Nanofibers from Blends of Polyvinyl Alcohol and ... - ACS Publications

Nov 22, 2010 - Institute of Physics, Martin Luther University Halle-Wittenberg, von ... Halle-Wittenberg, Heinrich-Damerow-Strasse 4, 06120 Halle (Saa...
0 downloads 0 Views 2MB Size
Biomacromolecules 2010, 11, 3413–3421

3413

Nanofibers from Blends of Polyvinyl Alcohol and Polyhydroxy Butyrate As Potential Scaffold Material for Tissue Engineering of Skin Ashraf Sh. Asran,*,†,§ Khashayar. Razghandi,‡ Neha Aggarwal,‡ Goerg H. Michler,† and T. Groth‡ Institute of Physics, Martin Luther University Halle-Wittenberg, von Danckelmann Platz 3, D-06099 Halle/ S., Germany, Biomedical Materials Group, Institute of Pharmacy, Martin Luther University Halle-Wittenberg, Heinrich-Damerow-Strasse 4, 06120 Halle (Saale), Germany, and National Research Centre, El Buhoth Street, 12311 Cairo, Egypt Received August 9, 2010; Revised Manuscript Received October 24, 2010

Nanofibers were prepared by electrospinning from pure polyvinyl alcohol (PVA), polyhydroxybutyrate (PHB), and their blends. Miscibility and morphology of both polymers in the nanofiber blends were studied by Fourier transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), transmission electron microscopy (TEM), and differential scanning calorimetry (DSC), revealing that PVA and PHB were miscible with good compatibility. DSC also revealed suppression of crystallinity of PHB in the blend nanofibers with increasing proportion of PVA. The hydrolytic degradation of PHB was accelerated with increasing PVA fraction. Cell culture experiments with a human keratinocyte cell line (HaCaT) and dermal fibroblast on the electrospun PHB and PVA/PHB blend nanofibers showed maximum adhesion and proliferation on pure PHB. However, the addition of 5 wt % PVA to PHB inhibited growth of HaCaT cells but not of fibroblasts. On the contrary, adhesion and proliferation of HaCaT cells were promoted on PVA/PHB (50/50) fibers, which inhibited growth of fibroblasts.

1. Introduction Skin is the largest organ of the human body and possesses many different functions including protection against heat, injury, and infections.1 Therefore, damage of larger skin areas is a life-threatening situation requiring intensive medical care. Basically, skin is built up of two main layers. The superficial one called the epidermis formed from keratinocytes acts as a barrier against infection and moisture loss. Below the epidermis lies the dermis, formed primarily from fibroblast cells. Living epidermal substitutes composed of keratinocytes only emerged in the 1980s2 for the treatment of burn victims and healing of chronic or deep wounds and diabetes-related ulcers;3-5 however, these epidermal sheets have certain disadvantages like fragility and poor take rates.6 Hence, attempts have been made by different groups to generate suitable scaffolds for living skin substitutes.7,8 In general, the ideal scaffold for skin substitution should have both a dermal and an epidermal component. The scaffold for dermal equivalents must promote adhesion, growth, and function of fibroblasts, whereas materials for the epidermal equivalent should solely promote keratinocytes but inhibit colonization with fibroblasts. One important feature of skin substitute scaffolds is their large inner surface area, which shall support attachment, proliferation, and migration of cells.3,9 Electrospun polymeric nanofiber scaffolds have attracted a great interest in tissue engineering because of their nanostructure morphology, which shall mimic the rather random distribution of fibrillar extracellular matrix (ECM) components like collagen.10 Moreover, electrospun nanofiber scaffolds have desir* To whom correspondence should be addressed. Tel: +49 3455525405. Fax: +49 34555 27149. E-mail: [email protected]. † Institute of Physics, Martin Luther University Halle-Wittenberg. ‡ Institute of Pharmacy, Martin Luther University Halle-Wittenberg. § National Research Centre.

able features as skin substitute materials because of their large aspect ratio and high porosity, which is required for nutrient delivery, fluid absorption and excretion, and oxygen supply.11 Over the past years, attention has been focused on the production, processing, and potential applications of polyhydroxyalkanoates (PHAs) and their copolymers in the biomedical field.12,13 Poly-3-hydroxybutyrate (PHB) is probably the most common type of PHA. It has been demonstrated that PHB is a biodegradable and biocompatible thermoplastic polymer with a high degree of crystallinity,10,14 displaying properties similar to polypropylene.15 It has been evaluated as a suitable material for a wide variety of medical applications including but not limited to controlled release systems, surgical sutures, wound dressings, orthopedic devices, bone tissue engineering, and skin substitute materials.10,16,17 Currently the main problem, which limits the widespread utilization of PHB, is its high crystallinity and brittleness.18 One approach to improve its physical and mechanical properties can be made through blending PHB with flexible or plasticizing polymers and compounds.19,20 Several reports showed that PHB is miscible with poly(ethy1ene oxide) (PEO),21 poly(viny1 acetate) (PVAc),22 poly(vinyl chloride) (PVC),23 and poly(viny1idene fluoride) (PVdF).24 Polyvinyl alcohol (PVA) is a water-soluble synthetic polymer that is used in a wide range of industrial, food, and medical applications.25 It is a hydrophilic, semicrystalline polymer with good chemical and thermal stability.26 Furthermore, it is a biodegradable, nontoxic, and noncarcinogenic polymer with excellent mechanical properties.27,28 Because of the interesting properties of PVA, attempts have been made to prepare films from PVA/ PHB blends.29,30 The polymer blend showed a partial miscibility of the two components in the amorphous region.31 Moreover, a lower crystallinity for both PVA and PHB was observed in the blend films.32 In addition, PVA significantly improved the me-

10.1021/bm100912v  2010 American Chemical Society Published on Web 11/22/2010

3414 Biomacromolecules, Vol. 11, No. 12, 2010

chanical properties of the blend compared with the pure PHB. It also contributed to a faster degradation of PHB.33 Mixing both phases of PVA and PHB at the macromolecular region is already difficult when the solution casting method is used because PVA is more hydrophilic than PHB. However, no information exists if good blending of both polymers will occur during formation of electrospun PVA/PHB nanofibers. Therefore, it was the aim of the present work to fabricate nanofibers from different PVA/PHB blends. In particular, the miscibility of both polymers in the macromolecular region should be investigated by a variety of physical methods. Furthermore, the obtained nanofiber scaffolds made from different PVA/PHB blends were investigated for their biocompatibility versus HaCaT cells (a keratinocyte cell line) and human dermal fibroblasts to learn about their potential application as a skin substitute material. Results are reported herein.

