Core/Shell Nanofibers with Embedded Liposomes as a Drug DeliverySystemAndrea Mickova,†,‡ Matej Buzgo,*,†,‡ Oldrich Benada,§ Michala Rampichova,†,‡ Zdenek Fisar,∥

Eva Filova,‡,⊥ Martina Tesarova,# David Lukas,○ and Evzen Amler†,‡,⊥

†Department of Biophysics, Second Faculty of Medicine, Charles University in Prague, V Uvalu 84, 150 06 Prague 5, Czech Republic‡Institute of Experimental Medicine, Academy of Sciences of the Czech Republic, v.v.i, Vídenska 1083,142 20 Prague 4, CzechRepublic§Laboratory of Molecular Structure Characterization, Institute of Microbiology, Academy of Sciences of the Czech Republic, v.v.i.,Vídenska 1083, 142 20 Prague 4, Czech Republic∥Department of Psychiatry, First Faculty of Medicine, Charles University in Prague, Czech Republic⊥Faculty of Biomedical Engineering, Czech Technical University, Sítna 3105, 272 01 Kladno, Czech Republic#Laboratory of Electron Microscopy, Institute of Parasitology, Biology Centre of the Academy of Sciences of the Czech Republic,Branisovska 31, 37005 Ceske Budejovice, Czech Republic○Department of Nonwovens, Technical University of Liberec, Studentska 2, 461 17 Liberec, Czech Republic

*S Supporting Information

ABSTRACT: The broader application of liposomes in regenerative medicineis hampered by their short half-life and inefficient retention at the site ofapplication. These disadvantages could be significantly reduced by theircombination with nanofibers. We produced 2 different nanofiber-liposomesystems in the present study, that is, liposomes blended within nanofibers andcore/shell nanofibers with embedded liposomes. Herein, we demonstrate thatblend electrospinning does not conserve intact liposomes. In contrast, coaxialelectrospinning enables the incorporation of liposomes into nanofibers. Wereport polyvinyl alcohol-core/poly-ε-caprolactone-shell nanofibers withembedded liposomes and show that they preserve the enzymatic activity ofencapsulated horseradish peroxidase. The potential of this system was alsodemonstrated by the enhancement of mesenchymal stem cell proliferation. In conclusion, intact liposomes incorporated intonanofibers by coaxial electrospinning are very promising as a drug delivery system.

■ INTRODUCTIONBiocompatible and biodegradable scaffolds prepared by electro-spinning are gaining popularity in the field of tissueengineering. Electrospinning is a fiber-forming nanotechnologyprocess that enables the creation of submicrometer fibers drawnfrom polymer solutions and melts.1,2 The principle ofelectrospinning technology has been described in detailpreviously.3,4 By modulating the properties of the polymersolution (e.g., concentration, viscosity, and molecular weight ofthe polymer), electrospinning setup (e.g., spinning electrode,collector, flow rate, collector distance, and applied voltage), andenvironmental properties (e.g., humidity and temperature),fibers with diameters in the order of a few micrometers down totens of nanometers can be produced.5,6 Thus, the electro-spinning process affords the opportunity to prepare tissueengineering scaffolds with micro- to nanoscale topography,tunable porosity, and large surface area-to-volume ratios.7,8

Nanofibers formed by electrospinning have been shown tomimic the structure and biological function of the extracellularmatrix (ECM) in terms of chemical and physical structure. The

diameter of electrospun fibers closely matches the size scale ofthe ECM, which is ideal for cell attachment, proliferation, anddifferentiation.9 With respect to the stereological properties ofnanofibers, fiber diameter, porosity, and pore size are the mostimportant parameters. Typically, an electrospun matrix hasporosity as high as 90% and a pore diameter down to 100 nm.Cell infiltration strongly depends on the pore size and porosityof the polymer mesh. The optimal pore size for cell attachment,proliferation, and migration varies from 5 to 500 μm.8 The poresize of electrospun scaffolds can be adjusted by modifying thefiber diameter, fiber alignment, or by introducing sacrificialcomponents.10,11

In addition to serving as scaffolds, nanofibrous meshes can bedesigned to serve as delivery vehicles for bioactive factors.12

The main aspects for the development of successful deliverysystems for tissue engineering are the bioactivity of the

Received: August 17, 2011Revised: March 7, 2012Published: March 8, 2012

Article

pubs.acs.org/Biomac

© 2012 American Chemical Society 952 dx.doi.org/10.1021/bm2018118 | Biomacromolecules 2012, 13, 952−962

biomolecules incorporated within the scaffolds and thecontrolled release of these biomolecules according to thetime frame of tissue regeneration.13

Functionalized nanofibers have been produced directly byelectrospinning polymer blends/mixtures with various chemicalagents such as antibiotics, cytostatics, proteins, DNA, and smallinterfering RNA.14,15 Recently, the successful encapsulation ofbacterial cells and viruses into blend nanofibers was reported.16

However, a common problem in this process is the loss ofactivity of the incorporated biomolecules (e.g., enzymes andgrowth factors) due to conformational changes in the organicsolution environment. This problem was overcome by theincorporation of proteins with precipitously degradablepolymers dissolved in polar solvents. Their ability to maintainenzyme activity was demonstrated by the incorporation ofnatural proteins such as lipase and casein,17 horseradishperoxidase (HRP),18 alkaline phosphatase, and β-galactosi-dase.19 The disadvantages of using such scaffolds are the lowaqueous stability of the nanofibers, burst release, and lowadhesion of cells to the hydrophilic polymer surface.13 Thus,there are several key challenges to be overcome before thepractical implementation of blend nanofibers as a sustaineddelivery system of bioactive factors in tissue engineering.12

Apart from the usage of blends, “core-shell electrospinning”,also known as “coaxial electrospinning”, is a sophisticatedmethod of producing second-generation nanofibers with astrictly organized core−shell structure.20 The spinningelectrode of the coaxial electrospinning apparatus consists oftwo capillaries that are placed together coaxially. The shellsolution is injected through the outer needle, which mostcommonly consists of a spinnable polymeric material, whereasthe core solution, which is injected through the inner needle,can be composed of other polymers or encapsulated masses,including simple liquids.21,22 The coaxial morphology of thenanofibers was demonstrated by the preparation of nanofiberswith a hollow interior.23 The effects of various processingparameters on coaxial electrospinning were evaluated by Sarafet al.24

Coaxial electrospinning is an attractive strategy for thedelivery of susceptible biomolecules because the producedcore−shell fibers have great potential to preserve proteinsduring the electrospinning process. The hydrophilic corepolymer facilitates the loading and preservation of proteinbioactivity, whereas the hydrophobic shell allows fiberformation.24 Nanofiber cores have been employed for theencapsulation of diverse bioactive substances such as anti-biotics, drugs, DNA, and proteins.14,15 Coaxial electrospinningwas successfully utilized for the encapsulation of proteins suchas lysozyme,25 platelet-derived growth factor-bb (PDGF-bb),26

nerve growth factor (NGF),27 and basic fibroblast growth factor(bFGF).28 In addition, coaxial electrospinning provideshomogeneous protein distribution throughout the fibers, andproteins can be delivered in a controlled manner.13

Recent progress in the preparation of protein-compatiblenanofibers has improved the preservation of protein activity;13

however, proteins still remain in polymeric solutions and arenot in their natural aqueous microenvironment. This short-coming can be solved by the use of liposomes. Conventionalliposomes are artificially prepared vesicles composed of a lipidbilayer that encapsulates an aqueous solution inside ahydrophobic membrane. Liposomes have received widespreadattention as carriers of therapeutically active compounds due totheir unique characteristics, for example, capability to

incorporate hydrophilic and hydrophobic drugs, good bio-compatibility, low toxicity, and targeted delivery of bioactivecompounds to the site of action.29 However, their short half-life, low stability, and poor control of drug release overprolonged periods limit their use for long-term drug delivery.30

The combination of liposomes with polymeric scaffolds couldovercome the limitations of conventional liposomes and extendtheir use in biomedical engineering.The aim of the present study was to develop a drug delivery

system based on liposome-enriched nanofibers. Blend andcoaxial electrospinning were employed for the incorporation ofliposomes into nanofibers. The morphology of nanofibers wasexamined using cryo-field emission scanning electron micros-copy (FESEM), while the morphology of embedded liposomeswas examined using confocal microscopy. This studyconcentrated primarily on the delivery efficiency and thepreservation of the enzymatic activity of the delivered proteins.Horseradish peroxidase (HRP) was selected as a model protein.The activity of HRP was compared in samples prepared byblend or coaxial electrospinning with or without liposomes. Inorder to examine if liposomes were able to maintain theirinternal aqueous environment, the water content in drynanofibers was measured. Finally, the potential of the preparedsystem was evaluated in a cultivation study using mesenchymalstem cells (MSCs). Given the unique properties of liposomesand nanofibers, this complex system could serve as aconvenient delivery vehicle for a variety of biologically activecompounds.

