J Mater Sci: Mater Med DOI 10.1007/s10856-014-5214-4
Optimization of fully aligned bioactive electrospun fibers for ‘‘in vitro’’ nerve guidance Valentina Cirillo • Vincenzo Guarino • Marco Antonio Alvarez-Perez • Marica Marrese Luigi Ambrosio
•
Received: 17 December 2013 / Accepted: 28 March 2014 Ó Springer Science+Business Media New York 2014
Abstract Complex architecture of natural tissues such as nerves requires the use of multifunctional scaffolds with peculiar topological and biochemical signals able to address cell behavior towards specific events at the cellular (microscale) and macromolecular (nanoscale) level. In this context, the electrospinning technique is useful to generate fiber assemblies having peculiar fiber diameters at the nanoscale and patterned by unidirectional ways, to facilitate neurite extension via contact guidance. Following a bio-mimetic approach, fully aligned polycaprolactone fibers blended with gelatin macromolecules have been fabricated as potential bioactive substrate for nerve regeneration. Morphological and topographic aspects of electrospun fibers assessed by SEM/AFM microscopy supported by image analyses elaboration allow estimating an increase of fully aligned fibers from 5 to 39 % as collector rotating rate increases from 1,000 to 3,000 rpm. We verify that fully alignment of fibers positively influences in vitro response of hMSC and PC-12 cells in neurogenic way. Immunostaining images show that the presence of topological defects, i.e., kinks—due to more frequent fiber crossing—in the case of randomly organized fiber assembly concurs to interfere with proper neurite outgrowth. On the contrary, fully aligned fibers without kinks offer a more efficient contact guidance to direct the orientation of nerve cells along the fibers respect to randomly organized ones, V. Cirillo V. Guarino (&) M. Marrese L. Ambrosio Institute of Polymers, Composites and Biomaterials, National Research Council of Italy, V.le Kennedy 54, Pad 20, Mostra d’Oltremare, 80125 Naples, Italy e-mail: [email protected] M. A. Alvarez-Perez TBL-DEPeI, Universidad Nacional Auto´noma de Me´xico (UNAM), Coyoaca´n, Mexico, DF, Mexico
promoting a high elongation of neurites at 7 days and the formation of bipolar extensions. So, this confirms that the topological cue of fully alignment of fibers elicits a favorable environment for nerve regeneration.
1 Introduction It is general understood that the complex architecture of scaffolds, as well as the characteristics of the surrounding environment, may affect cell fate and functions [1]. Recent experimental evidences have suggested that physical control of cell shape mediated by peculiar forces on the surrounding microstructure may act as a potent regulator of stem cell fate [2]. For instance, protein based micropatterns in various geometries (circle, square, rectangle and ellipses) recently confirmed how surface concavity may influence the human mesenchymal stem cells (hMSC) behavior, inducing greater attractive forces than those on convex surfaces [3], so directly forcing cells into osteogenic instead of adipogenic lineages. Other studies have also indicated that the characteristic size scale of structural units may also affect the mechanisms of cell metabolism, offering new insights into the role of textured materials in specific cell activities [4]. However, not only the scale but also the topographic organization in rigid aligned patterns can also relevantly contribute to the final cell behavior [5]. In order to study the ability of cells to recognize micro and/ or sub-microtopography, it is mandatory to design synthetic platforms with peculiar structural order and anisotropy. Currently, electrospinning is emerging as a powerful technique for the production of micro and nanotextured fibrous scaffolds with specific fiber patterning [6]. Traditionally, electrospun nanofibers have been usually collected
123
J Mater Sci: Mater Med
as nonwoven mats with randomly oriented fibers assembled to form interconnected porous networks that mimic the collagenous fibrous architecture of the extracellular matrix (ECM) which is naturally present in hard and soft tissues. These peculiar fiber features may be just satisfactory to offer multiple possibilities in surface functionalization and biomimicking [7]. In particular, high surface area of the nanofibers helps to overcome loading limitations that are normally encountered in current drug delivery methods [8], despite this effect may generally corroborated by the addition of surface pores able to promote the formation of larger number of binding sites for drug loading [9]. However, main features of most native ECMs analogues of natural tissues or organs (e.g. heart, tendons, blood vessels) have to reproduce the exact anisotropic architecture of tissue which is crucial to determine, case by case, the specific function of tissues. In this context, a fine control of the spatial orientation and fiber patterning by an accurate design of electrospinning process conditions may assure a more efficient mimesis of the ECM in terms of structural and functional properties thanks to the possibility to build different patterns which reply the structural order of several tissues including nerve and muscles. Recent studies have just demonstrated that the aligned electrospun nanofiber arrays may promote significant improvements in cell proliferation respect to random fiber configuration in the case of neural stem cells [10], primary Schwann cells [11] and endothelial cells [12], also affecting the cell-growing direction and morphologies [13, 14]. In the specific case of nerve regeneration, aligned fibers may provide signaling cues to the cells and act as contact guidance, i.e. cells on the scaffold are promoted to collectively align along the desired direction [15]. So, axons may be guided to their targets by the cues directly provided by the substrate, which could be either topographical or chemical and biological cues or their combinations [16, 17]. Hence, continuous straight and aligned fibers with tailored material compositions are suitable as artificial substrate for neurite alignment and outgrowth. Here, we propose to study the optimization of electrospinning process conditions to obtain random and aligned electrospun fibers by mixing two different polymers, i.e. poly-e-caprolactone (PCL) and gelatin for peripheral nerve regeneration. PCL is a semi crystalline linear polyester largely used as biomaterial. However, its poor hydrophilicity reduces cell adhesion, migration, proliferation, and differentiation [18–20]. So, the inclusion of a hydrophilic protein derived from collagen denaturation such as gelatin, may improve cell recognition in order to obtain fiber scaffold with more controlled cell response and degradation properties. In a previous study, we have just demonstrated that the gelatin integrated into PCL fibers
123
may act as biochemical signal improving significantly the biointeraction of PC-12 pheochromocytoma nerve cells with the substrate comparing to PCL fibers as shown by results of cell attachment, viability, and neurite outgrowth [17]. Now, we propose to investigate PCL/gelatin fibers with preferential fiber patterning to underline the effect of alignment in neurogenic way by the biological interaction of two types of cells with different differentiation state, i.e., undifferentiated hMSC and rat PC-12, offering new information about the role of fiber patterning on cell behavior.