2. Experimental Section 2.1. Materials. PVA, with an average molecular weight of Mw ) 85 000-124 000 g/mol (87-89% hydrolyzed), PHB with an average molecular weight of Mw ) 4.4 × 105 g/mol, and 1,1,1,3,3,3-hexafluoro2-propanol (HFIP) were purchased from Sigma-Aldrich (Deisenhofen, Germany) and used without further treatment or purification. 2.2. Fabrication of Electrospun PVA/PHB Blend Nanofibrous Scaffolds. PVA and PHB were separately dissolved in HFIP and vigorously mixed with a magnetic stirrer at room temperature for 4 h to prepare 3 wt % homogeneous solutions from each polymer. The prepared solutions were combined subsequently (w/w) to prepare PVA/ PHB blends of different compositions: PVA/PHB (05/95), PVA/PHB (10/90), PVA/PHB (30/70), and PVA/PHB (50/50) solutions. Electrospinning was carried out at room temperature in a vertical spinning configuration using voltages in the range from 16 to 20 kV, driven by a high voltage power supply Heinzinger PNC (Heinzinger electronic GmbH, Rosenheim, Germany) with a flow rate of 100 µL/h. The polymer solutions were filled in a 1 mL syringe equipped with a blunt steel needle of 0.8 mm inner diameter. A rectangular steel plate covered with aluminum foil was placed 15 cm away from the needle tip as counter electrode. For morphological investigations, a glass plate was placed over the counter electrode as collecting substrate. In addition, the electrospun fibers were collected on a grounded stainless steel network to obtain scaffolds for further investigations. 2.3. Characterization of PVA/PHB Blends. 2.3.1. Morphology of the Electrospun Nanofibers. The nanofibers morphology was studied using scanning electron microscopy (SEM). Samples for SEM were prepared by direct electrospinning of pure PVA, PHB, and the different PVA/PHB blends on glass slides, followed by Au sputtering to ∼20 nm thicknesses for a better conductivity during imaging. The size and morphology of the electrospun fibers were investigated with a JEOL JSM 6300 apparatus. We determined average fiber diameters and their size distributions by measuring over 200 fibers selected randomly from the SEM images using image analysis software (AnalySIS, Soft Imaging System, Münster, Germany). In addition, the scaffold porosity has been measured from the SEM images using Image J (National Institutes of Health, Bethesda, Maryland, USA) according to the method of Ghasemi-Mobarakeh et al.34 2.3.2. Phase BehaVior of the Electrospun PVA/PHB Nanofibers. Miscibility of both polymers in the electrospun PVA/PHB (50/50) nanofibers was investigated by conventional transmission electron microscopy (TEM) (JEOL 200CX operated at 200 kV). Samples were prepared by depositing nanofibers directly onto Cu grids of 3 mm diameter covered with an ultrathin carbon layer. The grid was then placed into a sealed chamber with osmium tetroxide (OsO4) vapor for 4 days. The vapors of OsO4 attack the hydroxyl group in PVA, increasing its contrast compared with PHB during TEM imaging.35 2.3.3. Contact Angle Measurements. Thin films of pure PVA, PHB and their blends (dissolved in HFIP) were prepared on glass slides by

Asran et al. spin coating (at 350 rpm) with “SPIN150-NPP spin coater by SPSEurope B.V” (Putten, The Netherlands). The static values of water contact angle (WCA) on the prepared films were measured in a humidity chamber with the contact angle unit OCA-15+ from Dataphysics Instruments (Filderstadt, Germany). The sample and needle of the device were enclosed in a chamber to control the humidity (relative humidity was of ∼50%). At least three measurements on different film locations were averaged for data analysis. 2.3.4. Zeta Potential Measurement. Streaming potential measurements were performed using a SurPASS (Anton Paar, Graz, Austria) to measure zeta potentials. Films of pure PVA, PHB, and their blends were prepared by solution casting (1.5 wt % polymer in HFIP). Two films were attached on the sample holders and introduced to the flow cell. We determined the zeta potential (ζ) by adjusting the gap of the flow cell to a distance where a flow rate between 100 and 150 mL/min at a maximum pressure of 300 mbar was reached. A flow check was performed to achieve a constant flow in both directions. Potassium chloride (1 mM) was used as electrolyte, whereas 0.1 N hydrochloric acid was used for pH titration. Before a measurement was started, the pH value of the electrolyte solution was adjusted to pH 10.5 using 1 N sodium hydroxide. Finally, an automated titration program performed the measurement from pH 10.5 to 2.25 using titration steps of 0.03 µL from pH 10.5 to 5.0 and 0.25 µL from pH 5.0 to 2.25. 2.3.5. Differential Scanning Calorimetry. Thermal characteristics such as glass-transition temperature, melting temperature, degree of crystallization, and thermal degradation of the pure PVA granules, PHB powder, electrospun PVA, PHB, and their blend nanofibers were investigated using differential scanning calorimetry (DSC). Thermal measurements have been done in heat flux mode under nitrogen as purge gas using (DSC Mettler Toledo 820, Giessen, Germany). About 3 mg of sample (pure polymers and electrospun fibers) was heated and cooled in the temperature range -30 to 200 °C in the DSC instrument at a rate of 10 K/min. First, the samples were heated from -30 to 200 °C and kept at 200 °C for 2 min to eliminate any thermal history and then cooled to -30 °C for complete crystallization of the matrix. The heat evolved during nonisothermal crystallization was recorded as a function of time. The specimens were heated again from -30 to 200 °C without prior cooling to obtain the DSC endotherms. The melting temperatures were determined from the maxima of the fusion peaks. Furthermore, the degree of relative crystallinity (Xc) was estimated from the endothermic area using eq 1.