■ MATERIALS AND METHODSLiposome Preparation. Unilamellar liposomes were prepared

from soybean-derived L-α-phosphatidylcholine (PC; Avanti PolarLipids, Inc.) using the extrusion method.31 Briefly, 25 mg of soybeanPC were dissolved in chloroform (1 mL) and subsequently evaporatedunder a flow of N2 at 4 °C to form a thin lipid film. The dried lipidfilms were then resuspended in 1 mL of Tris-buffered saline (TBS; 150mM NaCl, and 50 mM Tris, pH 7.4) for the preparation of emptyliposomes. Liposomes with encapsulated substances were prepared byresuspending the dry lipid films in 1 mL of TBS with 2 mg/mL HRP,25 mg/mL fluorescein, or 5 mg/mL FITC-dextran (10000 MW;purchased from Sigma-Aldrich) or with a mixture of recombinantgrowth factors (0.1 μg/mL TGF-β; 5 μg/mL bFGF, 10 μg/mL IGF-Iin phosphate-buffered saline (PBS)). To create unilamellar liposomes,the obtained multilamellar liposomes were extruded several timesthrough polycarbonate filters with a well-defined pore-size (1 μmdiameter) using an Avanti Mini-extruder (Avanti Polar Lipids Inc.).Unencapsulated fluorescein was separated on a Sephadex G-25 column(Sigma-Aldrich). Unencapsulated FITC-dextran, HRP, and themixture of growth factors were separated from the encapsulatedliposomes using a Sephacryl HR-500 column (GE Healthcare).

Measurement of Liposome Size. The average diameter and sizedistribution of the liposomes were measured using dynamic lightscattering on a Zetasizer Nano ZS (model ZEN3601; MalvernInstruments Ltd.). The experiments were carried out at 18 °C with adetection angle of 90°. The data were prepared using the Continmode, and the results were given in average mean diameters reportedas a function of the size distribution and intensity.

Blend Electrospinning of Nanofibers. A mixture of 12% (w/v)polyvinyl alcohol (PVA, Sloviol R) with empty liposomes or TBSbuffer was prepared for the quantification of lipid incorporation andwater content. A mixture of 12% (w/v) PVA with 271 μg/mL HRP or253 μg/mL HRP encapsulated into liposomes was prepared forenzymatic activity studies. A mixture of 12% (w/v) PVA with 5 mg/mL FITC-dextran (Sigma-Aldrich) encapsulated into liposomes wasprepared for visualization by confocal microscopy. Electrospinning wascarried out on a Nanospider device, as described previously in detail.32

Biomacromolecules Article

dx.doi.org/10.1021/bm2018118 | Biomacromolecules 2012, 13, 952−962953

A high-voltage source generated voltages of up to 55 kV, and thepolymer solutions were connected with the high-voltage electrode.The electrospun nanofibers were deposited on a grounded collectorelectrode. The distance between the tip of the syringe needle and thecollecting plate was 12 cm. All electrospinning processes wereperformed at room temperature (RT; ∼24 °C) and a humidity of∼50%.Coaxial Electrospinning of PCL/PVA Nanofibers. A 10% (w/v)

poly-ε-caprolactone (PCL) solution was used as the shell solution forall prepared samples. For confocal microscopy, the core solutionconsisted of 5% (w/v) PVA with fluorescein (25 mg/mL) orliposomes containing encapsulated fluorescein (25 mg/mL; Fluka,Sigma-Aldrich) or FITC-dextran (5 mg/mL) or liposomes (25 mg/mL) containing encapsulated FITC-dextran (5 mg/mL; Sigma-Aldrich). For the cell proliferation studies, the core solution containedeither a mixture of growth factors (0.1 μg/mL TGF-β; 5 μg/mLbFGF, 10 μg/mL IGF-I in PBS; GF) or liposomes containing anencapsulated mixture of growth factors (0.03 μg/mL TGF-β, 0,7 μg/mL bFGF, 3 μg/mL IGF-I, determined by ELISA; LIP-GF) dissolvedin 5% (w/v) PVA.The coaxial spinneret apparatus consisted of 2 needles placed

together coaxially.3 Two syringe pumps were utilized to deliver thecore and shell solutions, respectively. The flow rate was 5 mL/h for thecore polymer and 2 mL/h for the shell polymer. A high-voltage powersupply was used to generate voltages of up to 55 kV, and an aluminumplate collector was used as the receiving plate to collect theelectrospun nanofibers. The distance between the tip of the syringeneedle and the collecting plate was 12 cm. All electrospinningprocesses were performed at RT (∼24 °C) and a humidity of ∼50%.Coaxial Electrospinning of PVA/PVA Nanofibers. Readily

dissolvable coaxial PVA/PVA nanofibers were prepared for studies onwater content and enzymatic activity. A 12% (w/v) PVA (Sloviol R)solution was prepared as the shell solution for all prepared samples.For water content determination, the core solution contained emptyliposomes or TBS buffer dissolved in 5% (w/v) PVA. For thedetermination of enzymatic activity the core solution consisted of HRP(2 mg/mL) dissolved in 5% (w/v) PVA or liposomes withencapsulated HRP (888 μg/mL) dissolved in 5% (w/v) PVA. Thecoaxial spinneret apparatus consisted of 2 needles placed togethercoaxially.3 Two syringe pumps were used to deliver the core and shellsolutions, respectively. A high-voltage power supply was used togenerate voltages of up to 55 kV, and an aluminum plate collector wasused as the receiving plate to collect the electrospun nanofibers. Thedistance between the tip of the syringe needle and the collecting platewas 12 cm. All electrospinning processes were performed at RT (∼24°C) and a humidity of ∼50%.Cryo-Field Emission Scanning Electron Microscopy. Cryo-

field emission scanning electron microscopy (FESEM) was used tovisualize nanofiber morphology. Briefly, the sample was frozen rapidlyin liquid N2 (−210 °C). The sample was then transferred to the cryo-stage of the preparation chamber (ALTO2500), where it was freeze-fractured at −140 °C, freeze-etched by raising the sample temperatureuntil the sublimation of water started at −95 °C for 10 min, and thencoated with gold at −135 °C for 60 s. After this preparation, thesample was placed on a cold-stage microscope and examined in thefrozen state at a temperature of approximately −135 °C. The specimenwas observed at 1 kV in the GB-high mode and at 3 kV in the GB-lowmode on a Jeol 7401-FE microscope.Electrospun scaffolds were characterized in terms of fiber diameter

pore size and porosity using mathematical stereological methods, asdescribed in detail by Lukas et al.33 Briefly, stereological parameterswere measured from arbitrarily selected sections of the FESEM imagesin LUCIA software (Laboratory Imaging s.r.o.). The distribution offiber diameters was determined quantitatively from 200 measurements.The most probable void space radius, rp, that is, the most probableradius of a spherical pore, was estimated using the method ofMasounave et al.:34

=π

rN

12p

S

where NS is the number density of fiber hits in a random cross sectionof the specimen.

The probability density of a void space radius, r, was estimated usingthe formula

ρ = π − πr N re( ) 2 N rS S

2

Additionally, the size of the liposomes embedded in coaxial nanofiberswas also evaluated from SEM micrographs using Ellipse software(Version 2, 0, 7, 1; ViDiTo).

