitosan membrane was 454.8% 6 26.3%, signifi-cantly higher than 356.8% 6 14.7%, the value of theH-PLGA/chitosan membrane, but showed no signifi-cant difference from that of chitosan. It was sug-gested that the electrospun C-PLGA/chitosan mem-brane exhibited better hydrophilicity than the elec-trospun H-PLGA/chitosan membrane because of theexistence of large amount of chitosan in C-PLGA/chitosan, especially for chitosan being the shell com-ponent of the electrospun core/shell fibers.

To understand the hydrophilicity changes of theelectrospun H-PLGA/chitosan and C-PLGA/chito-san membranes with respect to PLGA and chitosan,the static water-contact angle on each of the mem-brane was also tested. Results showed that the elec-trospun PLGA membrane held a higher water-contact angle value of 128.198 6 8.338, exhibitinga hydrophobic surface. Unfortunately, data of thewater-contact angle on electrospun H-PLGA/chito-san, C-PLGA/chitosan, and chitosan membranescould not be obtained because of the strong wettabil-ity of them. Water drops were absorbed into the C-PLGA/chitosan membrane quickly in about 1–2 secand let the membrane get wet. Similar phenomenahappened when measuring the chitosan and H-PLGA/chitosan membranes, but it took a little lon-ger time in about 3–5 sec for water drops gettinginto the H-PLGA/chitosan membrane than the othertwo samples.

From the results of water uptake and water-con-tact angle measurements, it could be suggested thatthe H-PLGA/chitosan membrane showed signifi-cantly higher hydrophilicity than PLGA, but lowerthan C-PLGA/chitosan, which was similar to thechitosan membrane. The hydrophilicity of the fourkinds of electrospun membranes was closely relatedto the chitosan amount and the fiber structure inthem. From the component results (Table I) of theelectrospun membranes determined by dissolvingand weighing, the C-PLGA/chitosan membrane helda higher chitosan amount of about 91%. So, it couldbe easily understood that it was the higher chitosan

TABLE IIAtom Percentage of Electrospun Membranes by

XPS and Water Uptake

Samples

Atom percentage by XPSWater

uptake (%)C1s (%) O1s (%) N1s (%)

PLGA 59.7 40.3 – 75.8 6 6.3H-PLGA/CS 69.1 28.9 2.0 356.8 6 14.7C-PLGA/CS 61.8 32.8 5.4 454.8 6 26.3Chitosan 61.1 32.8 6.1 476.2 6 55.8

568 WU ET AL.


amount that contributed to the higher water-uptakeability and wettability of this sample. The hybridfiber structure of about 41% PLGA and about 59%chitosan in H-PLGA/chitosan made the membranealso show higher hydrophilicity.

Mechanical properties

Figure 4 presents typical stress–strain curves ofthe electrospun PLGA, H-PLGA/chitosan, C-PLGA/chitosan, and chitosan membranes in dry and wetstates, respectively. The values of Young’s modulus,tensile strength, and elongation of the samples areillustrated in Figure 5. Results suggested that thePLGA membrane endured a typical ductile fracturein dry state, whereas the dry chitosan membraneshowed brittleness [Fig. 4(a)]. The brittleness of theelectrospun chitosan membrane was generally unde-sirable in most cases when used independently.Because of the high amount of chitosan (about 59%),the H-PLGA/chitosan membrane in dry and wet

states lost its ductile and exhibited significantlysmaller elongations in comparison with the electro-spun PLGA membrane. However, because of theintroduction of PLGA component, the Young’s mod-ulus, tensile strength, and elongation of the H-PLGA/chitosan membrane in dry and wet stateshad been significantly improved with respect to theelectrospun chitosan membrane. Owing to the spe-cial core/shell structure, the electrospun C-PLGA/chitosan membrane in dry state exhibited signifi-cantly higher Young’s modulus (178.7 6 50.4 MPa)

Figure 4. Typical stress–strain curves of electrospunPLGA, H-PLGA/chitosan, C-PLGA/chitosan, and chitosanmembranes in dry (a) and wet (b) states, respectively.