Xc ) ∆Hf /∆Hfo

(1)

where ∆Hf is the measured enthalpy of fusion from DSC thermograms and ∆Hfo is the enthalpy of fusion for 100% crystalline polymer. Furthermore, the degree of crystallinity of PHB in the blends was calculated by assuming that the ∆Hfo value of 100% crystalline PHB is 146 J g-1 36 and ∆Hfo of 100% crystalline PVA is 156 J g-1.37 In addition, crystallization temperatures (Tc) and heat of crystallization (∆Hc) were determined from exothermal peaks in cooling run. Cold crystallization temperatures (Tcc), heat of cold crystallization (∆Hcc), and melting temperature (Tm) were obtained from the second endothermic heating runs. 2.3.6. Fourier Transform Infrared Spectroscopy (FTIR). Infrared spectra of electrospun PVA, PHB, and their blend nanofibers were obtained using FTIR spectrometer S2000, Perkin-Elmer equipped with a fixed 100 mm diameter aperture. A mercury cadmium-telluride (MCT) detector was used to analyze the absorbance in the wavenumber range of 500-4000 cm-1 with a resolution of 2 cm-1. FTIR spectroscopy was carried out to analyze any complex structural changes that might have occurred because of blending and also to elucidate the interaction (hydrogen bonding) between PVA and PHB in the blend nanofibers. 2.3.7. Degradation Measurements. Electrospun nanofibrous scaffolds of different compositions were cut into square shapes weighing ∼3-6 mg and placed in wells of 12-well plates. Wells were filled with phosphate-buffered saline (PBS; 2.7 mM KCl, 137 mM NaCl, 1.4 mM

Nanofibers from Blends of PVA and PHB

Biomacromolecules, Vol. 11, No. 12, 2010

3415

Figure 1. SEM micrographs of the electrospun fibers of (a) pure PHB, (b) PVA/PHB (50/50), and (c) pure PVA using HFIP as a solvent for both polymers.

KH2PO4, 4.3 mM Na2HPO4 dihydrate, pH 7.4). The samples were incubated at 37 °C to mimic the biological environment of the body. The PBS buffer was changed every 3 to 4 days, and samples were taken out after 1, 3, 6, 9, and 12 weeks and washed with micropure water and then dried to constant weight. The weight of samples was measured to estimate their weight loss. The degradation rate was determined by the ratio of the weight after incubation in PBS for a given time to the initial weight of samples, as shown in eq 2.

S)

Wo - Wt × 100 Wo

(2)

where S is the degradation ratio, Wo is the initial sample weight, and Wt is the weight of dried samples.

3. Cell Culture 3.1. Sterilization of Nanofiber Mats. The nanofiber mats were backed up by transparent polypropylene (PP) foils for better handling during cell culture. The transparent PP foils were washed before their use as support for 2 h in a solution mixture composed of ethanol/acetone (50/50) to remove any contaminations from their surface. The nanofiber mats supported by PP foils were cut into discs with a diameter of 14 mm. The prepared samples were placed in 24 well-plates and supported with glass rings to avoid their floating in the medium. Thereafter, samples were sterilized for 1 h with 70% ethanol followed by rinsing three times with sterilized PBS before cell seeding. 3.2. Cells Culture. Human Fibroblasts and the HaCaT (Keratinocytes) cell line were used to measure cell adhesion and surface coverage by cells after different days of culture. Primary fibroblasts were purchased from Cell Lining GmbH (Berlin, Germany). HaCaT cell lines were a kind gift from Prof. Dr. P. Boukamp from the German Cancer Research Centre (DKFZ, Heidelberg, Germany). Both cell types were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Biochrom AG, Germany) supplemented with 10% v/v fetal bovine serum (FBS, Biochrom AG, Germany), 1% Pen/Strep/Fungizone (AAs, PromoCell, Heidelberg, Germany), and 1% L-glutamine (Sigma, Germany) and were incubated in humidified incubator at 37 °C

with 5% CO2 (NuAire, Plymouth, Minnesota). For the experiments, cells from preconfluent cultures were harvested with 0.25% Trypsin/0.02% EDTA solution (PromoCell, Heidelberg, Germany) and resuspended in culture medium DMEM with 10% FBS. 3.3. Cells Adhesion and Proliferation Studies. Sterilized pure PHB and PVA/PHB blend nanofiber mats were placed in 24-well tissue culture plates for studies of cell attachment and growth. Plain tissue culture polystyrene (TCP) was used as a control. HaCaT cells were seeded at a concentration of 25.000 cells/mL, whereas human dermal fibroblasts were seeded at a density of 15 000 cells/mL. Samples were incubated at 37 °C, 5% CO2 for the times indicated. The medium was changed every 48 h. Attachment, distribution, and growth of cells were visualized by vital staining with fluorescein diacetate (FDA, Sigma, Deisenhofen, Germany) with confocal laser scanning microscopy “LSM 710” (Carl-Zeiss, Oberkochen, Germany). FDA staining was performed by exchanging the medium with 1 mL of DMEM plus 5 µL of FDA stock solution (5 mg/mL acetone), followed by an incubation for 5 min at 37 °C, 5% CO2. Thereafter, staining medium was replaced by DMEM only, and 15 images of each scaffold were taken by CLSM at an excitation wavelength of 485 nm and an emission of 520 nm. The percentage of surface coverage of samples with cells was calculated by image processing software “ImageJ, NIH” and used for quantitative analysis.