Scanning Electron Microscopy. Air-dried samples of blendelectrospun PVA nanofibers were mounted on aluminum stubs andsputter-coated with a layer of gold (∼60 nm thick) using a Polaronsputter-coater (SC510; Now Quorum Technologies Ltd.). Thesamples were examined in an Aquasem (Tescan) scanning electronmicroscope in the secondary electron mode at 15 kV.

Confocal Microscopy of Nanofibers. The distribution ofencapsulated FITC-dextran and fluorescein within the preparednanofibers was observed using a Zeiss LSM 5 DUO (FITCfluorescence, λex = 488 nm, λem = 520 nm) confocal laser scanningmicroscope.

Stability of Liposomes in PVA Solution. To quantify thedegradation rate, unilamellar liposomes with the encapsulatedfluorescent probe 8-aminonaphthalene-1,3,6-trisulfonic acid, disodiumsalt (ANTS) and quencher p-xylene-bis-pyridinium bromide (DPX;Invitrogen) dissolved in TBS were prepared by the extrusion method,as described in Liposome Preparation. Unilamellar liposomes withencapsulated ANTS/DPX were separated from unencapsulated ANTSand DPX on a Sephadex G-25 column (Sigma-Aldrich).

The liposomal suspension (1.25 mL) was mixed with 16% (w/v)PVA solution and TBS to obtain 5 mL of liposome-PVA suspensionwith a PVA concentration ranging from 1−12% (w/v). Liposomesdiluted with TBS buffer without the addition of PVA were used as acontrol sample. The solution was incubated for 30 min and thedegradation ratio was calculated using fluorescence spectroscopy on amicroplate reader (Synergy HT; λex 360−500 nm, λem 520−540 nm).First, 100 μL of the samples were diluted with 100 μL of TBS to lowerthe viscosity and fluorescence intensity of the liposomes in buffer(background intensity I0), and the liposomes diluted in PVA (I) weremeasured. The intact liposomes in the diluted samples were lysed bythe addition of 20 μL of 2% Triton X-100, and the I0TX for the controlsample and ITX for the liposomes diluted in PVA were determined.The degradation ratio was calculated as

=− ×

−I I C

I IDR(%)

( )0

TX 0

where C represents the background correction factor calculated as

=CII0TX

TX

All values represent the mean ± standard deviation (SD) of at leastfour independent experiments.

Phosphorus Determination. To quantify the incorporation ofphospholipids into PVA nanofibers prepared by blend electrospinning,we measured the amount of phosphorus in the nanofiber mesh. Thenanofiber samples containing empty liposomes were first weighed andcut into four small pieces to determine the total amount ofincorporated phospholipids. The samples were placed into calibratedtubes containing 0.3 mL of 70% perchloric acid. For combustion, thetubes were placed in a sand bath heated to 180 °C for 60 min (thestandards did not require combustion). The resulting metaphosphateswere converted to orthophosphate by adding 2 mL of redistilled waterto the cooled samples and mixing thoroughly. Subsequently, 0.25 mLof 2.5% ammonium molybdate solution followed by 0.25 mL of 10%ascorbic acid solution (freshly prepared) were added, mixedthoroughly, and heated for 5 min in a boiling water bath. Aftercooling the samples in a cold water bath for 5 min, the optical densityat 820 nm was measured using a Unicam SP1700 UV/visspectrophotometer. In each series of measurements, parallel

Biomacromolecules Article

dx.doi.org/10.1021/bm2018118 | Biomacromolecules 2012, 13, 952−962954

determinations were made of the blank values of the reagent solutionand a standard preparation. The weight of the incorporatedphospholipids (mincorporated = mg phospholipids/mg sample) wascalculated from the amount of measured phosphorus (1 mol ofphospholipids = 1 mol of phosphorus):

= ×m

n Mmincorporated

sample

where n represents the amount of detected phosphorus, M representsthe molecular weight of phosphatidylcholine (760.09 g/mol), andmsample represents the weight of sample used for the analysis.The incorporation of phospholipids (IR) was calculated as a ratio

between the measured weight of phospholipids in the samples(mincorporated = mg phospholipids/mg sample) and the weight of thetotal added phospholipids in the samples (mtotal = mg phospholipids/mg total sample):

= ×m

mIR(%) 100

incorporated

total

Measurement of the FITC-Dextran Release Profile. To studythe release profile of FITC-dextran, nanofiber meshes with or withoutliposomes were cut into round patches (6 mm diameter) andincubated with 1 mL of TBS at RT. At specific intervals, the TBS waswithdrawn and replaced with fresh buffer. The time interval wasdetermined according to the balance between the release of adetectable amount of FITC-dextran and the maintenance of the sinkcondition. Drug release was quantified using fluorescence spectrosco-py. Briefly, 200 μL of samples and blank samples were measured on amultiplate fluorescence reader (Synergy HT; λex 480−500 nm, λem520−540 nm) and a background subtraction was performed. All valuesrepresent the mean ± SD of at least three independent experiments.The cumulative release profile of FITC-dextran was obtained, and therelease half-time was determined as the time at which the initialfluorescence intensity (I0) decreased to I = I0·e

−1.Determination of HRP Activity. Nanofiber samples prepared

either by blend electrospinning (PVA/HRP, PVA/liposomes withencapsulated HRP) or by coaxial electrospinning (PVA/PVA withHRP, PVA/PVA with HRP encapsulated in liposomes) were dissolvedin dH2O, vortexed, and incubated overnight at 4 °C. The total proteinconcentration was determined by the Bradford assay (BradfordQuickStart; BioRad) using an enzyme-linked immunosorbent assay(ELISA) reader at 590 nm (Biotec, Synergy HT). Enzyme activity wasmeasured colorimetrically using tetramethylbenzidine (TMB) as thesubstrate. The enzyme substrate reaction was terminated by theaddition of 50 μL of 2 M H2SO4 after 30 s. Enzyme activity wasmeasured at 450 nm using an ELISA reader (Biotec, Synergy HT). Allvalues represent the mean ± SD of at least three independentexperiments.Water Content Determination. To determine residual water

content inside embedded liposomes, coulometric Karl Fisher titrationwas performed. Dry PVA nanofiber meshes with empty liposomeswere prepared by coaxial or blend electrospinning, whereas the controlsamples were prepared without liposomes, simply with TBS buffer. Allsamples were cut into small pieces of equal size and weighed (∼20mg/sample). Subsequently, the samples were wrapped in filter paperand measured. Blind samples prepared without nanofibers consistedsimply of filter paper. Water content was determined by coulometricKarl Fisher titration on an 831 KF Coulometer (Methrom AG) with adiaphragm-free cell and HYDRANAL-Coulomat AG (Sigma-Aldrich)as the anolyte. All samples were measured in triplicate and thebackground was subtracted. The water content corresponding toliposomes in nanofibers was calculated as the difference between thevalues measured in the samples with and without liposomes. Overallwater retention (%) was calculated as a ratio of the measured watercontent (mg) to the theoretical overall water content in the liposomes(mg).Enzyme-Linked Immunosorbent Assay. To determine the

cumulative release profile of growth factors from nanofiber mesheswith (LIP-GF) and without liposomes (GF), the meshes were cut into

round patches (6 mm in diameter) and a time dependent release studywas performed. Samples were incubated with 1 mL of sterile PBS at 37°C. At specific intervals, the PBS was withdrawn and replaced withfresh buffer. The concentrations of TGF-β, IGF-I, and bFGF in themedium were determined by sandwich enzyme-linked immunosorbentassay (ELISA; DuoSet, RnD systems) following the manufacturer’sinstructions. Briefly, the samples were prepared by acidification of thecollected medium with 1 M HCl for 10 min and neutralized by theaddition of 1.2 M NaOH with 0.5 M HEPES. Captured antibodieswere precoated onto 96-well plates and blocked by 5% (w/v) BSAdissolved in PBS-T overnight and washed twice with PBS-T. Then,100 μL of samples or standards were added to each of the wells intriplicate, and the plates were incubated for 2 h at 4 °C and washedtwice with PBS-T, which was followed by the addition of 100 μL ofbiotinylated primary antibodies, and the plates were incubated for 2 hat 4 °C. The plates were washed twice with PBS-T, and each well wasincubated with 100 μL of streptavidin-conjugated HRP (20 min, 4 °C)and again washed twice with PBS-T. The antigen−antibody complexwas detected by colorimetric reactions initiated by the addition of 100μL of TMB substrate solution (RnD Systems) to each well, and thereaction was stopped after 20 min with 50 μL of 2 N H2SO4. Theabsorbance of the samples was measured at 470 nm with an ELISAreader (SynergyHT, Biotek) and the concentration was determined bya six-point calibration curve. The cumulative release profile isrepresented as a mass of released growth factor per scaffold in time.All values represent the mean ± SD of at least three independentexperiments.