Figure 5. Dry and wet mechanical properties of electro-spun PLGA, H-PLGA/chitosan, C-PLGA/chitosan, andchitosan membranes. (a) Young’s modulus; (b) strength;(c) elongation.



and tensile strength (2.73 6 0.30 MPa) (p < 0.05)than those of the electropsun H-PLGA/chitosan(40.48 6 4.07, 1.44 6 0.12 MPa) and chitosan (38.676 7.61, 1.10 6 0.23 MPa) membranes, respectively,but presented similar elongation (p > 0.05).

In respect that PLGA was hydrophobic and waterwas not able to get into the center of the fibers, theelectrospun PLGA membrane could uptake a certainamount of water because of its high porosity. Afterincubation in PBS, the electrospun PLGA membraneexhibited a decrease tendency in its mechanicalproperties with respect to those in dry state, both ofthe tensile strength and elongation decreased slightlywithout significant difference, whereas Young’smodulus dropped significantly. On the contrary, chi-tosan was a good hydrophilic natural polymer andwhen incubated in PBS, the interactions between chi-tosan molecules were weakened by water, which sat-urated into the inside of chitosan fibers, causing thedecrease of tensile strength and Young’s modulus inwet state. With the introduction of PLGA compo-nent, the hydrophilicity of the H-PLGA/chitosanmembrane was increased but still lower than chito-san. Their Young’s modulus and tensile strengthwere 13.13 6 2.97 and 1.17 6 0.15 MPa in wet state,respectively, exhibiting significant decreasing extentsin comparison with the dry state, whereas the elon-gation increased significantly from 4.38% 6 0.89% to10.66% 6 2.49%. In the case of the C-PLGA/chitosanmembrane, the Young’s modulus and the tensilestrength in wet state decreased sharply from 178.7 650.4 and 2.73 6 0.30 MPa to 2.42 6 0.54 and 0.43 60.10 MPa, respectively, comparing with its dry state.This was due to its special structure with PLGA asthe core and chitosan as the shell. In dry state,PLGA worked similar like a strengthening additivein the membrane. On incubation in PBS, waterdestroyed the interactions between PLGA and chito-san molecules owing to the good hydrophilicity ofthe latter, and as a result, the tensile strength wasgreatly decreased. However, in wet state, PLGAcould work similar to a plasticizer and helped the C-PLGA/chitosan membrane keep a relatively highelongation. In this study, the shell amount of chito-san was much higher (about 91%) so that theC-PLGA/chitosan membrane showed lower Young’smodulus and tensile strength in wet state. It wassupposed that the weakness of the core/shell mem-branes in mechanical properties could be furtherstrengthened by increasing the core amount ofPLGA.

Cell adhesion

Adhesion of tissue cells was an important factor todetermine the biocompatibility of biomaterials and it

was thought that cells were more easily to attach onthe biocompatible surfaces. In the present work, theadhesion of hESFs on electrospun PLGA, H-PLGA/chitosan, C-PLGA/chitosan, chitosan membranes aswell as TCPS as controls was studied and shown inFigure 6 in a function of culture time.

Surface characteristics of the membrane, includingtheir topography, charge balance, hydrophilicity/hydrophobicity ratio, played essential roles in celladhesion on biomaterials.17,27,28 Cells in contact witha surface would firstly attach, adhere, and spread,and this first phase depended on adhesion of pro-teins. Generally, hydrophobic surfaces are able toadhere to more proteins in the first step to facilitatethe cell attachment.18,27,28 The electrospun PLGAmembrane contained a large number of hydrophobicgroups so that the electrospun PLGA membraneadhered to significantly more hESFs than the othersat 1 h. Higher hydrophilicity of the H-PLGA/chito-san, C-PLGA/chitosan, and chitosan membraneshampered the protein adhesion at the first phase sothat their adhesion rates were lower at 1 h.

As for highly hydrophobic materials, the confor-mation of proteins needed adjustment when proteinswere absorbed, the process of which might cause thedenaturation of proteins. Therefore, the adhesionnumbers of hESFs on PLGA became the least at thefollowing intervals of 3, 5, and 7 h. The introductionof chitosan component made the hydrophilicity ofthe H-PLGA/chitosan membrane increase to a sig-nificantly higher level, so the adhesion rate reachedthe highest at 3 h and 5 h, but it only showed a sig-nificant difference with PLGA membrane at 5 h. Astime passed to 7 h, the super high hydrophilicity ofthe C-PLGA/chitosan membrane became the mostimportant factor, which caused the adhesion ratehigher at 7 h. The surface hydrophilicity of

Figure 6. Adhesion of hESFs on the electrospun mem-branes and TCPS.