2 Materials and methods 2.1 Materials PCL (Mn 45 kDa) and gelatin (Type B from calf skin, *225 Bloom) were purchased from Sigma-Aldrich (Italy) and 1,1,1,3,3,3-hexafluoro-2-propanol (HFP) from Fluka (Italy) was used as received without further purification. High glucose-Dulbecco’s modified Eagle’s medium (HG-DMEM); Eagle’s alpha minimum essential medium (a-MEM), trypsin–EDTA solution, L-glutamin, fetal bovine serum, horse serum were purchased from Gibco. Antibiotic solution, nerve growth factor 2.5S from murine submaxillary gland (NGF), brain derived neurotrophic factor (BDNF), dimethyl sulfoxide (DMSO), b-mercaptoethanol, triton X-100 were supplied by Sigma. Epidermal growth factor (EGF) and basic fibroblast growth factor (bFGF) were purchase from Chemicon. Anti-Growth associated protein-43 (GAP-43) antibody was purchased from Millipore. 2.2 Electrospinning process PCL and gelatin were separately dissolved in HFP at a concentration of 0.1 g ml-1 and kept under magnetic stirring at room temperature for 24 h. Then, polymer solutions were mixed to obtain a 50/50 wt/wt PCL/gelatin solution. Electrospun fiber membranes were obtained by using a commercially available electrospinning setup (Nanon01, MECC, Japan). The solution was dispensed from a 5 ml syringe connected to a 18 Ga needle. Different working parameters were selected in order to get the best fibers morphology. In particular, a voltage of 13 kV, a distance of 8 cm and a flow rate of 0.5 ml h-1 were used to produce randomly dispersed fibers. Longitudinal alignment of PCL/ gelatin fibers was achieved by using a rotating drum with a diameter of 19 cm as collector. The rotation speed has been varied from 1,000 to 3,000 rpm to optimize the degree of anisotropy. The process was carried out at room temperature within a range of relative humidity (RH = 45–50 %).
J Mater Sci: Mater Med
2.3 Morphological analysis The morphology of fibrous membranes was studied by scanning electron microscopy (SEM; QuantaFEG 200, FEI, The Netherlands) under high vacuum conditions (*10-5 mbar). To improve the sample conductivity, specimens were preliminary coated with a Pd–Au nanolayer by using a sputter coater (Emitech K550, Italy). To analyze fiber alignment, SEM images were imported into ImageJ software and transformed using the two-dimensional FFT function. This function converted the image into a spatial distribution corresponding to the changes in pixel intensity across the sample. Transformed images were rotated by 90° to match the alignment of the original images. From FFT images, the distribution of fibers orientation has been calculated by using OrientedJ plugin, in order to quantify the amount of fully aligned fibers in the range between -5° and 5°. Further morphological investigation and surface profile of fibers were determined using Tapping-Mode Atomic Force Microscopy (TM-AFM; Innova System AFM, Bruker Nano Instruments, Santa Barbara, CA, USA). In this case, electrospun fibers have been collected on a glass substrate for 30 s and 2 min for random and aligned fibers, respectively. All AFM procedures were conducted, in air, at room temperature (25 °C) and with an RTESPA silicon cantilever (Bruker Corporation, Santa Barbara, USA) with a rotated tip of radius 8 nm. The resonance frequency of the cantilever (320 kHz) was confirmed by the Thermal Tune Calibration on a glass substrate, and the spring constant was calibrated to be 42.76 N m-1, in general agreement with the nominal values (300–400 kHz and 40–80 N m-1). The average sensitivity of an optical detection system for a cantilever was calculated to be 14 nm V-1. The scan rate was 0.3 Hz while the scan size was 20 9 20 lm2. According to the Brandsch et al. [21] description of tapping conditions, all AFM analysis were mainly performed at moderate tapping (csp = 0.4–0.7) whereas csp is the attenuation ratio of the set point amplitude to the free oscillation, in order to optimize tip/sample interaction. The images of interest were first captured by AFM raster scanning, processed using the NanoScope Analysis data processing software 1.40 (Bruker Corporation, USA) and then graphically reported in 2D and 3D form. 2.4 Cells culture PC-12 cells—derived from pheochromocytoma of the rat adrenal medulla—[kindly donated by the group of Dr. Cerchia Laura (Instituto per l’Endocrinologia e l’Oncologia Sperimentale del CNR Gaetano Salvatore, Naples, Italy] and bone-marrow-derived hMSC line (PT-2501)
obtained from Lonza were selected as the testing cell models. PC-12 cells, after 10 passages and hMSC, after 3–6 ones, were used for all the experimental procedures. PCL/gelatin randomly organized and aligned fibers were used as the experimental group. Prior to the biological assays, samples were sterilized by immersion in 70 % of ethanol (v/v) with antibiotic solution (100 lg/ml streptomycin and 100 U/ml penicillin) for 10 min, washed three times with phosphate-buffered saline (PBS) and air dried. PC-12 cells were routinely cultured in tissue culture flask surfaces with HG-DMEM, containing 100 units ml-1 penicillin, 100 lg ml-1 streptomycin sulfate (Sigma), 5 % fetal bovine serum and 15 % of horse serum. hMSCs were cultured in Eagle’s a-MEM supplemented with 10 % fetal bovine serum, antibiotic solution (streptomycin 100 lg ml-1 and penicillin 100 U ml-1, Sigma Chem. Co) and 2 mM L-glutamin. Both cell lines were incubated at 37 °C in a humidified atmosphere with 95 % air and 5 % CO2. When the cells became almost confluent, they were released by treating with 0.25 % trypsin–EDTA solution for 3 min at 37 °C and resuspended in their respective medium with a final concentration of 2 9 105 cells ml-1. 2.5 In vitro neuronal cell differentiation assay For in vitro differentiation assay, PC-12 were seeded in triplicate onto electrospun fibres mats of PCL/gelatin scaffolds placed onto 24 cell culture plates and cultured with RPMI 1640 medium with 1 % of horse serum and 50 ng ml-1 of neuronal growth factor (NGF) for 7 days. For neuronal induction, hMSCs were pre-treated with aMEM containing 0.0001 % b-mercaptoethanol, IBMX, DMSO, 10 ng ml-1 EGF and 10 ng ml-1 b-FGF (Chemicon, Temecula, CA, USA) previously seeded in triplicate onto electrospun membranes for 24 h. After pre-treatment, hMSCs were washed carefully with PBS and then incubated with 50 ng ml-1 of NGF and 20 ng ml-1 of BDNF for 7 days. After induction, both PC-12 and hMSC cells cultures onto electrospun membranes scaffolds were fixed with 4 % paraformaldehyde and permeabilized with PBS containing 0.5 % Triton X-100. Cells were incubated at 4 °C overnight in a 1:300 dilution of the rabbit (IgG) polyclonal antibody against rat GAP-43 in PBS containing 2 mg ml-1 of bovine serum albumin (BSA). Scaffolds were washed with ice-cold PBS for 10 min at room temperature and incubated for 1 h at 4 °C with goat-anti-rabbit immunoglobulin secondary antibody conjugated with FITC for PC-12 and hMSC (3 mg ml-1, Sigma Chemical, St. Louis, MO), diluted 1:50 in PBS. Control groups were the PC-12 and hMSC cell culture incubate without any presence of growth factors in cell culture media and for immunoassay the lack of the first antibody. The
123
J Mater Sci: Mater Med
experiments were two fold repeated on three PCL/gelatin fiber scaffolds in each experiment. Immunoassaying was visualized by Laser Scanning Confocal Microscopy (Carl Zeiss) (LSCM). On selected images, with higher contrast and best focused, ImageJ Analysis Software has been used in order to estimate neurites length and cell elongation by measuring the aspect ratio (Ar), defined as the ratio between cell width and length of each cell. All results were reported as mean ± standard deviation (SD). The statistical significance (set at P \ 0.05) was determined with Student’s t test by using GraphPad Prism Software.