MSC Isolation, Culture, and Seeding. Blood marrow wasaspirated from the iliac wing of a minipig into a 5 mL syringecontaining 1 mL of PBS and 25 IU heparin under general anesthesia.The mononuclear cells were isolated using gradient separation withthe plasma substitute Gelofusine. Briefly, blood marrow was mixedwith 1.25 mL of Gelofusine. After 30 min, the upper and mediumlayers containing plasma, mononuclear cells, and erythrocytes wereaspirated and centrifuged at 270 × g for 15 min. Subsequently, themedium layer with mononuclear cells was aspirated and seeded intissue culture flasks. Adherent cells were cultured in MinimumEssential Medium (MEM; with L-glutamine, PAA) containing 10%fetal bovine serum (Mycoplex, PAA), 100 IU/mL penicillin, and 100μg/mL streptomycin. The cells were passaged using the trypsin−EDTA method before confluence was reached. The cells from thethird passage were used for the cell culture study.

PCL/PVA scaffolds were prepared for the cell culture studies asdescribed in the Coaxial electrospinning of PCL/PVA nanofiberssection. Three scaffolds of each sample (i.e., nanofibers with a mixtureof growth factors (GF) or liposomes containing an encapsulatedmixture of growth factors (LIP-GF) were cut into circles with adiameter of 6 mm, sterilized using ethylene oxide at 37 °C, and putinto a 96-well plate. The scaffolds were seeded with MSCs (37 × 103

cells/cm2) and cultured in 200 μL of MEM (Sigma-Aldrich, E15−825), 1% fetal bovine serum, and penicillin/streptomycin for 14 days.The medium was not changed during cultivation. After 1, 3, 7, or 14days, cell viability, proliferation, and DNA content were determined.

DNA Measurement. DNA content was determined using thePicoGreen Assay Kit (Invitrogen Ltd.) on days 1, 3, 7, and 14. Toprocess the samples for the analysis of DNA content, 250 μL of celllysis solution (0.2% v/v Triton X-100, 10 mM Tris (pH 7.0), 1 mMEDTA) was added to each well containing a scaffold sample. Toprepare the cell lysate, the samples were processed through a total ofthree freeze/thaw cycles, that is, the scaffold sample was first frozen at−70 °C and then thawed at RT. Between each freeze/thaw cycle, thescaffolds were roughly vortexed. The prepared samples were stored at−70 °C until analysis. To quantify the number of cells on the scaffolds,a cell-based standard curve was prepared using samples with knowncell numbers (range 100−106 cells). DNA content was determined bymixing 100 μL of PicoGreen reagent and 100 μL of DNA sample.Samples were loaded in triplicate and florescence intensity wasmeasured on a multiplate fluorescence reader (Synergy HT, λex =480−500 nm, λem = 520−540 nm).

Biomacromolecules Article

dx.doi.org/10.1021/bm2018118 | Biomacromolecules 2012, 13, 952−962955

Viability of Seeded MSCs. Cell viability was determined using theCellTiter 96 Aqueous One Solution Cell Proliferation Assay (MTSassay, Promega). At 1, 3, 7, and 14 days, the scaffolds were transferredto a new 96-well plate containing 100 μL of fresh medium per well and20 μL of CellTiter 96 Aqueous One Solution Reagent. The formazanabsorbance in 100 μL of the solution was measured (λsample = 490 nm,λreference = 690 nm) after a 2 h incubation at 37 °C and 5% CO2 usingan ELISA reader (EL800; BioTek). The absorbance of the sampleswithout cells was deducted from the cell-seeded samples.Cell Proliferation Evaluation. The proliferation of MSCs seeded

on the nanofibrous scaffolds was determined using a colorimetricimmunoassay based on the measurement of 5-bromo-2″-deoxyuridine(BrdU), which is incorporated during DNA synthesis (cellproliferation ELISA, BrdU, colorimetric; Roche Diagnostics GmbH,Germany). The assay was performed according to the manufacturersinstructions. Briefly, on days 3 and 7, 120 μL BrdU-labeling solutionwas added to each well containing a scaffold and was allowed toincorporate into the cells in a CO2-incubator at 37 °C for 2 h.Subsequently, the supernatant in each well was removed, and thescaffolds were incubated with FixDenat solution to fix the cells anddenature the DNA at RT for 30 min. The supernatant was removedand, subsequently, 100 μL of anti-BrdU-peroxidase (dilution ratio =1:100) was added and kept at RT for 60 min. After removing theunbound antibody conjugate, 100 μL of substrate solution was added,allowed to stand for 6 min, and the reaction was completed by adding25 μL of H2SO4 solution (1 M). Then, 100 μL of solution wastransferred to a 96-well plate and measured within 5 min at 450 nmwith a reference wavelength of 690 nm using an ELISA plate reader(EL 800; BioTek). The blank corresponded to scaffolds without cellswith or without BrdU.Fluorescence Confocal Microscopy of Cell-Seeded Nano-

fibers. After 1, 7, and 14 days of cultivation, the scaffolds seeded withcells were fixed with frozen methanol (−20 °C), rinsed twice withPBS, incubated in 10 μg/mL 3,3′-dihexyloxacarbocyanine iodide(DiOC6; Invitrogen) for 45 min at RT, and then incubated in 5μg/mL propidium iodide in PBS for 10 min. The scaffold was rinsedtwice with PBS and scanned the same day. DiOC6 staining was usedto visualize mitochondria and inner membranes, while propidiumiodide staining was used to visualize cell nuclei. A Zeiss LSM 5 DUOconfocal microscope at λex = 488 and 560 nm and λem = 505−550 and575−650 nm was used for DiOC6 and propidium iodide detection,respectively.Statistical Analysis. The results were evaluated statistically using

one-way analysis of the variance (ANOVA) and the Student−Newman−Keuls method. The level of significance was set at 0.05.The data are presented as the mean ± SD.

■ RESULTS AND DISCUSSION

Blend Electrospinning Does Not Conserve IntactPhosphatidylcholine Liposomes. First, blend electrospin-ning was employed for the encapsulation of liposomes into

nanofibers. Unilamellar liposomes were dispersed in a 12%aqueous solution of PVA, and nanofibers with liposomes wereelectrospun using needleless electrospinning. To demonstratethat the blended nanofibers contained lipids, the total amountof incorporated phospholipids was determined (SupportingInformation, Table 1). The incorporation ratio of phospholi-pids into nanofibers prepared by blend electrospinning was57.1 ± 5.8% (w/w). Visualization of nanofibers with liposomescontaining FITC-dextran was achieved by means of SEM(Figure 1a) and confocal microscopy (Figure 1c). Theseanalyses did not show any intact liposomes; moreover, FITC-dextran was distributed in nonfibrous areas. The results indicatethat the spinning technology used here caused the liposomes tobreak and release their encapsulated material.