570 WU ET AL.


H-PLGA/chitosan, C-PLGA/chitosan, and chitosanmembranes were similar between each other, so thecell adhesion ability on them at 7 h was found.

Cell viability

Figure 7 shows the cell viability measured byMTT assay of hESFs cultured on different electro-spun membranes and TCPS controls within 7 days.The cell viability on the PLGA membrane was thesignificantly lowest all the time, and addition of thechitosan component could facilitate the hESFs viabil-ity. After 1 day, cells on TCPS, PLGA, H-PLGA/chi-tosan, C-PLGA/chitosan, and chitosan membranesall adhered to some extent and already began to pro-liferate, among them cells on H-PLGA/chitosan, C-PLGA/chitosan, and chitosan membranes showedsignificantly higher viability (p < 0.01) than those onthe PLGA membrane and even TCPS. After 3 and 5days of culture, the cells on all substrates increased,but the cell viability on the PLGA membrane exhib-ited the significant lowest growth rate (p < 0.01)than that on other samples. Cells on the PLGA mem-brane exhibited the lowest viability, similar to that ofcell adhesion at 5 h and 7 h. The number of cells onH-PLGA/chitosan grew almost fastest as it was

Figure 7. Viability of hESFs cultured on the electrospunmembranes and TCPS.

Figure 8. SEM micrographs (31k) of hESFs cultured for 3 days on the electrospun membranes of PLGA (a), H-PLGA/chitosan (b), C-PLGA/chitosan (c), and chitosan (d).



supposed that its surface hydrophilicity was suitablefor the cells to grow and proliferate.17,26 The cell via-bility on C-PLGA/chitosan and chitosan membraneswas also increased to higher levels but there wereno significant differences between the cell viabilityvalues on these three samples at 3 and 5 days of cul-ture.

It was assumed that both kinds of electrospunPLGA and chitosan fibers were located in the H-PLGA/chitosan membrane, suggesting the hydrophi-licity/hydrophobicity ratio and the charge balanceon the H-PLGA/chitosan membrane surface werebetween those of PLGA fibers and chitosanfibers.17,26 On the other hand, because of the shellchitosan component of the C-PLGA/chitosan fibers,the hydrophilicity/hydrophobicity ratio and thecharge balance on the surface were almost equal tothe electrospun chitosan fibers. So, the cells on H-PLGA/chitosan and C-PLGA/chitosan exhibited nosignificantly different viability values.

Cell morphology

Cell morphology and interaction between cellsand the electrospun membranes were studied in vitrofor 5 days. Figures 8 and 9 present SEM micrographsshowing the morphology of hESFs on the mem-branes cultured for 3 and 5 days, respectively. Onday 3, hESFs attached and spread on all the fourelectrospun membranes, but showed different mor-phology. The hESFs on the PLGA membrane wasspread to a certain extent, but they all kept their nor-mal shapes of spindle or unsymmetrical triangle onthe surface of the H-PLGA/chitosan membrane. Af-ter 5 days of culture, the number of hESFs on all thefour membranes tended to grow and most of thecells connected to each other as shown in Figure 9.The cells cultured on H-PLGA/chitosan, C-PLGA/chitosan, and chitosan membranes had connectedinto patches and the interactions between cells andelectrospun membranes became tighter. The electro-

Figure 9. SEM micrographs (3500) of hESFs cultured for 5 days on the electrospun membranes of PLGA (a), H-PLGA/chitosan (b), C-PLGA/chitosan (c), and chitosan (d).

572 WU ET AL.


spun fibers looked perforated from parts of the cellson the H-PLGA/chitosan membrane, and the cellspresented the current of ingrowth. It could be seenthat the hESFs on the electrospun chitosan-contain-ing membranes spread widely in comparison withthose on the electropsun PLGA membrane.