3 Results and discussion An important component of the overall cell response onto biomaterials is related to cell attachment which may often determine cell fate, including viability, proliferation, and function [22]. In the case of in vitro culture onto bio inspired scaffolds for tissue engineering, different cell behavior may occur. When cells do not attach, they will likely undergo apoptosis and die. Secondly, cells may attach, but not spread by also incurring to a similar fate. Sometimes, adherent cells may also form fibrotic scar tissue and resistance to cell attachment can help to prevent this behavior [23]. However, mechanisms of cell/ scaffold interaction may be partially controlled in vitro by the use of properly designed electrospun scaffolds able to guide the formation of the ECM in response to specific chemical and topographical cues. However, it is extremely hard to distinguish between these effects and identify the relationship between cell attachment and specific chemical and topographical features on the surrounding materials. In previous works, we have demonstrated that gelatin macromolecules, blended with PCL, are able to improve the bioactivity of electrospun fibers [24]. Indeed, the introduction of a natural polymer,—i.e. gelatin to the PCL—drastically alters its hydrophobic behavior, due to a strong hydrophilicity related to the presence in its structure of amine and carboxylic functional groups which are absent in PCL [24]. Hence, gelatin intrinsically enhances the cell affinity by exposing many integrin sites able to favor cell adhesion and differentiation, thus creating a favorable environment which reproduces the chemical cues of natural ECM. Starting from previous studies, we propose the investigation of the in vitro response of nerve cells on PCL/gelatin electrospun fibers with preferential fiber patterning to underline the effect of alignment on the morphology of cell body and cytoskeleton filaments at the early phase of culture. At the preliminary stage, we have optimized the solution and the process parameters such as flow rate, applied
123
voltage and tip–target distance, to deposit on a metal plate a random network of defect-free PCL/gelatin fibers (Fig. 1a). The surface morphology of PCL/gelatin fibers appears smooth and the average diameter results equal to 1.00 ± 0.30 lm. To impart the desired anisotropy degree, a preferential deposition of PCL/gelatin fibers has been obtained by properly setting the rate of the rotating mandrel from 1,000 to 3,000 rpm while keeping the other parameters such as concentration, flow rate, needle gauge, applied voltage and tip–target distance constant. At 1,000 rpm, fibers still appear randomly organized (Fig. 1b) so indicating that this rate is not sufficient to longitudinally orient the fibers. As the rotating rate increased to 2,000 rpm (Fig. 1c), the fibers begin to be progressively aligned, showing different arrangement in diagonal direction as function of the rotating rate. At 3,000 rpm, SEM images show the best alignment of fibers (Fig. 1d), due to the maximum value of stretching forces exerted along the polymer jet during the mechanical winding onto the rotating drum. Therefore, 3,000 rpm is set as the best tested rotation condition to achieve maximum longitudinal orientation. SEM images of random and aligned fiber mats were analyzed using the FFT function to quantify the fiber alignment as a function of the rotation rate. Randomly oriented fibers usually generate an output FFT image where the pixel intensities seem to be uniformly distributed in a circular pattern (Fig. 2a). This characteristic distribution occurs as the frequency related to specific pixel intensities is theoretically the same one in any direction. On the contrary, aligned fibers patterns generate as output a not homogeneous FFT image with a group of pixel intensities oriented in a preferential direction (Fig. 2b, d). Here, a graphical depiction of the FFT frequency distribution has been calculated by using ImageJ Software which provides a radial summation of the pixel intensities encountered along that specific radius in the FFT output images. The sum of the pixel intensity is then plotted between 0° and 180°, taking into account the shape of the FTT spectra is symmetric (Fig. 2). The height and overall shape of the peak offered a quantitative parameter of the fiber alignment in the original data image, thus confirming that the control of the rotation rate allows modulating the structural anisotropy of electrospun scaffolds. A measure of the fibers alignment has been further estimated by the identification of fully aligned fibers characterized by a deviation of the fiber angle below 5° respect to a reference line [25]. Thus, the total frequency of occurrence (OF) of fully aligned fibers with orientation falling down in the range from -5° and 5° has been calculated as a function of the rotation rate (Table 1). In the case of rotation speed of 3,000 rpm, 39 % of fully aligned fibers is estimated while just at 2,000 rpm only 7 % of fibers resulted aligned in the same direction,
J Mater Sci: Mater Med
Fig. 1 SEM micrographs of PCL/gelatin fibers randomly collected on a plate (a) and collected on a drum rotating at 1,000 (b), 2,000 (c) and 3,000 rpm (d)
Fig. 2 FFT images and distribution of orientation plot of a random PCL/gelatin fibers and of PCL/gelatin fibers collected on a drum rotating at 1,000 (b), 2,000 (c) and 3,000 rpm (d)
123
J Mater Sci: Mater Med Table 1 Frequency of occurrence of aligned fiber from -5° to 5° Rotation speed (rpm)
0
1000
2000
3000
OF
5
5
7
39
(%)
indicating that 2,000 rpm as rate is not sufficient to impart an ordered deposition of electrospun fibers. Once set the best electrospinning parameters for fibers alignment, the mean fibers diameter has been calculated on selected SEM images. No significant difference has been found comparing to random PCL/gelatin fibers, i.e. 0.98 ± 0.40 lm. Several studies have demonstrated that the diameter of the fiber scaffold is related to the axonal diameter. As a function of the body site (i.e., vertebrae, peripheral nerve, etc.), there is a large variation in the size of axons, ranging between less than 1 lm and up to 50 lm (i.e., vertebrae, peripheral nerve) [26]. Fibers with sizes comparable to axon diameters are a prerequisite for nerve scaffold design. In this case, fiber diameter is properly on the same scale as the neurites for peripheral nerve applications. The topography of random and aligned PCL/gelatin fibers has been estimated by using TM-AFM, by a tip scanning perpendicular to the fiber axis to assess fiber profile. In the case of random fiber organization, the continuous overlapping of the fibers during the electrospinning process leads to larger z-scale range, i.e. 3.0 lm (Fig. 4a) comparing to the aligned mat, i.e. 0.5 lm (Fig. 4b). As shown in Figs. 3 and 4a, in fact, fibers form a mesh-like network with branches and crossing point among fibers, i.e., kinks, with average diameter falling in the micron range, according to qualitative SEM measurement. In Figs. 3 and 4b, a regular pattern of longitudinally aligned