Intact Phosphatidylcholine Liposomes Can Be In-corporated into Nanofibers by Coaxial Electrospinning.To overcome this problem with blend electrospinning ofconventional liposomes, we used the coaxial electrospinningmethod to prepare PVA-core/PCL-shell nanofibers withembedded liposomes. Core−shell nanofibers with embeddedliposomes have not yet been described. The core mediumcontained liposomes dispersed in a 5% PVA solution. Twoneedles that were placed together coaxially to deliver the coreand shell solutions, respectively, created the main part of thecoaxial spinneret apparatus. To visualize the outcome, weapplied FESEM.Stereological measurements identified two main populations

of nanofibers in the fibrous PCL/PVA mesh without liposomes.The first population had an average fiber diameter that peakedat 50 nm and the second at 150 nm. The sample also containeda fraction of microfibers. The most probable void space radius,rp, that is, the most probable radius of a spherical pore, was 0.33± 0.07 μm. Stereological analysis of the PCL/PVA fiber meshwith embedded liposomes showed a similar radius distributionof nanofibers. Thin nanofibers with an average radius of 50 nmdominated, followed by nanofibers with an average radius of150 and 350 nm. The mean porosity was 90% and the mostprobable void space radius, rp, was 0.17 ± 0.04 μm. Bothsamples also contained a small fraction of microfibers.Compared to control samples prepared without liposomes

(Figure 2, inset), the morphology of the fibers indicated thatthe liposomes were encapsulated in the PVA/PCL nanofibers(Figure 2a) because of the presence of round bulges. Thesebulges, with a size of 233.4 ± 36.9 nm, indicated the location ofthe embedded individual liposomes. The size of the liposomesprepared for the electrospinning process was characterized by

Figure 1. SEM and confocal microscopy analysis of blend electrospun PVA nanofibers with liposomes: (A) SEM analysis of nanofibers withincorporated liposomes shows a large number of nonfibrous areas. The blend spinning technology causes the liposomes to break and consequentlyrelease their encapsulated material. The large nonfibrous areas are probably the result of the dissolution of PVA nanofibers caused by liposomebreakage. (B) PVA nanofibers without liposomes as a control. (C) Visualization of liposomes containing FITC-dextran by confocal microscopy.Blend electrospinning of PVA nanofibers with liposomes does not keep the liposomes intact; consequently, FITC-dextran leaks from the liposomesand shows signals that are distributed throughout the sample. Scale bars indicate (A) 50 μm and (B, C) 20 μm.

Biomacromolecules Article

dx.doi.org/10.1021/bm2018118 | Biomacromolecules 2012, 13, 952−962956

dynamic light scattering measurements. It was confirmed thattheir size ranged from 100 nm to 1 μm (SupportingInformation, Figure 2). Detailed observations of SEM micro-graphs using Ellipse software revealed nanofibers withembedded liposomes with a size of 835.5 ± 2.8 nm (Figure 2b).To investigate whether the embedded liposomes remained

intact in the mesh, another batch of PVA-core/PCL-shellnanofibers was prepared. In this case, 10% PCL nanofibers wereused as the shell solution, whereas the core solution consistedof either liposomes with encapsulated fluorescein dissolved in5% PVA or liposomes with encapsulated FITC-dextrandissolved in the same solution. The dry PVA/PCL nanofibermesh with embedded liposomes was visualized by means ofconfocal microscopy (Figure 3), revealing distinct areas ofliposomes containing either fluorescein (Figure 3a) or FITC-dextran (Figure 3c). In contrast, nanofibers prepared by coaxialelectrospinning without liposomes, but with the addition offluorescein (Figure 3b) or FITC-dextran (Figure 3d) in 5%PVA, showed a uniform distribution of the fluorescentsubstances throughout the nanofiber mesh. Therefore, thelocalization of the fluorescent substances in the liposomeswithin a dry nanofiber mesh suggests that the liposomesremained intact during and after coaxial electrospinning.This effect can be explained by the higher osmotic pressure

of the blend PVA samples and lower shear stress during thecoaxial electrospinning process compared to noncoaxialelectrospinning.35 Osmotic pressure reportedly affects theviability of electrospun bacterial cells, indicating that thisparameter may play a role for membrane systems, includingliposomes.36 To examine the effect of the PVA polymersolution on liposomes, the degradation ratio of liposomes in a1−12% (w/v) PVA solution was determined. We observed ahigher degree of liposome degradation as the polymerconcentration increased (Supporting Information, Figure 1).The degradation ratio of liposomes (dissolved in 12% PVA)used for blend electrospinning was 53.7 ± 8.7%, indicating thatmore than half of the liposomes were disintegrated before theelectrospinning process started. For the liposomes dissolved in5% PVA, the degradation ratio was lower (29.3 ± 2.5%), that is,∼70% of the liposomes remained intact.The second factor affecting liposomal stability during the

electrospinning process is shear stress.35 We hypothesized thatduring the process of blend electrospinning, the hydrodynamicforces disrupt the liposomal membrane and the large amount ofwater that evaporates during the fiber-forming process keepsthe liposomes from reforming. On the other hand, during

coaxial electrospinning, the hydrodynamic forces in the core ofthe jet are lowered,35 thereby preventing the disruption of theliposomal membrane.Recently, the successful incorporation of cerasomes into

blend nanofibers was reported by Zha et al.37 Cerasomes areorganic−inorganic lipid carriers that were shown to surviveencapsulation during blend electrospinning due to their higherstability compared to conventional liposomes. Modification ofliposomal composition, which promotes a more stablestructure, enables their incorporation into nanofibers. Onother hand, coaxial electrospinning enables the incorporation ofunilamellar phosphatidylcholine liposomes and opens the wayfor their application in long-term drug delivery. The use of suchnanofibers as drug delivery systems can combine the uniqueadvantages of nanofibers and liposomes.

FITC-Dextran Release is Prolonged from CoaxialNanofibers with Intact Liposomes. The developed systemwas tested for drug release. To study the release profile fromcoaxially electrospun nanofiber meshes with or withoutliposomes, FITC-dextran incorporated into the nanofiber corewas employed as the monitoring fluorescence probe. TheFITC-dextran samples were incubated at RT in TBS buffer,which was subsequently replaced with fresh buffer. The releaseprofile of FITC-dextran was then obtained. The collectedfractions were analyzed by fluorescence spectroscopy and thecumulative release profile was calculated (Figure 4).Core/shell nanofibers containing FITC-dextran (without

liposomes) showed an intense burst release (60.6% of FITC-dextran released in 24 h), and the fluorescence in the collectedTBS buffer was undetectable after 240 h. The release half-timewas calculated as tr = 20 h. The release half-time from coaxial

Figure 2. FESEM of coaxially electrospun PVA-core/PCL-shellnanofibers with encapsulated liposomes: (A) FESEM of liposomesembedded within PVA-core/PCL-shell nanofibers. (Inset) Pure PVA-core/PCL-shell nanofibers mesh without liposomes as a control. (B)Details of an incorporated intact liposome in a coaxially preparednanofiber with a smaller nanofiber attached to the surface of the PCLnanofiber. Scale bars indicate (A, B) 300 nm and (inset) 2 μm.

Figure 3. Confocal microscopy of PVA-core/PCL-shell nanofibersprepared by coaxial electrospinning: Dry coaxially electrospun PVA-core/PCL-shell nanofiber mesh with encapsulated liposomes contain-ing either (A) fluorescein or (C) FITC-dextran with signals distributedinside the intact liposomes visualized by confocal microscopy. Coaxialnanofibers prepared without liposomes, but with the addition of either(B) fluorescein or (D) FITC-dextran show signals that are distributedthroughout the entire fibers. Scale bars indicate (A, B) 20 μm and (C,D) 50 μm.

Biomacromolecules Article

dx.doi.org/10.1021/bm2018118 | Biomacromolecules 2012, 13, 952−962957

nanofibers is strongly dependent on the presence of a water-soluble core polymer,38 in our case, PVA. Rapid FITC-dextranrelease from the nanofiber core could be explained either bywater sorption into the nanofiber core39 and by the formationof submicrometer pores in the nanofiber shell.40 A comparisonof dry (Figure 5b) and wet (Figure 5a) coaxial nanofiber

meshes indicated the formation of submicrometer pores in thePCL shell in wet nanofiber samples (Figure 5a). Therefore, wehypothesized that pore formation affected drug release from thenanofiber core. This observation supports the hypothesisintroduced by Tiwari et al.40

Interestingly, core/shell nanofibers with liposomes encapsu-lating FITC-dextran showed a slower initial release. Only ∼20%of the FITC-dextran was detected after 24 h of incubation,which shows the protective function of the liposomes. Therelease half-time was shifted to tr = 112 h for samples withFITC-dextran entrapped in liposomes. This phenomenon canbe explained by the temporary stability of the liposomes in thePVA core. The release half-time from nanofibers with

encapsulated liposomes depends on many physical andchemical factors, especially on the chemical composition andsize of the liposomes.41,42 Nevertheless, the release from thesystem depends on the stability of the core and shell polymers13

and the morphology of the fibers.24 However, the observedresults demonstrated that nanofibers with liposomes in theircore are appropriate means of controlled drug delivery;therefore, the design of coaxial nanofibers with the desiredrelease kinetics for practical tissue engineering applicationsshould be further examined.