Taking into account of all the results of cell adhe-sion, cell viability and cell morphology, it could bediscovered that the PLGA membrane showed goodcell compatibility, and could accelerate cell adhesionand spread on the surface of scaffolds. The introduc-tion of chitosan greatly improved the cell compatibil-ity of the scaffolds and enhanced the interactionbetween cells and electrospun membranes. Becausethe surface component of the C-PLGA/chitosan elec-trospun membrane was similar to that of the electro-spun chitosan membrane, the growing state of hESFson its surface was similar, which meant the core/shell structure could keep the advantages of chitosanand suitable for cells to adhere and proliferate.

CONCLUSIONS

In this article, ultrafine fibrous membranes of chi-tosan, PLGA, H-PLGA/chitosan, and C-PLGA/chito-san were produced by electrospinning, coelectrospin-ning, and coaxial electrospinning, respectively.Results of component analysis indicated that theH-PLGA/chitosan and C-PLGA/chitosan mem-branes contained about 59% and 91% of chitosan,respectively. Measurements of water uptakes of bothH-PLGA/chitosan and C-PLGA/chitosan mem-branes suggested that their hydrophilicity was sig-nificantly higher than that of PLGA because of theintroduction of chitosan, and the C-PLGA/chitosanmembrane exhibited similar hydrophilicity with thechitosan membrane. The H-PLGA/chitosan and C-PLGA/chitosan membranes showed higher tensileproperties than those of chitosan, though these prop-erties of each membrane got worse in wet state thanin dry state. The evidence on cytocompatibility man-ifested that the H-PLGA/chitosan and C-PLGA/chitosan membranes could facilitate cell adhesionand viability, positively promote cell-scaffold andcell–cell interactions and encourage cells to migratethrough porous membranes to integrate with thesurrounding fibers. The obtained H-PLGA/chitosanand C-PLGA/chitosan membranes would haveadvantages for their potential applications in drugcontrol release and skin restoration.

References

1. Bashur CA, Dahlgren LA, Goldstein AS. Effect of fiber diam-eter and orientation on fibroblast morphology and prolifera-

tion on electrospun poly(D,L-lactic-co-glycolic acid) meshes.Biomaterials 2006;27:5681–5688.

2. Kim GH, Cho YS, Kim WD. Stability analysis for multi-jetselectrospinning process modified with a cylindrical electrode.Eur Polym J 2006;42:2031–2038.

3. Huang ZM, Zhang YZ, Kotaki M, Ramakrishna S. A reviewon polymer nanofibers by electrospinning and their applica-tions in nanocomposites. Comp Sci Tech 2003;63:2223–2253.

4. Ignatova M, Manolova N, Rashkov I. Electrospinning ofpoly(vinyl pyrrolidone)-iodine complex and poly(ethyleneoxide)/poly(vinyl pyrrolidone)-iodine complex – a prospec-tive route to antimicrobial wound dressing materials. EurPolym J 2007;43:1609–1623.

5. Huang ZM, He CL, Yang AZ, Zhang YZ, Han XJ, Yin JL, WuQS. Encapsulating drugs in biodegradable ultrafine fibersthrough co-axial electrospinning. J Biomed Mater Res Part A2006;77A:169–179.

6. Chew SY, Mi RF, Hoke A, Leong KW. Aligned protein-poly-mer composite fibers enhance nerve regeneration: A potentialtissue-engineering platform. Adv Funct Mater 2007;17:1288–1296.

7. Tsuji H, Nakano M, Hashimoto M, Takashima K, Katsura S,Mizuno A. Electrospinning of poly(lactic acid) stereocomplexnanofibers. Biomacromolecules 2006;7:3316–3320.

8. Zhou HJ, Green TB, Joo YL. The thermal effects on electro-spinning of polylactic acid melts. Polymer 2006;47:7497–7505.

9. Xin XJ, Hussain M, Mao JJ. Continuing differentiation ofhuman mesenchymal stem cells and induced chondrogenicand osteogenic lineages in electrospun PLGA nanofiber scaf-fold. Biomaterials 2007;28:316–325.

10. Zong XH, Ran SF, Fang DF, Hsiao BS, Chu B. Control ofstructure, morphology and property in electrospun poly(gly-colide-co-lactide) non-woven membranes via post-draw treat-ments. Polymer 2003;44:4959–4967.