Fig. 3 TM-AFM images of random (a) and aligned (b) PCL/gelatin fibers by top view. The intensity of the color reflects the height of the fibers and the topography section analysis is also shown
123
fibers is also reported. AFM angle measurements on 12 different aligned fibers show a degree of fiber angle of 4.6° in agreement with image analyses data. A biological validation with PC-12 and hMSC cells was performed in order to qualitatively identify the effect of the different pattern due to fibers spatial disposition on cells behavior. SEM pictures showed the spreading and cell attachment of cells after neuronal induction media for hMSC (Fig. 5, left column) and PC-12 (Fig. 5, right column). After 7 days in culture with NGF-rich medium, PC-12 and hMSC cells showed a good spreading and attachment to the fibers, inducing the formation of characteristic neurites with growth conical ends due to physical contact with electrospun fibers surface. Cells seeded on randomly oriented fibers show a polygonal shape and less organized growth patterns (Fig. 6, up-row), while cells seeded on the aligned electrospun fibers appear more elongated in a bipolar way, growing parallel to aligned fibers (Fig. 6, down-row). In particular, immunostaining images prove the preliminary differentiation of hMSC cells and PC-12 cultured onto random and aligned fibers scaffolds by detection of neurite marker GAP-43. After 7 days in neuronal induction medium, morphological appearance of PC-12 and hMSC confirms the correct advancement of differentiation process: cells sprout out neurites that extend and interact with the substratum with growth cones from them onto fibrous scaffolds. Moreover, neurites from hMSC and PC-12 cells grew along parallel directions to fiber axes, in the case of aligned fibers, thus recognizing the topological guidance dictated by the oriented fiber patterning. To quantify the cell elongation degree, the aspect ratio (Ar) of the cells (width/length) has been measured and
J Mater Sci: Mater Med Fig. 4 AFM 3-D view of random (a) and aligned (b) PCL/gelatin fibers on a glass substrate with a scan size of 20 9 20 lm2
Fig. 5 SEM images of hMSC (left) and PC-12 (right) cell morphology after 7 days culture with NGF addition on random (a, b) and aligned (c, d) fibers
summarized in Fig. 7. In the case of randomly organized fibers, the cytoskeletal morphology of hMSC and PC-12 cells do not show a significant elongation as confirmed by values of (0.53 ± 0.26) and (0.72 ± 0.23) respectively, while the Ar strongly decreases in the case of cells seeded onto aligned electrospun fibers, e.g. (0.10 ± 0.02) for hMSC and (0.23 ± 0.14) for PC-12. Moreover, neurites length also varies for both cells types depending on membranes topography: aligned fibers provide a direct guide for neurite outgrowth while randomly organized fibers could act as barrier to the linear extension of cells protrusion as a consequence of the large presence of kinks. As shown in Fig. 8, in the case of differentiated hMSC cells, neurites length varied from
6.50 ± 1.69 to 28.10 ± 8.24 lm when seeded on random and aligned fibers respectively, while for PC12 cells, length is increased passing from 3.68 ± 2.19 to 24.95 ± 8.44 lm. In summary, all these results univocally indicate that the homogeneous distribution of fiber diameters at micrometric scale promote the creation of microtextured surfaces able to interact more effectively with cells, due to the higher surface volume ratio. The presence of the protein cue also provides specific signals to promote cell growth and functionality. Schnell et al. [27] just reported that a biochemical interaction between cells and collagen exposed on the surface of the nanofibers is mediated by the ECM glycoprotein fibronectin, which binds to collagen I and to
123
J Mater Sci: Mater Med Fig. 6 Confocal images of hMSC (a–c) and PC-12 (b– d) cells morphology after 7 day culture with NGF addition on random (up-row) and aligned (down-row) fibers
Fig. 7 Evaluation of hMSC and PC-12 cells elongation by measuring Ar as the ratio between cell width and length after 7 days in culture. Error bars represent mean ± SD. Asterisk denotes significant differences (P \ 0.05) as determined by Student’s t test
integrin receptors on cell membranes. Indeed, gelatin macromolecules, directly blended inside the fibers, guarantee more favorable conditions in cell adhesion, able to influence the response of cells’ physiological and biological functions, proliferation and differentiation [24]. Meanwhile, fully aligned fibers may indicate a preferential direction to neurite elongation with effects on the rate and the extent of neurite elongation. Previous studies have shown that highly aligned fibers with diameters in the 1–2 lm range have great potential in guiding in vitro neurite outgrowth from DRG explants in preferential ways [28]. However, other ones have also shown that neurite extension may tend to extend perpendicularly to the
123
aligned fibers due to the presence of defects or local misalignment of fibers [29]. Here, we verify that fully aligned fibers obtained by properly set the rotating rates during the electrospinning process allows to minimize topological defects (i.e., kinks) which may interfere to unidirectional outgrowth of neurites. According to previous studies [30] we verify that randomly oriented fibers characterized by the presence of more frequent kinks due to fiber crossing may potentially hinder the natural progress of neurite outgrowth. Wang et al. [31] have just investigated the ability of randomly organized fibers to divert neurite elongation by correlating the fiber density to Schwann cell response. They have demonstrated that the use of aligned
J Mater Sci: Mater Med
References
Fig. 8 Measurements of neurites length from hMSC and PC-12 cells after 7 days of culture. Error bars represent mean ± SD. Asterisk denotes significant differences (P \ 0.05) as determined by Student’s t test
fibers with higher packing degree increases Schwann cell attachment also supporting more neurite extension. Accordingly, we suggest that the fully aligned fibers with low formation of crossing kinks, may also support the attachment of nerve-like cells, further directing the alignment of neurites and avoiding misdirectional axonal growth.