Intact Encapsulated Liposomes Preserve EnzymaticActivity. To verify the protective effect of liposomes in thenanofiber core on enzymatic activity, the enzyme activity ofHRP was tested using four types of nanofiber samples. The firsttwo were prepared by blend electrospinning (B-HRP wasprepared from PVA solution containing HRP and B-LIP wasprepared from PVA solution containing HRP encapsulated inthe liposomes), while the third and fourth samples wereprepared by coaxial electrospinning, where the shell solutionwas made from PVA and the core solution consisted of eitherPVA with HRP (K-HRP) or PVA containing HRP encapsu-lated in the liposomes (K-LIP; Figure 6). The sampleconcentration was correlated with the homogeneously dis-tributed incorporation value to calculate the incorporation ratio.Proteins were encapsulated effectively and the total amount ofHRP in the nanofibers was comparable. The incorporation ratiowas ∼50% in both blend and coaxial systems (Figure 6a). HRPactivity was detected in all samples; however, the nanofibersprepared by coaxial electrospinning preserved HRP activity at asignificantly higher level than those prepared by blendelectrospinning (Figure 6b). The preserved HRP-specificactivity of 62.3 ± 7.9% in liposomes encapsulated in thenanofiber core was an order of magnitude higher than that insamples prepared by blend electrospinning.Electrospun nanofiber meshes prepared with empty lip-

osomes or without liposomes were tested for water content in adry state. The highly sensitive and water-specific titrationtechnique for water determination, that is, coulometric KarlFisher titration, was employed. Nanofibers prepared by blendelectrospinning with liposomes contained 36.2 ± 0.9% water.Interestingly, samples prepared by coaxial electrospinning withliposomes showed 65.2 ± 0.8% water retention (Table 1).These results indicate the preservation of intact liposomes filledwith water, even in a dry state.This observation opens up new perspectives on the

preservation of native enzymes. The liposomes entrapped inthe nanofiber core maintained a water environment afterseveral weeks of shelf storage; moreover, the system apparentlydid not leak water and could probably be used for even longerperiods of storage (i.e., in terms of months). Thus, thenanofiber core appears to form a water-impermeable barrier inthe dry state because of the interaction between the corehydrophilic polymer and the liposomes inside the core solution.

Liposomes with Encapsulated Growth Factors Stim-ulate MSC Proliferation. Electrospun nanofibers wereutilized for the delivery of various growth factors for theinduction of MSC viability and proliferation. Sahoo et al.28

prepared nanofibrous scaffolds for the controlled delivery ofbFGF and showed increased MSC proliferation on coaxialnanofibers compared to blend electrospun nanofibers.To demonstrate the potential of the coaxial system with

embedded liposomes for drug delivery of growth factors, weexamined the viability and proliferation of MSCs on prepared

Figure 4. Cumulative release profile of FITC-dextran from nanofibers:Nanofibers prepared by coaxial electrospinning of PCL/PVA with(LIP) and without (DEX) encapsulated liposomes are shown. Therelease profile of the DEX sample shows an intense burst release(60.6%) that is subsequently followed by a slow release. On the otherhand, the release profile of the LIP sample shows a lower initial release(20%) that is followed by a sustained release. The release half-time (tr)is 20 h for the DEX sample and 112 h for the LIP sample. Theseresults show the protective activity of liposomes and the sustainedrelease of encapsulated FITC-dextran.

Figure 5. Comparison of dry and wet nanofibers indicates theformation of submicrometer pores in the nanofiber shell: Nanofibersprepared by the coaxial electrospinning of PCL as a shell polymer withPVA and liposomes as the core polymer. (A) FESEM visualization ofwet nanofibers shows the formation of submicrometer pores on thenanofiber surface (see arrows). (B) Dry nanofibers visualized byFESEM microscopy exhibit smooth surfaces without pore structures.These results indicate that the release of core content occurs throughnewly formed pores. Scale bars indicate (A, B) 1 μm.

Biomacromolecules Article

dx.doi.org/10.1021/bm2018118 | Biomacromolecules 2012, 13, 952−962958

nanofibers. TGF-β, bFGF, and IGF-I have been shown toincrease MSC proliferation and viability,44,45 thus, MSCs wereseeded on PCL and PVA coaxial nanofibers with the addition ofa mixture of these growth factors (bFGF, TGF-β, and IGF-I)either dispersed in the core PVA solution (GF) or encapsulatedin liposomes (LIP-GF). Concentrations of released growth

factors were detected by ELISA. Results showed controlledrelease of growth factor during the incubation period.Concentration of all three growth factors from LIP-GF scaffoldwas lower than from GF scaffold (Figure 7). From scaffolds

with embedded liposomes, 7.07 ± 0.65 ng of TGF-β, 420.2 ±57.5 ng of IGF-I, and 131.37 ± 4.21 ng of bFGF was releasedduring the 21 day study. In contrast, from coaxial nanofiberswithout liposomes, 11.69 ± 0.44 ng of TGF-β, 880.96 ± 10.27ng of IGF-I, and 239.23 ± 7.83 ng of bFGF was released. Toincrease the effect of the delivered growth factors, the cells werecultivated in a cell culture medium containing only 1% fetalbovine serum, and the medium was not refreshed during thecultivation period. The number of cells on the scaffolds wasquantified using a dsDNA-specific assay on days 1, 3, 7, and 14(Figure 8a). Statistically significant differences between

Figure 6. Incorporation ratio and enzymatic activity of HRP: (A)Comparison of HRP incorporation into nanofibers with (K-LIP and B-LIP) and without (K-HRP and B-HRP) liposomes prepared by coaxialelectrospinning (K-LIP and K-HRP) or blend electrospinning (B-LIPand B-HRP). The incorporation of HRP into the nanofibers wasdetermined by the Bradford assay. The highest HRP incorporationratio was measured in nanofibers prepared by blend electrospinningwithout liposomes (B-HRP). (B) The enzyme activity in nanofibersprepared by coaxial electrospinning (K-LIP and K-HRP) or blendelectrospinning (B-LIP and B-HRP) was determined by theconversion of the TMB substrate and was correlated with the proteinconcentration. The highest activity was in coaxial nanofibers withliposomes (K-LIP).

Table 1. Water Content of the Nanofibersa

sample

sampleweight(mg)

maximal watercontent (μg water/

mg sample)

measured watercontent (μgwater/mgsample)

incorporationratio (%)

K-LIP 129 2.759 1.8 65.23 ± 0.95B-LIP 121 4.413 1.6 36.26 ± 0.78aNanofibers prepared by blend electrospinning with liposomes (B-LIP) contain 36.2 ± 0.95% water. Nanofibers prepared by coaxialelectrospinning with liposomes (K-LIP) contain 65.2 ± 0.78% water.

Figure 7. Cumulative release profile of growth factors: Release ofgrowth factors was determined using ELISA. (A) Release profile ofTGF-β1 released from coaxial nanofibers with (LIP-GF) and without(GF) liposomes. (B) Release profile of IGF-I released from coaxialnanofibers with (LIP-GF) and without (GF) liposomes. (C) Releaseprofile of bFGF released from coaxial nanofibers with (LIP-GF) andwithout (GF) liposomes.