11. Thomas V, Jose MV, Chowdhury S, Sullivan JF, Dean DR,Vohra YK. Mechano-morphological studies of aligned nanofi-brous scaffolds of polycaprolactone fabricated by electrospin-ning. J Biomater Sci Polym Ed 2006;17:969–984.

12. Wutticharoenmongkol P, Sanchavanakit N, Pavasant P, Supa-phol P. Novel bone scaffolds of electrospun polycaprolactonefibers filled with nanoparticles. J Nanosci Nanotechnol2006;6:514–522.

13. OhkawaK, ChaDI, KimH,NishidaA, YamamotoH. Electrospin-ning of chitosan.Macromol RapidCommun2004;25:1600–1605.

14. Duan B, Dong CH, Yuan XY, Yao KD. Electrospinning of chi-tosan solutions in acetic acid with poly(ethylene oxide). J Bio-mater Sci Polym Ed 2004;15:797–811.

15. Zhong SP, Teo WE, Zhu X, Beuerman RW, Ramakrishna S,Yung LYL. An aligned nanofibrous collagen scaffold by elec-trospinning and its effects on in vitro fibroblast culture.J Biomed Mater Res Part A 2006;79A:456–463.

16. Choktaweesap N, Arayanarakul K, Aht-Ong D, MeechaisueC, Supaphol P. Electrospun gelatin fibers: Effect of solventsystem on morphology and fiber diameters. Polym J 2007;39:622–631.

17. Duan B, Wu LL, Li XR, Yuan XY, Li XL, Zhang Y, Yao KD.Degradation of electrospun PLGA-chitosan/PVA membranesand their cytocompatibility in vitro. J Biomater Sci Polym Ed2007;18:95–115.

18. Duan B, Wu LL, Yuan XY, Li XR, Hu Z, Li XL, Zhang Y, YaoKD, Wang M. Hybrid nanofibrous membranes of PLGA/chi-tosan fabricated via an electrospinning array. J Biomed MaterRes Part A 2007;83A:868–878.

19. Sun B, Duan B, Yuan XY. Preparation of core/shell PVP/PLA ultrafine fibers by coaxial electrospinning. J Appl PolymSci 2006;102:39–45.

20. Liao IC, Chew SY, Leong KW. Aligned core-shell nanofibersdelivering bioactive proteins. Nanomedicine 2006;1:465–471.



21. Jiang HL, Hu YQ, Zhao PC, Li Y, Zhu KJ. Modulation of pro-tein release from biodegradable core-shell structured fibersprepared by coaxial electrospinning. J Biomed Mater Res BAppl Biomater 2006;79B:50–57.

22. Kim K, Yu M, Zong XH, Chiu J, Fang DF, Seo YS, Hsiao BS,Chu B, Hadjiargyrou M. Control of degradation rate andhydrophilicity in electrospun non-woven poly(D,L-lactide)nanofiber scaffolds for biomedical applications. Biomaterials2003;24:4977–4985.

23. Ishihara M, Nakanishi K, Ono K, Sato M, Kikuchi M, SaitoY, Yura H, Matsui T, Hattori H, Uenoyama M, Kurita A.Photocrosslinkable chitosan as a dressing for wound occlu-sion and accelerator in healing process. Biomaterials 2002;23:833–840.

24. Liu XF, Chen Q, Pan HM. Rheological behavior of chitosanderivative/cellulose polyblends from N-methylmorpholineN-oxide/H2O solution. J Mater Sci 2007;42:6510–6514.

25. Choi WS, Ahn KJ, Lee DW, Byun MW, Park HJ. Preparationof chitosan oligomers by irradiation. Polym Degrad Stab2002;78:533–538.

26. Li S, Sun B, Li XR, Yuan XY. Characterization of electrospuncore/shell poly(vinyl pyrrolidone)/poly(L-lactide-co-e-capro-lactone) fibrous membranes and their cytocompatibility invitro. J Biomater Sci Polym Ed 2008;19:245–258.

27. Anselme K. Osteoblast adhesion on biomaterials. Biomaterials2000;21:667–681.

28. Vogler EA. Structure and reactivity of water at biomaterialsurfaces. Adv Colloid Interface Sci 1998;74:69–117.

574 WU ET AL.


composite fibrous membranes of plga and chitosan prepared by coelectrospinning and coaxial...

Documents