4 Conclusion Fully aligned and kinkless fibers have significant potential to promote nerve cell recognition and to direct neurite outgrowth, being promising for nerve regeneration applications. The accurate control of electrospinning process condition allowed to improve the fiber alignment, minimizing the formation of topological defects which may alter the recognition of the surface patterning. This is extremely necessary to promote nerve regeneration more efficiently. Here, we have verified that the topographic features of fully aligned fibers are able to stimulate hMSC and PC-12 cells to form neurites and their elongation. Furthermore, peculiar topological signals offered by fully fiber alignment corroborated by the effect of bioactive signals—i.e. gelatin endowed with PCL fibers—concur to guide the biological response of nerve cells, thus resulting a really promising system for nerve regeneration. Acknowledgments This work has been financially supported by MERIT n. RBNE08HM7T and FIRB-NEWTON RBAP11BYNP. Scanning Electron Microscopy was supported by the Transmission and Scanning Electron Microscopy Labs (LAMEST) of the National Research Council.
1. Ingber DE. Tensegrity-based mechanosensing from macro to micro. Prog Biophys Mol Biol. 2008;97:163–79. 2. Wozniak MA, Chen CS. Mechanotransduction in development: a growing role for contractility. Nat Rev Mol Cell Biol. 2009;10: 34–43. 3. Ruiz SA, Chen CS. Emergence of patterned stem cell differentiation within multicellular structures. Stem Cells. 2008;26:2921–7. 4. Guaccio A, Guarino V, Alvarez-Perez MA, Cirillo V, Netti PA, Ambrosio L. Influence of electrospun fiber mesh size on hMSC oxygen metabolism in 3D collagen matrices: experimental and theoretical evidences. Biotechnol Bioeng. 2011;108(8):1965–76. 5. Stevens MM, George JH. Exploring and engineering the cell surface interface. Science. 2005;310:1135–8. 6. Li D, Ouyang G, McCann JT, Xia Y. Collecting electrospun nanofibers with patterned electrodes. Nano Lett. 2005;5:913–6. 7. Guarino V, Raucci MG, Ambrosio L. Micro/Nanotexturing and bioactivation strategies to design composite scaffolds and ECMlike analogues. Macromol Symp. 2013;331–332:65–70. 8. Agarwal S, Wendorff JH, Greiner A. Use of electrospinning technique for biomedical applications. Polymer. 2008;49(26): 5603–21. 9. Dayal P, Liu J, Kumar S, Kyu T. Experimental and theoretical investigations of porous structure formation in electrospun fibers. Macromolecules. 2007;40(21):7689–94. 10. Yang F, Murugan R, Wang S, Ramakrishna S. Electrospinning of nano/micro scale poly(L-lactic acid) aligned fibers and their potential in neural tissue engineering. Biomaterials. 2005;26:2603–10. 11. Chew SY, Mi R, Hoke A, Leong KW. The effect of the alignment of electrospun fibrous scaffolds on Schwann cell maturation. Biomaterials. 2008;29:653–61. 12. Ma Z, He W, Yong T, Ramakrishna S. Grafting of gelatin on electrospun poly(caprolactone) nanofibers to improve endothelial cell spreading and proliferation and to control cell orientation. Tissue Eng. 2005;11:1149–58. 13. Riboldi SA, Sadr N, Pigini L, Neuenschwander P, Simonet M, Mognol P, Sampaolesi M, Cossu G, Mantero S. Skeletal myogenesis on highly orientated microfibrous polyesterurethane scaffolds. J Biomed Mater Res A. 2008;84:1094–101. 14. Rockwood DN, Akins RE Jr, Parrag IC, Woodhouse KA. Culture on electrospun polyurethane scaffolds decreases atrial natriuretic peptide expression by cardiomyocytes in vitro. Biomaterials. 2008;29:4783–91. 15. Arras MLM, Grasl C, Bergmeister H, Schima H. Electrospinning of aligned fibers with adjustable orientation using auxiliary electrodes. Sci Technol Adv Mater. 2012;13:035008 (8 pp). 16. Prabhakaran MP, Vatankhah E, Ramakrishna S. Electrospun aligned PHBV/collagen nanofibers as substrates for nerve tissue engineering. Biotechnol Bioeng. 2013;110(10):2775–84. 17. Alvarez-Perez MA, Guarino V, Cirillo V, Ambrosio L. Influence of gelatin cues in PCL electrospun membranes on nerve outgrowth. Biomacromolecules. 2010;11:2238–46. 18. Ghasemi-Mobarakeh L, Prabhakaran MP, Morshed M, Nasr-Esfahani MH, Ramakrishna S. Electrospun poly(3-caprolactone)/ gelatin nanofibrous scaffolds for nerve tissue engineering. Biomaterials. 2008;29:4532–9. 19. Kim CH, Khil MS, Kim HY, Lee HU, Jahng KY. An improved hydrophilicity via electrospinning for enhanced cell attachment and proliferation. J Biomed Mater Res B. 2006;78B:283–90. 20. Li WJ, Cooper JA Jr, Mauck RL, Tuan RS. Fabrication and characterization of six electrospun poly(a-hydroxyester)-based nanofibrous scaffolds for tissue engineering applications. Acta Biomater. 2006;2:377–85.
123
J Mater Sci: Mater Med 21. Brandsch R, Bar G, Whangbo MH. On the factors affecting the contrast of height and phase images in tapping mode atomic force microscopy. Langmuir. 1997;13:6349–53. 22. Han N, Johnson JK, Bradley PA, Parikh KS, Lannutti JJ, Winter JO. Cell attachment to hydrogel–electrospun fiber mat composite materials. J Funct Biomater. 2012;3:497–513. 23. Anderson JM, Rodriguez A, Chang DT. Foreign body reaction to biomaterials. Semin Immunol. 2008;20:86–100. 24. Guarino V, Alvarez-Perez MA, Cirillo V, Ambrosio L. hMSC interaction with PCL and PCL/gelatin platforms: a comparative study on films and electrospun membranes. J Bioact Compat Polym. 2011;26(2):144–60. 25. Schneider T, Kohl B, Sauter T, Kratz K, Lendlein A, Ertel W, Schulze-Tanzil G. Influence of fiber orientation in electrospun polymer scaffolds on viability, adhesion and differentiation of articular chondrocytes. Clin Hemorheol Microcirc. 2012;52:325–36. 26. Rydmark M. Nodal axon diameter correlates linearly with intermodal axon diameter in spinal roots of the cat. Neurosci Lett. 1981;24:247–50.