Biomacromolecules Article

dx.doi.org/10.1021/bm2018118 | Biomacromolecules 2012, 13, 952−962959

scaffolds were detected on all assayed days. Interestingly, whilethe cell number decreased on coaxial PCL/PVA scaffoldswithout liposomes during the study period, the concentrationof DNA significantly increased in samples with embeddedliposomes utilized as a growth factor delivery vehicle duringcultivation. To determine cell viability, the MTS assay wasperformed on days 1, 3, 7, and 14 (Figure 8b). On day 1, therewas no difference in cell viability among the samples. However,the MTS assay showed a significantly higher viability of MSCscultured on LIP-GF scaffolds compared to GF scaffolds on days3 and 14, indicating that cells incubated on a sample withliposomes had a higher viability. The MTS assay showed asignificant increase of MSC viability on LIP-GF scaffoldscompared to GF scaffolds. The increased DNA concentrationof MSCs was indicative of higher rates of cell proliferation. Cellproliferation on the scaffolds was examined using the BrdUassay and confocal microscopy. The BrdU assay showedsignificantly higher proliferation rates on days 3 and 7 ofcultivation on PCL/PVA scaffolds with embedded liposomes(Figure 8c). Additionally, confocal microscopy showed that

there was no difference between the GF and LIP-GF scaffoldson day 1, and the cells were spread, isolated, or in small groupsand randomly distributed on the scaffold (Figure 9a,c); low cell

numbers were observed for the cells cultured on GF scaffolds,even on day 14. The cells were organized in small isolatedgroups, indicating poor cell growth on the scaffold (Figure 9b).In contrast, the LIP-GF scaffolds showed increased cell density(Figure 9d). The well-spread cells formed large groups,indicating increased cell proliferation. Confocal microscopysupported the results of the quantitative assays and confirmedthat core/shell nanofibers with embedded liposomes enhancedMSC proliferation and viability more than coaxial nanofiberswithout liposomes.In addition, data from confocal microscopy showed poor cell

infiltration into the scaffold. This result could be explained bythe small pore size of the scaffold. Coaxial nanofibers withembedded liposomes had a mean pore size of 0.17 ± 0.04 μm,while coaxial nanofibers without liposomes had a mean poresize of 0.33 ± 0.07 μm. Such pore sizes are insufficient for cellinfiltration. If the scaffolds made via electrospinning have a poresize <10 μm, the cells cannot easily infiltrate the nanofibers andmake a three-dimensional shape, as exists in the ECM.43 Thus,the nanofiber mesh will essentially behave as a two-dimensionalsheet, where the cells can only proliferate on its surface. Toovercome this limitation, several approaches have beenreported. For example, systems combining insoluble fibers

Figure 8. Viability and proliferation of MSCs on PVA/PCL coaxialnanofibers: (A) Determination of dsDNA content using thePicoGreen method showed an increase in cell number on LIP-GFscaffolds compared to GF scaffolds. Statistically significant differenceswere detected between scaffolds at all assayed days. (B) The MTSassay revealed significantly higher viability of MSCs on PVA/PCLcoaxial scaffolds with liposomes (LIP-GF) on days 3 (indicated by *)and 14 (indicated by **) compared to coaxial scaffolds withoutliposomes (GF). (C) The BrdU assay showed a significantly highercell proliferation rate on LIP-GF scaffolds compared to GF scaffolds atdays 3 and 7. The level of statistical significance for the assays isdesignated above the mean values (p < 0.05).

Figure 9. Confocal microscopy of MSCs on PVA/PCL coaxialscaffolds: Confocal microscopy on day 1 after seeding did not showany difference between PVA/PCL coaxial scaffolds containing amixture of growth factors (GF) (A) and PVA/PCL coaxial scaffoldswith liposomes containing growth factors (LIP-GF) (C). Themicrographs show the increased proliferation of MSCs on LIP-GFcoaxial scaffolds at 14 d after seeding (D) compared to GF coaxialscaffolds containing a growth factor solution in the core (B). Thedetection of mitochondria and inner membranes of MSCs usingDiOC6 staining (green) and of cell nuclei using propidium iodidestaining (red). A maximum intensity projection of a 30−40 μm thickstack of images was applied using a 20× objective (NA = 0.80). Scalebar = 50 μm.

Biomacromolecules Article

dx.doi.org/10.1021/bm2018118 | Biomacromolecules 2012, 13, 952−962960

and sacrificial cofibers were reported46,47 or combiningelectrospun scaffolds with electrosprayed hydrogels werereported.48 Additionally pore size and cell infiltration ofnanofibrous scaffolds can be regulated with use of macro-molecules, for example, hyaluronan.49 Alternatively, pores canbe introduced by salt particles that are subsequently leachedout50 or by introducing ice crystals.51 After the removal of thesacrificial part of the system, void spaces are introduced and thepore size of the scaffold increases. Physical manipulation, thatis, ultrasonication in an aqueous solution can also enhancecellular infiltration and enlarge the porosity of electrospunnanofibers.52 Beside these methods, pore size could be adjustedby regulating electrospinning processes4 or by using specialcollectors.53,54 Such approaches could be combined with theproposed system; however, further study is necessary tooptimize their use for specific tissue engineering applications.The results of the cell culture study showed the protection of

susceptible compounds by embedded liposomes in coaxialnanofibers. This drug delivery system seems to be a promisinggrowth factor delivery strategy for use in regenerative medicine.However, further optimization of the proposed system forparticular tissue engineering applications is necessary.

■ CONCLUSIONSWe produced liposomes dispersed within blended nanofibersand core/shell nanofibers with embedded liposomes. Thehighest potential was shown by intact liposomes incorporatedinto nanofibers by coaxial electrospinning in contrast withblend electrospinning. Interestingly, coaxial electrospinningenabled the retention of an aqueous environment inside intactliposomes embedded in nanofibers. Enzymes encapsulated inliposomes can better survive the electrospinning process,probably because of the shielding effect of the lipid sphere.Finally, nanofibrous scaffolds containing embedded liposomeswith encapsulated recombinant growth factors were morepotent at stimulating MSC proliferation than coaxial nanofiberswithout liposomes. In conclusion, the results presented in thecurrent study show the potential of this drug delivery systemusing intact liposomes embedded in coaxial nanofibers invarious fields of tissue engineering and regenerative medicine.

■ ASSOCIATED CONTENT*S Supporting InformationCalculation of the phospholipids incorporation into PVAnanofibers prepared by blend electrospinning (Table 1.).Additionally, the effect of 1−12% (w/v) PVA solution onliposomal degradation is presented in Figure 1. Finally, the sizedistribution of prepared unilamellar liposomes detected bydynamic light scattering is presented in Figure 2. This materialis available free of charge via the Internet at http://pubs.acs.org.

■ AUTHOR INFORMATIONCorresponding Author*Tel.: (+420) 296 442 387. E-mail: buzgo@labdemo.cz.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis research was supported by the Academy of Sciences of theCzech Republic (Institutional Research Plans AV0Z50390703and AV0Z50390512), the Ministry of Education, Youth, andSports of the Czech Republic (Research Programs NPV II

2B06130 and 1M0510, Grant No. MSM0021620849, andProject MSM Grant No. 4977751303), the Grant Agency of theAcademy of Sciences (Grant No. IAA500390702), grants fromthe Czech Science Foundation (Grants Nos. GA202/09/1151and P304/10/1307), the Grant Agency of Charles University(Grant Nos. 96610, 80009, 97110, 330611, 384311, and164010), the Grant Agency of the Czech Republic (Grant No.106/09/P226), Internal Grant Agency of the Ministry ofHealth of the Czech Republic (Grant No. NT12156), and theMinistry of Education of the Czech Republic (Project ERA-NET CARSILA, No. ME 10145). Cactus Communications Inc.provided language help and proofreading of the manuscript(www.editage.com).