4. Results and Discussion 4.1. Morphology of Electrospun PVA/PHB Blend Nanofibers. Figure 1 shows the morphology of electrospun pure PVA, PHB, and PVA/PHB (50/50). It is presented that PHB and PVA/PHB (50/50) nanofibers have a uniform structure without any sign of beads formation. In addition, the average size diameter of PHB is ∼680 nm, and PVA/PHB (50/50) is ∼615 nm (Figure 2). Nevertheless, the majority of fiber diameters in both cases were in the range of 200-1100 nm. An increase in PVA fraction of >50 wt % in the blend resulted in a nonhomogeneous structure with small beads and ribbonlike morphology. Pure PVA also showed nonhomogeneous

3416

Biomacromolecules, Vol. 11, No. 12, 2010

Asran et al.

Figure 2. Fiber diameter distribution of electrospun mats of (a) pure PHB and (b) PVA/PHB (50/50) blend nanofibers.

Figure 3. TEM micrograph of chemically stained PVA/PHB (50/50) nanofibers, stained with OsO4 vapor for 4 days (a,b) in a lower and higher magnification (fibers have an average diameter 1.5 µm) showed phase separation between the two polymers (Figure 4). The nonhomogeneity of the thicker fibers can be related to the tendency of PVA to be agglomerated on the fiber surface, making microphase separation.39 This result revealed that the reduction of the fibers diameter decreases the molecular segregation of the polymers in the blend. 4.3. Water Contact Angle Measurements. Films of PVA, PHB, and their blends prepared by spin-coating were used for WCA measurements. The static WCA values are summarized in Table 1, where zeta potential values are also shown. It is evident that pure PHB exhibited a less hydrophilic surface with

Figure 4. TEM micrograph of chemically stained PVA/PHB (50/50) nanofibers, stained with OsO4 vapor for 4 days (fibers have an average diameter >1.5 µm).

WCA of ∼70°. However, the addition of PVA to PHB gradually decreased the values of WCA for the blend films. PVA/PHB (50/50) possessed a more hydrophilic surface possibly because of the segregation of PVA fraction on the surface of the blend film. Moreover, PVA/PHB blend films containing 30 and 50 wt % PVA were found to have similar WCA of ∼41°. This indicates that the surface is covered and saturated with PVA molecules. This assumption is in agreement with the results of Ikejima et al.40 4.4. Zeta Potential. Zeta potential is related to the quantity and dissociation of charged groups on the material surface and dependent on ionic strength and pH value of electrolyte solutions.41 Results of zeta potential measurements in dependence of pH value for different blend compositions are presented in Figure 5 and summarized in Table 1. It is shown that pure

Nanofibers from Blends of PVA and PHB

Biomacromolecules, Vol. 11, No. 12, 2010

Table 1. Water Contact Angle, Zeta Potential and pH Values at Zero Charge for Spin-Coated Films of PHB, PVA, and Their Blend with Different Compositions

sample name

WCA (deg)

pure PHB PVA/PHB (05/95) PVA/PHB (10/90) PVA/PHB (30/70) PVA/PHB (50/50) pure PVA

70 62 54.8 41.7 41 37.3

zeta potential at pH 7.54 (mV)

pH (point of zero charge)

-16 -12.6 -6.9 -2.3 -2.1

6.9 6.5 5.9 5.6 5.1

PHB has a zeta potential of ∼45 mV at acidic pH, whereas at alkaline pH, PHB becomes negatively charged (∼ -42 mV). It was previously reported that hydroxyl ions preferentially adsorb onto aqueous-hydrophobic interfaces, resulting in a net negative surface charge density of nonpolar surfaces.42 The addition of a small quantity of PVA (5 wt %) to PHB resulted in a significant decrease in zeta potential values for the blend films, and these values further decreased with increasing PVA fraction. This decrease in zeta potential seems to be a reflection of the decrease in the number of surface charges attributable to the presence of PVA molecules.43 Furthermore, zeta potentials for films with PVA content of 30 and 50 wt % were almost the same, which can be interpreted as another sign that the surface of the films becomes saturated with PVA molecules. Zeta potential values for all polymer compositions were found to be negative at pH 7.5 (biological pH), as shown in Table 1. However, the negative magnitude decreased gradually with increasing PVA fraction, obviously due to the presence of hydroxyl groups (Table 1). Altankov et al. also demonstrated with self-assembled monolayers on glass that polar hydroxyl groups (OH) diminish the negative magnitude of zeta potential, whereas carboxylic group (COOH) increased it.44 Table 1 demonstrates in addition that the isoelectric point (or point of zero charge) was in the acidic pH range and shifted to lower pH values with increasing PVA fraction compared with pure PHB. Such changes might be attributed to the specific adsorption of anions (hydroxyl ions), as previously discussed.45 Both WCA and zeta potential data strongly indicate that PVA molecules concentrate on the surface of the films. 4.5. Fourier Transform Infrared Spectroscopy. FTIR measurements have been performed to analyze the structural changes that might occur upon blending, as well as the

Figure 5. Zeta potential versus pH for films of (]) pure PHB, (×) PVA/PHB (05/95), (/) PVA/PHB (10/90), (4) PVA/PHB (30/70), and (O) PVA/PHB (50/50).