123
27. Schnell E, Klinkhammer K, Balzer S, Brook G, Kleeb D, Dalton P, Mey J. Guidance of glial cell migration and axonal growth on electrospun nanofibers of poly-e-caprolactone and a collagen/ poly-e-caprolactone blend. Biomaterial. 2007;28:3012–25. 28. Wang HB, Mullins ME, Cregg JM, McCarthy CW, Gilbert RJ. Varying the diameter of aligned electrospun fibers alters neurite outgrowth and Schwann cell migration. Acta Biomater. 2010;6: 2970–8. 29. Xie J, MacEwan MR, Li X, Sakiyama-Elbert SE, Xia Y. Neurite outgrowth on nanofiber scaffolds with different orders, structures, and surface properties. ACS Nano. 2009;3:1151–9. 30. Corey JM, Lin DY, Mycek KB, Chen Q, Samuel S, Feldman EL, Martin DC. Aligned electrospun nanofibers specify the direction of dorsal root ganglia neurite growth. J Biomed Mater Res A. 2007;83:636–45. 31. Wang HB, Mullins ME, Cregg JM, Hurtado A, Oudega M, Trombley MT, Gilbert RJ. Creation of highly aligned electrospun poly-L-lactic acid fibers for nerve regeneration. J Neural Eng. 2009;6:016001.
Optimization of fully aligned bioactive electrospun fibers for ‘‘in vitro’’ nerve guidance Valentina Cirillo • Vincenzo Guarino • Marco Antonio Alvarez-Perez • Marica Marrese Luigi Ambrosio
•
Received: 17 December 2013 / Accepted: 28 March 2014 Ó Springer Science+Business Media New York 2014
Abstract Complex architecture of natural tissues such as nerves requires the use of multifunctional scaffolds with peculiar topological and biochemical signals able to address cell behavior towards specific events at the cellular (microscale) and macromolecular (nanoscale) level. In this context, the electrospinning technique is useful to generate fiber assemblies having peculiar fiber diameters at the nanoscale and patterned by unidirectional ways, to facilitate neurite extension via contact guidance. Following a bio-mimetic approach, fully aligned polycaprolactone fibers blended with gelatin macromolecules have been fabricated as potential bioactive substrate for nerve regeneration. Morphological and topographic aspects of electrospun fibers assessed by SEM/AFM microscopy supported by image analyses elaboration allow estimating an increase of fully aligned fibers from 5 to 39 % as collector rotating rate increases from 1,000 to 3,000 rpm. We verify that fully alignment of fibers positively influences in vitro response of hMSC and PC-12 cells in neurogenic way. Immunostaining images show that the presence of topological defects, i.e., kinks—due to more frequent fiber crossing—in the case of randomly organized fiber assembly concurs to interfere with proper neurite outgrowth. On the contrary, fully aligned fibers without kinks offer a more efficient contact guidance to direct the orientation of nerve cells along the fibers respect to randomly organized ones, V. Cirillo V. Guarino (&) M. Marrese L. Ambrosio Institute of Polymers, Composites and Biomaterials, National Research Council of Italy, V.le Kennedy 54, Pad 20, Mostra d’Oltremare, 80125 Naples, Italy e-mail: [email protected] M. A. Alvarez-Perez TBL-DEPeI, Universidad Nacional Auto´noma de Me´xico (UNAM), Coyoaca´n, Mexico, DF, Mexico
promoting a high elongation of neurites at 7 days and the formation of bipolar extensions. So, this confirms that the topological cue of fully alignment of fibers elicits a favorable environment for nerve regeneration.
1 Introduction It is general understood that the complex architecture of scaffolds, as well as the characteristics of the surrounding environment, may affect cell fate and functions [1]. Recent experimental evidences have suggested that physical control of cell shape mediated by peculiar forces on the surrounding microstructure may act as a potent regulator of stem cell fate [2]. For instance, protein based micropatterns in various geometries (circle, square, rectangle and ellipses) recently confirmed how surface concavity may influence the human mesenchymal stem cells (hMSC) behavior, inducing greater attractive forces than those on convex surfaces [3], so directly forcing cells into osteogenic instead of adipogenic lineages. Other studies have also indicated that the characteristic size scale of structural units may also affect the mechanisms of cell metabolism, offering new insights into the role of textured materials in specific cell activities [4]. However, not only the scale but also the topographic organization in rigid aligned patterns can also relevantly contribute to the final cell behavior [5]. In order to study the ability of cells to recognize micro and/ or sub-microtopography, it is mandatory to design synthetic platforms with peculiar structural order and anisotropy. Currently, electrospinning is emerging as a powerful technique for the production of micro and nanotextured fibrous scaffolds with specific fiber patterning [6]. Traditionally, electrospun nanofibers have been usually collected
123
J Mater Sci: Mater Med
as nonwoven mats with randomly oriented fibers assembled to form interconnected porous networks that mimic the collagenous fibrous architecture of the extracellular matrix (ECM) which is naturally present in hard and soft tissues. These peculiar fiber features may be just satisfactory to offer multiple possibilities in surface functionalization and biomimicking [7]. In particular, high surface area of the nanofibers helps to overcome loading limitations that are normally encountered in current drug delivery methods [8], despite this effect may generally corroborated by the addition of surface pores able to promote the formation of larger number of binding sites for drug loading [9]. However, main features of most native ECMs analogues of natural tissues or organs (e.g. heart, tendons, blood vessels) have to reproduce the exact anisotropic architecture of tissue which is crucial to determine, case by case, the specific function of tissues. In this context, a fine control of the spatial orientation and fiber patterning by an accurate design of electrospinning process conditions may assure a more efficient mimesis of the ECM in terms of structural and functional properties thanks to the possibility to build different patterns which reply the structural order of several tissues including nerve and muscles. Recent studies have just demonstrated that the aligned electrospun nanofiber arrays may promote significant improvements in cell proliferation respect to random fiber configuration in the case of neural stem cells [10], primary Schwann cells [11] and endothelial cells [12], also affecting the cell-growing direction and morphologies [13, 14]. In the specific case of nerve regeneration, aligned fibers may provide signaling cues to the cells and act as contact guidance, i.e. cells on the scaffold are promoted to collectively align along the desired direction [15]. So, axons may be guided to their targets by the cues directly provided by the substrate, which could be either topographical or chemical and biological cues or their combinations [16, 17]. Hence, continuous straight and aligned fibers with tailored material compositions are suitable as artificial substrate for neurite alignment and outgrowth. Here, we propose to study the optimization of electrospinning process conditions to obtain random and aligned electrospun fibers by mixing two different polymers, i.e. poly-e-caprolactone (PCL) and gelatin for peripheral nerve regeneration. PCL is a semi crystalline linear polyester largely used as biomaterial. However, its poor hydrophilicity reduces cell adhesion, migration, proliferation, and differentiation [18–20]. So, the inclusion of a hydrophilic protein derived from collagen denaturation such as gelatin, may improve cell recognition in order to obtain fiber scaffold with more controlled cell response and degradation properties. In a previous study, we have just demonstrated that the gelatin integrated into PCL fibers
123
may act as biochemical signal improving significantly the biointeraction of PC-12 pheochromocytoma nerve cells with the substrate comparing to PCL fibers as shown by results of cell attachment, viability, and neurite outgrowth [17]. Now, we propose to investigate PCL/gelatin fibers with preferential fiber patterning to underline the effect of alignment in neurogenic way by the biological interaction of two types of cells with different differentiation state, i.e., undifferentiated hMSC and rat PC-12, offering new information about the role of fiber patterning on cell behavior.