■ REFERENCES(1) Cipitria, A.; Skelton, A.; Dargaville, T. R.; Dalton, P. D.;Hutmacher, D. W. J. Mater. Chem. 2011, 21, 9419−9453.(2) Garg, K.; Bowlin, G. L. Biomicrofluidics 2011, 5, 13403.(3) Lukas, D.; Sarkar, A.; Martinova, L.; Vodsedalkova, K.; Lubasova,D.; Chaloupek, J.; Pokorny, P.; Chvojka, J.; Komarek, M. Textile Prog.2009, 41, 59−140.(4) Pham, Q. P.; Sharma, U.; Mikos, A. G. Tissue Eng. 2006, 12,1197−1211.(5) Datta, N.; P. Pham, Q.; Sharma, U.; Sikavitsas, V. I.; Jansen, J. A.;Mikos, A. G. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 2488−2493.(6) Teo, W. E.; Inai, R.; Ramakrishna, S. Sci. Technol. Adv. Mater.2011, 12, 013002.(7) Ma, Z.; Kotaki, M.; Inai, R.; Ramakrishna, S. Tissue Eng. 2005, 11,101−109.(8) Pham, Q. P.; Sharma, U.; Mikos, A. G. Biomacromolecules 2006, 7,2796−2805.(9) Keun Kwon, I.; Kidoaki, S.; Matsuda, T. Biomaterials 2005, 26,3929−3939.(10) Vaquette, C.; Cooper-White, J. J. Acta Biomater. 2011, 7, 2544−2557.(11) Rnjak-Kovacina, J.; Weiss, A. S. Tissue Eng., Part B 2011, 17,365−372.(12) Dahlin, R. L.; Kasper, F. K.; Mikos, A. G. Tissue Eng., Part B2011, 17, 349−364.(13) Ji, W.; Sun, Y.; Yang, F.; van den Beucken, J. J.; Fan, M.; Chen,Z.; Jansen, J. A. Pharm. Res. 2011, 28, 1259−1272.(14) Sill, T. J.; von Recum, H. A. Biomaterials 2008, 29, 1989−2006.(15) Wenguo, C.; Yue, Z.; Jiang, Ch. Sci. Technol. Adv. Mater. 2010,11, 014108.(16) Zussman, E. Polym. Adv. Technol. 2011, 22, 366−371.(17) Xie, J. B.; Hsieh, Y. L. J. Mater. Sci. 2003, 38, 2125−2133.(18) Patel, A. C.; Li, S.; Yuan, J. M.; Wei, Y. Nano Lett. 2006, 6,1042−1046.(19) Dror, Y.; Kuhn, J.; Avrahami, R.; Zussman, E. Macromolecules2008, 41, 4187−4192.(20) Song, T.; Zhang, Y.; Zhou, T.; Lim, C. T.; Ramakrishna, S.; Liu,B. Chem. Phys. Lett. 2005, 415, 317−322.(21) Yarin, A. L. Polym. Adv. Technol. 2011, 22, 310−317.(22) Moghe, A. K.; Gupta, B. S. Polym. Rev. 2008, 48, 353−377.(23) Dror, Y.; Salalha, W.; Avrahami, R.; Zussman, E.; Yarin, A. L.;Dersch, R.; Greiner, A.; Wendorff, J. H. Small 2007, 3, 1064−1073.(24) Saraf, A.; Lozier, G.; Haesslein, A.; Kasper, F. K.; Raphael, R. M.;Baggett, L. S.; Mikos, A. G. Tissue Eng., Part C 2009, 15, 333−344.(25) Yang, D.-Z.; Long, Y.-h.; Nie, J. Front. Mater. Sci. China 2008, 2,261−265.(26) Liao, I. C.; Chew, S. Y.; Leong, K. W. Nanomedicine (London,U.K.) 2006, 1, 465−471.(27) Wang, C. Y.; Liu, J. J.; Fan, C. Y.; Mo, X. M.; Ruan, H. J.; Li, F.F. J. Biomater. Sci., Polym. Ed. 2012, 23, 167−184.(28) Sahoo, S.; Ang, L. T.; Goh, J. C.-H.; Toh, S.-L. J. Biomed. Mater.Res. A 2010, 93A, 1539−1550.

Biomacromolecules Article

dx.doi.org/10.1021/bm2018118 | Biomacromolecules 2012, 13, 952−962961

(29) Mufamadi, M. S.; Pillay, V.; Choonara, Y. E.; Du Toit, L. C.;Modi, G.; Naidoo, D.; Ndesendo, V. M. K. J. Drug Delivery 2011, 2011,119.(30) Sharma, A.; Sharma, U. S. Int. J. Pharm. 1997, 154, 123−140.(31) Mayer, L. D.; Bally, M. B.; Hope, M. J.; Cullis, P. R. Chem. Phys.Lipids 1986, 40, 333−345.(32) Lukas, D.; Sarkar, A.; Pokorny, P. J. Appl. Phys. 2008, 103,084309.(33) Lukas, D.; Chaloupek, J. In Thermal and Moisture Transport inFibrous Materials; Pan, N., Gibson, P., Eds.; Woodhead PublishingLimited: Cambridge, 2006; p 42−101.(34) Masounave, J.; Rollin, A. L.; Denis, R. J. Microsc. 1981, 121, 99−110.(35) Reznik, S. N.; Yarin, A. L.; Zussman, E.; Bercovici, L. Phys. Fluids2006, 18, 062101.(36) Salalha, W.; Kuhn, J.; Dror, Y.; Zussman, E. Nanotechnology2006, 17, 4675.(37) Zha, Z.; Leung, S. L.; Dai, Z.; Wu, X. Appl. Phys. Lett. 2012, 100(3), 033702−3.(38) Jiang, H.; Hu, Y.; Zhao, P.; Li, Y.; Zhu, K. J. Biomed. Mater. Res.,Part B 2006, 79, 50−57.(39) Srikar, R.; Yarin, A. L.; Megaridis, C. M.; Bazilevsky, A. V.;Kelley, E. Langmuir 2008, 24, 965−974.(40) Tiwari, S. K.; Tzezana, R.; Zussman, E.; Venkatraman, S. S. Int. J.Pharm. 2010, 392, 209−217.(41) Chang, H. I.; Yeh, M. K. Int. J. Nanomed. 2012, 7, 49−60.(42) Torchilin, V. P. Nat. Rev. Drug Discovery 2005, 4, 145−160.(43) Shabani, I.; Haddadi-Asl, V.; Seyedjafari, E.; Babaeijandaghi, F.;Soleimani, M. Biochem. Biophys. Res. Commun. 2009, 382, 129−133.(44) Pei, M.; He, F.; Vunjak-Novakovic, G. Differentiation 2008, 76,1044−1056.(45) Holland, T. A.; Mikos, A. G. J. Controlled Release 2003, 86, 1−14.(46) Baker, B. M.; Gee, A. O.; Metter, R. B.; Nathan, A. S.; Marklein,R. A.; Burdick, J. A.; Mauck, R. L. Biomaterials 2008, 29, 2348−2358.(47) Guimaraes, A.; Martins, A.; Pinho, E. D.; Faria, S.; Reis, R. L.;Neves, N. M. Nanomedicine (London, U.K.) 2010, 5, 539−554.(48) Ekaputra, A. K.; Prestwich, G. D.; Cool, S. M.; Hutmacher, D.W. Biomacromolecules 2008, 9, 2097−2103.(49) Li, L.; Qian, Y.; Jiang, Ch.; Lv, Y.; Liu, W.; Zhong, L.; Cai, K.; Li,S.; Yang, L. Biomaterials 2012, 33, 3428−3445.(50) Nam, J.; Huang, Y.; Agarwal, S.; Lannutti, J. Tissue Eng. 2007,13, 2249−2257.(51) Simonet, M.; Schneider, O. D.; Neuenschwander, P.; Stark, W. J.Polym. Eng. Sci. 2007, 47, 2020−2026.(52) Lee, J. B.; Jeong, S. I.; Bae, M. S.; Yang, D. H.; Heo, D. N.; Kim,Ch. H.; Alsberg, E; Kwon, K. Tissue Eng., Part A 2011, 17, 21−22.(53) Li, D.; Ouyang, G.; McCann, J. T.; Xia, Y. Nano Lett. 2005, 5,913−916.(54) Zussman, E. Appl. Phys. Lett. 2003, 82, 973.

Biomacromolecules Article

dx.doi.org/10.1021/bm2018118 | Biomacromolecules 2012, 13, 952−962962