3417

interactions between (OH) groups of PVA and (CdO) groups of PHB, which consequently affect the miscibility and crystallinity of the blend nanofibers. The absorption peaks of the observed frequencies and their assignments are summarized in Table 2. Results obtained showed that the most evident characteristic bands for PHB and PVA were the carbonyl and hydroxyl group bands centered at ∼1720 and ∼3278 cm-1, respectively. These two molecular vibrations are important in infrared range absorption peaks; the investigation of these peaks in the blend gives important information about the intermolecular interaction (e.g., hydrogen bond) between two polymers in the blend and their compatibility.29 The other characteristic bands for PHB were observed at ∼1276, ∼1180, and ∼1057 cm-1, which can be assigned to the ester groups of the polymer.46 The bands centered at 980, 1227, 1276, and 1720 cm-1 were shown to arise from the crystalline phase of PHB. Pure PVA nanofibers showed two bands centered at 1714 and 1089 cm-1, which are corresponding to C-O of the remaining acetyl groups present on the partially hydrolyzed PVA back bone.39 The sensitivity of FTIR spectrum to crystallization of PHB has been investigated by Bloembergen et al.47 Moreover, analyzing the spectra of FTIR would give crucial evidence of both crystallization and miscibility of the blend polymers.48 In the present study, the electrospun PVA/PHB (50/50) blend nanofibers showed an increment in the crystalline-related carbonyl group bands of PHB to ∼1723 and 1284 cm-1. (See Table 2.) The increment of carbonyl band order can be attributed to the absence of an ordered structure (hydrogen-bonding effects are reduced), and thereby the wavenumber of absorbance will increase.49 Furthermore, the intensity of bands observed at 980 and 1227 cm-1 was decreased in the blend PVA/PHB (50/50) fibers. In addition, the absorption band of PVA hydroxyl group shifted to a higher frequency at ∼3311 cm-1 because of the presence of intermolecular hydrogen bonding interaction between the two polymers.50 Moreover, in the pure PHB nanofibers, a crystalline-related peak such as the peak at ∼2976 cm -1 for CH3 asymmetric and symmetric stretching disappeared in the blend fibers, which can be assumed to be a sign of depression of crystallization of PHB by the addition of PVA. Likewise, this phenomenon can be also verified by the disappearance of the band centered at ∼2871 cm-1 (CH stretching of PHB) and band at ∼1687 cm-1 (stretching of CdO of PHB). The crystalline-related peaks at 1417 and 1326 cm-1 in pure PVA disappeared in the blend fibers (Table 2), which can be a sign of interaction of the carbonyl groups of PHB with hydroxyl groups of PVA. Thereby, restrictions of these bands were obtained, and disappearance of wagging and bending movement of the hydroxyl groups were observed. The FTIR results revealed that the crystallinity of electrospun PVA/PHB (50/50) nanofibers was suppressed compared with pure PHB. Such a decrease in the total crystallinity can be considered to be a very important factor for acceleration of fiber degradation, decreasing the brittleness and improving the mechanical properties of the nanofibers. The intermolecular hydrogen bonds between CdO groups of PHB and OH groups of PVA found in the blend PVA/ PHB (50/50) fibers are considered to be evidence of the miscibility of the polymers. 4.6. Differential Scanning Calorimetry. The data derived from the second heating scans of DSC for PVA granules, PHB powder, electrospun PVA, PHB, and their blend nanofibers are given in Table 3. Results obtained show that PVA exhibited a first endothermic peak at ∼41.7 °C, which corresponds to the glass-transition temperature (Tg). Furthermore, crystallization

3418

Biomacromolecules, Vol. 11, No. 12, 2010

Asran et al.

Table 2. Assignments of FTIR Absorption Bands for Electrospun Pure PVA, Pure PHB, and PVA/PHB (50/50) Blend Nanofibers peaks for electrospun nanofibers (cm-1) PVA

PHB

3278 2912.82

1714.36 1656 1556.21

2976 2936.18 2871 1720

PVA/PHB (50/50)

assignment

3311.70

stretching of OH C-CH3 asymmetric stretching of CH2 CH stretching CdO carbonyl group stretching vibration of the crystalline carbonyl group stretching of CdO (acetate group) (crystalline) OdH, CdC CdC end group asymmetric deformation of CH3 O-H, C-H bending, γ(CH2), δ(OH) symmetric wagging of CH3 C-H bending or CH2 wagging C-O-H bending symmetric C-O-C stretching symmetric C-O-C stretching + C-H deformation C-O-C (acetate group) C-O-C stretching asymmetric C-O-C stretching symmetric stretching vibration of C-O-C group symmetric C-O-C stretching stretching of C-O and bending of OH (amorphous sequence of PVA) C-O stretching and CH2 rocking C-C stretching (crystalline) bending of CH2 rocking of CH2

2941.38 1723.23

1687 1452.30

1454.57

1380.24

1381.24

1276.67 1258.92

1284.23 1263.14

1227.44 1180.40 1130.11 1100.90

shoulder 1182.67 1130.43 1198.96

1057.10 980.85

1058 shoulder

1417 1374 1326.70 1233.61

1089.92

918.66 831.87

828.03

Table 3. DSC Data and the Characteristics Observed for PVA Granules, PHB Powder, Electrospun PVA, PHB, and Their Blend Nanofibers with Different Compositions sample

Tm1 (°C)

PVA granules PVA nanofibers PHB powder PHB nanofibers PVA/PHB (05/95) PVA/PHB (10/90) PVA/PHB (30/70) PVA/PHB (50/50)

218 219 147.2 152.2 134.9 133.4 116.6 102

a

Tm for PVA fraction.

b

Tm2 (°C)

Tc first crystallization temperature

∆Hm (J · g-1)

crystallinity (%)

159.68 164.50 152.19 151.59 134.34 144.15 186.51a 122.26 135.29 186.15a

179 190 76.1 81.4 65.4 82.3 161.9 159.6

65.3 61.4 97 98.5 84 74 53.5 24

47 44 66.43 67.46 60.56b 56.31b 52.34b 32.87b

Crystallinity of PHB fraction in the blend nanofibers.