2 Materials and methods 2.1 Materials PCL (Mn 45 kDa) and gelatin (Type B from calf skin, *225 Bloom) were purchased from Sigma-Aldrich (Italy) and 1,1,1,3,3,3-hexafluoro-2-propanol (HFP) from Fluka (Italy) was used as received without further purification. High glucose-Dulbecco’s modified Eagle’s medium (HG-DMEM); Eagle’s alpha minimum essential medium (a-MEM), trypsin–EDTA solution, L-glutamin, fetal bovine serum, horse serum were purchased from Gibco. Antibiotic solution, nerve growth factor 2.5S from murine submaxillary gland (NGF), brain derived neurotrophic factor (BDNF), dimethyl sulfoxide (DMSO), b-mercaptoethanol, triton X-100 were supplied by Sigma. Epidermal growth factor (EGF) and basic fibroblast growth factor (bFGF) were purchase from Chemicon. Anti-Growth associated protein-43 (GAP-43) antibody was purchased from Millipore. 2.2 Electrospinning process PCL and gelatin were separately dissolved in HFP at a concentration of 0.1 g ml-1 and kept under magnetic stirring at room temperature for 24 h. Then, polymer solutions were mixed to obtain a 50/50 wt/wt PCL/gelatin solution. Electrospun fiber membranes were obtained by using a commercially available electrospinning setup (Nanon01, MECC, Japan). The solution was dispensed from a 5 ml syringe connected to a 18 Ga needle. Different working parameters were selected in order to get the best fibers morphology. In particular, a voltage of 13 kV, a distance of 8 cm and a flow rate of 0.5 ml h-1 were used to produce randomly dispersed fibers. Longitudinal alignment of PCL/ gelatin fibers was achieved by using a rotating drum with a diameter of 19 cm as collector. The rotation speed has been varied from 1,000 to 3,000 rpm to optimize the degree of anisotropy. The process was carried out at room temperature within a range of relative humidity (RH = 45–50 %).
J Mater Sci: Mater Med
2.3 Morphological analysis The morphology of fibrous membranes was studied by scanning electron microscopy (SEM; QuantaFEG 200, FEI, The Netherlands) under high vacuum conditions (*10-5 mbar). To improve the sample conductivity, specimens were preliminary coated with a Pd–Au nanolayer by using a sputter coater (Emitech K550, Italy). To analyze fiber alignment, SEM images were imported into ImageJ software and transformed using the two-dimensional FFT function. This function converted the image into a spatial distribution corresponding to the changes in pixel intensity across the sample. Transformed images were rotated by 90° to match the alignment of the original images. From FFT images, the distribution of fibers orientation has been calculated by using OrientedJ plugin, in order to quantify the amount of fully aligned fibers in the range between -5° and 5°. Further morphological investigation and surface profile of fibers were determined using Tapping-Mode Atomic Force Microscopy (TM-AFM; Innova System AFM, Bruker Nano Instruments, Santa Barbara, CA, USA). In this case, electrospun fibers have been collected on a glass substrate for 30 s and 2 min for random and aligned fibers, respectively. All AFM procedures were conducted, in air, at room temperature (25 °C) and with an RTESPA silicon cantilever (Bruker Corporation, Santa Barbara, USA) with a rotated tip of radius 8 nm. The resonance frequency of the cantilever (320 kHz) was confirmed by the Thermal Tune Calibration on a glass substrate, and the spring constant was calibrated to be 42.76 N m-1, in general agreement with the nominal values (300–400 kHz and 40–80 N m-1). The average sensitivity of an optical detection system for a cantilever was calculated to be 14 nm V-1. The scan rate was 0.3 Hz while the scan size was 20 9 20 lm2. According to the Brandsch et al. [21] description of tapping conditions, all AFM analysis were mainly performed at moderate tapping (csp = 0.4–0.7) whereas csp is the attenuation ratio of the set point amplitude to the free oscillation, in order to optimize tip/sample interaction. The images of interest were first captured by AFM raster scanning, processed using the NanoScope Analysis data processing software 1.40 (Bruker Corporation, USA) and then graphically reported in 2D and 3D form. 2.4 Cells culture PC-12 cells—derived from pheochromocytoma of the rat adrenal medulla—[kindly donated by the group of Dr. Cerchia Laura (Instituto per l’Endocrinologia e l’Oncologia Sperimentale del CNR Gaetano Salvatore, Naples, Italy] and bone-marrow-derived hMSC line (PT-2501)
obtained from Lonza were selected as the testing cell models. PC-12 cells, after 10 passages and hMSC, after 3–6 ones, were used for all the experimental procedures. PCL/gelatin randomly organized and aligned fibers were used as the experimental group. Prior to the biological assays, samples were sterilized by immersion in 70 % of ethanol (v/v) with antibiotic solution (100 lg/ml streptomycin and 100 U/ml penicillin) for 10 min, washed three times with phosphate-buffered saline (PBS) and air dried. PC-12 cells were routinely cultured in tissue culture flask surfaces with HG-DMEM, containing 100 units ml-1 penicillin, 100 lg ml-1 streptomycin sulfate (Sigma), 5 % fetal bovine serum and 15 % of horse serum. hMSCs were cultured in Eagle’s a-MEM supplemented with 10 % fetal bovine serum, antibiotic solution (streptomycin 100 lg ml-1 and penicillin 100 U ml-1, Sigma Chem. Co) and 2 mM L-glutamin. Both cell lines were incubated at 37 °C in a humidified atmosphere with 95 % air and 5 % CO2. When the cells became almost confluent, they were released by treating with 0.25 % trypsin–EDTA solution for 3 min at 37 °C and resuspended in their respective medium with a final concentration of 2 9 105 cells ml-1. 2.5 In vitro neuronal cell differentiation assay For in vitro differentiation assay, PC-12 were seeded in triplicate onto electrospun fibres mats of PCL/gelatin scaffolds placed onto 24 cell culture plates and cultured with RPMI 1640 medium with 1 % of horse serum and 50 ng ml-1 of neuronal growth factor (NGF) for 7 days. For neuronal induction, hMSCs were pre-treated with aMEM containing 0.0001 % b-mercaptoethanol, IBMX, DMSO, 10 ng ml-1 EGF and 10 ng ml-1 b-FGF (Chemicon, Temecula, CA, USA) previously seeded in triplicate onto electrospun membranes for 24 h. After pre-treatment, hMSCs were washed carefully with PBS and then incubated with 50 ng ml-1 of NGF and 20 ng ml-1 of BDNF for 7 days. After induction, both PC-12 and hMSC cells cultures onto electrospun membranes scaffolds were fixed with 4 % paraformaldehyde and permeabilized with PBS containing 0.5 % Triton X-100. Cells were incubated at 4 °C overnight in a 1:300 dilution of the rabbit (IgG) polyclonal antibody against rat GAP-43 in PBS containing 2 mg ml-1 of bovine serum albumin (BSA). Scaffolds were washed with ice-cold PBS for 10 min at room temperature and incubated for 1 h at 4 °C with goat-anti-rabbit immunoglobulin secondary antibody conjugated with FITC for PC-12 and hMSC (3 mg ml-1, Sigma Chemical, St. Louis, MO), diluted 1:50 in PBS. Control groups were the PC-12 and hMSC cell culture incubate without any presence of growth factors in cell culture media and for immunoassay the lack of the first antibody. The
123
J Mater Sci: Mater Med
experiments were two fold repeated on three PCL/gelatin fiber scaffolds in each experiment. Immunoassaying was visualized by Laser Scanning Confocal Microscopy (Carl Zeiss) (LSCM). On selected images, with higher contrast and best focused, ImageJ Analysis Software has been used in order to estimate neurites length and cell elongation by measuring the aspect ratio (Ar), defined as the ratio between cell width and length of each cell. All results were reported as mean ± standard deviation (SD). The statistical significance (set at P \ 0.05) was determined with Student’s t test by using GraphPad Prism Software.