Figure 6. Degradation of electrospun nanofibrous mats of PHB, PVA/PHB (05/95), PVA/PHB (10/90), PVA/PHB (30/70), and PVA/PHB (50/50). Degradation was measured by weighing the nanofibers scaffolds after degradation in phosphate-buffered saline within 3 months.

temperature (Tc) and melting temperature (Tm) of PVA granules were observed at ∼179 and ∼218 °C, respectively. The Tg of PVA nanofibers was found to be ∼48.3 °C with a slight increment after electrospinning. Similarly, electrospun PVA

nanofibers demonstrated a low increment in Tc and Tm peaks to ∼190 and ∼219 °C, respectively. These increments indicate that the PVA chains were highly confined in the electrospun fibers, which limits the chain mobility and thereby increases Tg, Tc,

Nanofibers from Blends of PVA and PHB

Biomacromolecules, Vol. 11, No. 12, 2010

3419

Figure 7. HaCaT cell growth on the electrospun PVA/PHB nanofibers mats with different blend compositions after 3, 7, and 14 days.

and Tm values.51 Likewise, an increment of Tc, Tm, and degree of crystallinity were observed for the electrospun PHB nanofibers compared with PHB powder. The increasing degree of crystallinity for PHB fibers can be also attributed to the orientation of polymer molecules in the fiber direction during electrospinning.52 Tg of PHB and their blends with PVA were not detectable because of the high crystallinity of PHB. Xing et al. suggested that Tg of blends must be between glasstransition temperatures of the two polymers.30 Furthermore, pure PHB exhibited a cold crystallization peak at ∼76 °C, suggesting that PHB crystallizes more rapidly at a lower temperature compared with PVA. In addition, PHB exhibited double melting peaks at ∼147 and ∼159 °C. The double or multiple melting behavior of PHB can be attributed to secondary crystallization (melting-recrystallization-remelting), which can occur during the melting process as well as during storage at room temperature.53,54 Furthermore, in the electrospun PHB and their blends with PVA, the cold crystallization of PHB was observed, which normally takes place above Tg of the blend, where crystallizable polymer chains have enough segmental mobility to crystallize. The crystallinity of both PVA and PHB was reduced by the blending (even with 5 wt % PVA) because of some specific interactions between them (e.g., hydrogen bonds). The interaction between PVA and PHB occurs only in the amorphous phase, whereas the crystalline fractions of both components do not influence each other. Another factor other than hydrogen bonding that lowers the degree of crystallinity of PHB is the difference in the molecular mobility between PHB and PVA. The difference in the glass-transition temperature between PVA and PHB (∼60-80 °C) influences the degree of crystallinity of their blends. The “glassy” environment of PVA seems to trap the molecules of PHB in the amorphous phase and hinders their mobility required for crystallization, which results in a lower degree of crystallinity. Therefore, the influence of PVA on the crystallinity of PHB can be considered to be proof of good miscibility (compatibility) of the two polymers. Others found as well that PHB is partially miscible with PVA in the amorphous region.40 Moreover, Yoshie et al. found that PVA and PHB are compatible only when the blend contains a large amount of PVA.31 A similar behavior has been previously

Figure 8. HaCaT cells surface coverage on the electrospun PVA/PHB nanofibers mats with different blend compositions after 3, 7, and 14 days.

observed for blends of PHB with chitin and chitosan.55 The DSC results in this work revealed that blending PVA with PHB suppressed the crystallinity of both polymers in the blend fibers. In addition, an interaction between both polymers in the electrospun PVA/PHB blend fibers was evident. Such interaction can be considered to be a proof of miscibility of the two polymers in the macromolecular region. 4.7. In Vitro Degradation of Electrospun PVA/PHB Nanofibrous Scaffold. Figure 6 shows the hydrolytic degradation of electrospun PHB and PVA/PHB nanofibers mats. It was found that all type of fibers degraded during the time of observation. Degradation was detectable already after 1 week for the nanofiber mats of PVA-rich composition (e.g., 50/50). A sudden onset of degradation of pure PHB nanofibers occurred after 9 weeks. PHB nanofibers containing 30 and 50 wt % PVA showed a weight loss of ∼8 and ∼28% after 12 weeks, respectively. Pure PHB nanofibers showed ∼9% weight loss after 12 weeks, which is slightly higher than PVA/PHB nanofibers containing 30% PVA. This can

3420

Biomacromolecules, Vol. 11, No. 12, 2010

Asran et al.

Figure 9. Dermal fibroblast growth on the electrospun PVA/PHB nanofibers mats with different blend compositions after 3, 7, and 14 days.

be attributed to the different degradation mechanism in pure PHB nanofibers and PVA-containing nanofibers.27 The accelerated degradation of PVA/PHB nanofibers (containing 30 and 50 wt % PVA) corresponds to the lower degree of crystallinity of these blend nanofibers. Previous reports on biodegradation of PVA/PHB blend films related the biodegradation profile to the water solubility of the PVA components.33,55 The degradation results shown here prove that PVA accelerates the degradation of PHB. 4.8. Adhesion and Proliferation of Epidermal and Dermal Cells. 4.8.1. HaCaT Cells. Figures 7 and 8 show HaCaT cells adhesion and growth on the electrospun PHB and its blends with different PVA fractions. Figure 7 demonstrates that HaCaT cells attached and grew well on pure PHB nanofibers. Moreover, the surface coverage after 7 days was almost the same as that on the control surface (TCP), as demonstrated in Figure 8. Peschel et al. have shown that hydrophobic polymers have a negative effect on HaCaT cell growth.56 However, others recently suggested that PHB nanofibers may mimic the nanostructure of the native ECM, which might compensate the negative effect of hydrophobic PHB on the behavior of HaCaT cells.57,58 The addition of only 5 wt % PVA resulted in a significant decrease in the number of viable HaCaT cells on the nanofibers mats. On the contrary, further increasing PVA fractions in the nanofiber blends resulted in an increase in viable cells on the nanofibers scaffolds. Such increase in HaCaT cell numbers could be attributed to the saturation of PVA on the surface, as was found by surface analysis (WCA and zeta potential measurements). As previously discussed, HaCaT cells prefer more hydrophilic surfaces.56 Hence an increasing PVA content resulted in good adhesion and proliferation of HaCaT cells on the nanofibers scaffolds. Overall, HaCaT cells showed highest proliferation on electrospun pure PHB nanofibers (comparable to TCP) and relatively high adhesion and proliferation on PHB/PVA 50/50 nanofiber mats after 1 and 2 weeks, promising further proliferation and growth upon longer incubation times. 4.8.2. Dermal Fibroblasts. Qualitative and quantitative measurements of the adhesion and growth of dermal fibroblast are presented in Figures 9 and 10. Figure 9 clearly demonstrates

Figure 10. Dermal fibroblast cells surface coverage on the electrospun PVA/PHB nanofibers mats with different blend compositions after 3, 7, and 14 days.

that an increasing PVA content of the nanofibers had a strong negative impact on the attachment and proliferation of fibroblasts. It was previously reported that hydrophilic substrata have a negative effect on fibroblast proliferation and function; better cellular interaction was found on intermediate wettable surfaces.44 The decreased number of viable fibroblasts with increasing substrate hydrophilicity can be related to the high water uptake of PVA, probably reducing adsorption of attachment proteins.59 Water uptake of PVA/PHB blend nanofibers seems to have no negative impact on HaCaT cells, which could be related to their strong cell-cell adhesions absent in fibroblasts.60 Also, here pure PHB nanofibers were found to promote highest fibroblasts adhesion and proliferation compared with other blend compositions. However, nanofiber mats with 5-10% PVA content were also found to support adhesion and proliferation of fibroblasts but to a lower extent than pure PHB.

Nanofibers from Blends of PVA and PHB

5. Conclusions Biocompatible PVA/PHB blend nanofibers of different compositions were prepared by electrospinning technique for potential application in tissue engineering of skin. Morphological characterization and crystallization behavior indicated a good miscibility of PVA and PHB in the amorphous phase. Furthermore, PVA suppressed the crystallinity of PHB with increasing PVA fraction. In addition, PHB nanofibers and their blends with PVA were found to be degradable in a simulated body fluid. Degradation increased with increasing PVA content. The cell culture experiments demonstrated that the moderately hydrophobic surface of pure PHB nanofibers promoted highest cell adhesion and proliferation for both HaCaT and fibroblasts cells. An increase in the PVA fraction in the blend had a negative impact on fibroblast adhesion and growth, whereas that of HaCaT cells were increased. The obtained PVA/PHB nanofiber blend represents promising materials for fabrication of bilayered nanofiber scaffold for tissue engineering of skin. A top layer composed of PVA/PHB (50/50) blend fibers could promote adhesion and growth of HaCaT cells but could inhibit fibroblasts to generate an epidermal layer. The dermal equivalent could be prepared from pure PHB or PVA/PHB (10/90) to promote fibroblast adhesion and growth only. Such kind of bilayered nanofibers scaffold would mimic not only the morphological structure of native ECM of the skin but also its specific separation into an epidermal and dermal layer. Acknowledgment. We would like to thank Dr. Andre´ Wutzler for conducting the FTIR experiments. Special thanks are addressed to Mrs. S. Goerlitz for chemical staining and TEM imaging of the nanofibers.

References and Notes (1) Balasubramani, M.; Kumar, T. R.; Babu, M. Burns 2001, 27, 534– 544. (2) Hata, K.-i. J. Artif. Organs 2007, 10, 129–132. (3) Metcalfe, A. D.; Ferguson, M. W. J. J. R. Soc., Interface 2007, 4, 413–437. (4) van der Veen, V. C.; van der Wal, M. B.; van Leeuwen, M. C.; Ulrich, M. M.; Middelkoop, E. Burns 2009, 36, 305–321. (5) Gibbs, S.; van den Hoogenband, H. M.; Kirtschig, G.; Richters, C. D.; Spiekstra, S. W.; Breetveld, M.; Scheper, R. J.; de Boer, E. M. Br. J. Dermatol. 2006, 155, 267–274. (6) Lee, K. H. Yonsei Med. J. 2000, 41, 774–779. (7) Shores, J. T.; Gabriel, A.; Gupta, S. AdV. Skin Wound Care 2007, 20, 493–508. (8) Pham, C.; Greenwood, J.; Cleland, H.; Woodruff, P.; Maddern, G. Burns 2007, 33, 946–957. (9) Sheridan, R. L.; Morgan, J. R.; Cusick, J. L.; Petras, L. M.; Lydon, M. M.; Tompkins, R. G. Burns 2001, 27, 421–424. (10) Chen, G.-Q.; Wu, Q. Biomaterials 2005, 26, 6565–6578. (11) Venugopal, J.; Low, S.; Choon, A. T.; Ramakrishna, S. J. Biomed. Mater. Res., Part B 2008, 84, 34–48. (12) Mu¨ller, H.-M.; Seebach, D. Angew. Chem., Int. Ed. 1993, 32, 477– 502. (13) Philip, S.; Keshavarz, T.; Roy, I. J. Chem. Technol. Biotechnol. 2007, 82, 233–247. (14) Zhijiang, C. J. Mater. Sci.: Mater. Med. 2006, 17, 1297–1303. (15) Harding, K. G.; Dennis, J. S.; von Blottnitz, H.; Harrison, S. T. L. J. Biotechnol. 2007, 130, 57–66. (16) Galego, N.; Rosza, C.; Sanchez, R.; Fung, J.; Vazquez, A.; Tomas, J. Polym. Test. 2000, 19, 485–492. (17) Wu, Q.; Wang, Y.; Chen, G. Q. Artif. Cells, Blood Substitutes, Immobilization Biotechnol. 2009, 37, 1–12. (18) Reis, K. C.; Pereira, J.; Smith, A. C.; Carvalho, C. W. P.; Wellner, N.; Yakimets, I. J. Food Eng. 2008, 89, 361–369. (19) Songling, X.; Rongcong, L.; Linping, W.; Kaitian, X.; Guo-Qiang, C. J. Appl. Polym. Sci. 2006, 102, 3782–3790. (20) Lucia, H. I.-M.; Julio, R. B.; Rodrigo, C. B. Macromol. Symp. 2003, 197, 77–88. (21) Avella, M.; Martuscelli, E. Polymer 1988, 29, 1731–1737.

Biomacromolecules, Vol. 11, No. 12, 2010

3421

(22) Greco, P.; Martuscelli, E. Polymer 1989, 30, 1475–1483. (23) McCarthy, S. P.; Gross, R. Proceedings of Environmentally Degradable Polymers: Technical, Business, and Public Perspectives; Chelmsford, MA, 1991. (24) Kaito, A.; Li, Y.; Shimomura, M.; Nojima, S. J. Polym. Sci., Part B: Polym. Phys. 2009, 47, 381–392. (25) Chen, J.; Zhang, Y.; Du, G.-C.; Hua, Z.-Z.; Zhu, Y. Enzyme Microb. Technol. 2007, 40, 1686–1691. (26) Koski, A.; Yim, K.; Shivkumar, S. Mater. Lett. 2004, 58, 493–497. (27) Chiellini, E.; Corti, A.; D’Antone, S.; Solaro, R. Prog. Polym. Sci. 2003, 28, 963–1014. (28) Paradossi, G.; Cavalieri, F.; Chiessi, E.; Spagnoli, C.; Cowman, M. J. Mater. Sci.: Mater. Med. 2003, 14, 687–691. (29) Zhao, L.; Tsuchiya, K.; Inoue, Y. Macromol. Biosci. 2004, 4, 699– 705. (30) Xing, P.; Ai, X.; Dong, L.; Feng, Z. Macromolecules 1998, 31, 6898– 6907. (31) Yoshie, N.; Yoichiro, A.; Minoru, S.; Yoshio, I. J. Appl. Polym. Sci. 1995, 56, 17–24. (32) Azuma, Y.; Yoshie, N.; Sakurai, M.; Inoue, Y.; Chuˆjoˆ, R. Polymer 1992, 33, 4763–4767. (33) Ikejima, T.; Yoshie, N.; Inoue, Y. Polym. Degrad. Stab. 1999, 66, 263–270. (34) Ghasemi-Mobarakeh, L.; Semnani, D.; Morshed, M. J. Appl. Polym. Sci. 2007, 106, 2536–2542. (35) Michler, G. H. Electron Microscopy of Polymers; Springer Laboratory: Berlin, 2008. (36) Barham, P. J.; Keller, A.; Otun, E. L.; Holmes, P. A. J. Mater. Sci. 1984, 19, 2781–2794. (37) Tubbs, R. K. J. Polym. Sci., Part A: Polym. Chem. 1965, 3, 4181– 4189. (38) Xie, J.; Li, X.; Xia, Y. Macromol. Rapid Commun. 2008, 29, 1775– 1792. (39) Andrade, G.; Barbosa-Stancioli, E. F.; Mansur, A. A. P.; Vasconcelos, W. L.; Mansur, H. S. Biomed. Mater 2006, 1, 221–234. (40) Ikejima, T.; Cao, A.; Yoshie, N.; Inoue, Y. Polym. Degrad. Stab. 1998, 62, 463–469. (41) Mu¨ller, R. H.; Davis, S. S.; Illum, L.; Mark, E. In Targeting of Drugs with Synthetic Systems; Gregoriadis, G., Senior, J., Poste, G., Eds.; Plenum: New York, 1986, pp 239-263. (42) Tandon, V.; Bhagavatula, S. K.; Nelson, W. C.; Kirby, B. J. Electrophoresis 2008, 29, 1092–1101. (43) Isci, S.; Unlu, C. H.; Atici, O.; Gungor, N. Bull. Mater. Sci. 2006, 29, 449–456. (44) Altankov, G.; Richau, K.; Groth, T. Materialwiss. Werkstofftech. 2003, 34, 1120–1128. (45) Somasundaran, P. Encyclopedia of Surface and Colloid Science; CRC Press: New York, 2006. (46) Kecskemeti, G.; Smausz, T.; Kresz, N.; Toth, Z.; Hopp, B.; Chrisey, D.; Berkesi, O. Appl. Surf. Sci. 2006, 253, 1185–1189. (47) Bloembergen, S.; Holden, D. A.; Hamer, G. K.; Bluhm, T. L.; Marchessault, R. H. Macromolecules 1986, 19, 2865–2871. (48) Guo, L.; Sato, H.; Hashimoto, T.; Ozaki, Y. Macromolecules 2010, 43, 3897–3902. (49) Xu, J.; Guo, B.; Yan, R.; Wu, Q.; Chen, G.; Zhang, Z. Polymer 2002, 43, 6893–6899. (50) Yi, J. Z.; Goh, S. H. Polymer 2005, 46, 9170–9175. (51) Zong, X.; Kim, K.; Fang, D.; Ran, S.; Hsiao, B. S.; Chu, B. Polymer 2002, 43, 4403–4412. (52) Wong, S.-C.; Baji, A.; Leng, S. Polymer 2008, 49, 4713–4722. (53) El-Hadi, A.; Schnabel, R.; Straube, E.; Mu¨ller, G.; Henning, S. Polym. Test. 2002, 21, 613–733. (54) Gunaratne, L. M. W. K.; Shanks, R. A. Thermochim. Acta 2005, 430, 183–190. (55) Ikejima, T.; Inoue, Y. Carbohydr. Polym. 2000, 41, 351–356. (56) Peschel, G.; Dahse, H. M.; Konrad, A.; Wieland, G. D.; Mueller, P. J.; Martin, D. P.; Roth, M. J. Biomed. Mater. Res., Part A 2008, 85, 1072–1081. (57) Li, X. T.; Zhang, Y.; Chen, G. Q. Biomaterials 2008, 29, 3720–3728. (58) Sombatmankhong, K.; Sanchavanakit, N.; Pavasant, P.; Supaphol, P. Polymer 2007, 48, 1419–1427. (59) Groth, T.; Seifert, B.; Malsch, G.; Albrecht, W.; Paul, D.; Kostadinova, A.; Krasteva, N.; Altankov, G. J. Biomed. Mater. Res. 2002, 61, 290– 300. (60) Ren, Q.; Kari, C.; Quadros, M. R. D.; Burd, R.; McCue, P.; Dicker, A. P.; Rodeck, U. Cancer Res. 2006, 66, 5209–5215.

BM100912V