3 Results and discussion An important component of the overall cell response onto biomaterials is related to cell attachment which may often determine cell fate, including viability, proliferation, and function [22]. In the case of in vitro culture onto bio inspired scaffolds for tissue engineering, different cell behavior may occur. When cells do not attach, they will likely undergo apoptosis and die. Secondly, cells may attach, but not spread by also incurring to a similar fate. Sometimes, adherent cells may also form fibrotic scar tissue and resistance to cell attachment can help to prevent this behavior [23]. However, mechanisms of cell/ scaffold interaction may be partially controlled in vitro by the use of properly designed electrospun scaffolds able to guide the formation of the ECM in response to specific chemical and topographical cues. However, it is extremely hard to distinguish between these effects and identify the relationship between cell attachment and specific chemical and topographical features on the surrounding materials. In previous works, we have demonstrated that gelatin macromolecules, blended with PCL, are able to improve the bioactivity of electrospun fibers [24]. Indeed, the introduction of a natural polymer,—i.e. gelatin to the PCL—drastically alters its hydrophobic behavior, due to a strong hydrophilicity related to the presence in its structure of amine and carboxylic functional groups which are absent in PCL [24]. Hence, gelatin intrinsically enhances the cell affinity by exposing many integrin sites able to favor cell adhesion and differentiation, thus creating a favorable environment which reproduces the chemical cues of natural ECM. Starting from previous studies, we propose the investigation of the in vitro response of nerve cells on PCL/gelatin electrospun fibers with preferential fiber patterning to underline the effect of alignment on the morphology of cell body and cytoskeleton filaments at the early phase of culture. At the preliminary stage, we have optimized the solution and the process parameters such as flow rate, applied
123
voltage and tip–target distance, to deposit on a metal plate a random network of defect-free PCL/gelatin fibers (Fig. 1a). The surface morphology of PCL/gelatin fibers appears smooth and the average diameter results equal to 1.00 ± 0.30 lm. To impart the desired anisotropy degree, a preferential deposition of PCL/gelatin fibers has been obtained by properly setting the rate of the rotating mandrel from 1,000 to 3,000 rpm while keeping the other parameters such as concentration, flow rate, needle gauge, applied voltage and tip–target distance constant. At 1,000 rpm, fibers still appear randomly organized (Fig. 1b) so indicating that this rate is not sufficient to longitudinally orient the fibers. As the rotating rate increased to 2,000 rpm (Fig. 1c), the fibers begin to be progressively aligned, showing different arrangement in diagonal direction as function of the rotating rate. At 3,000 rpm, SEM images show the best alignment of fibers (Fig. 1d), due to the maximum value of stretching forces exerted along the polymer jet during the mechanical winding onto the rotating drum. Therefore, 3,000 rpm is set as the best tested rotation condition to achieve maximum longitudinal orientation. SEM images of random and aligned fiber mats were analyzed using the FFT function to quantify the fiber alignment as a function of the rotation rate. Randomly oriented fibers usually generate an output FFT image where the pixel intensities seem to be uniformly distributed in a circular pattern (Fig. 2a). This characteristic distribution occurs as the frequency related to specific pixel intensities is theoretically the same one in any direction. On the contrary, aligned fibers patterns generate as output a not homogeneous FFT image with a group of pixel intensities oriented in a preferential direction (Fig. 2b, d). Here, a graphical depiction of the FFT frequency distribution has been calculated by using ImageJ Software which provides a radial summation of the pixel intensities encountered along that specific radius in the FFT output images. The sum of the pixel intensity is then plotted between 0° and 180°, taking into account the shape of the FTT spectra is symmetric (Fig. 2). The height and overall shape of the peak offered a quantitative parameter of the fiber alignment in the original data image, thus confirming that the control of the rotation rate allows modulating the structural anisotropy of electrospun scaffolds. A measure of the fibers alignment has been further estimated by the identification of fully aligned fibers characterized by a deviation of the fiber angle below 5° respect to a reference line [25]. Thus, the total frequency of occurrence (OF) of fully aligned fibers with orientation falling down in the range from -5° and 5° has been calculated as a function of the rotation rate (Table 1). In the case of rotation speed of 3,000 rpm, 39 % of fully aligned fibers is estimated while just at 2,000 rpm only 7 % of fibers resulted aligned in the same direction,
J Mater Sci: Mater Med
Fig. 1 SEM micrographs of PCL/gelatin fibers randomly collected on a plate (a) and collected on a drum rotating at 1,000 (b), 2,000 (c) and 3,000 rpm (d)
Fig. 2 FFT images and distribution of orientation plot of a random PCL/gelatin fibers and of PCL/gelatin fibers collected on a drum rotating at 1,000 (b), 2,000 (c) and 3,000 rpm (d)
123
J Mater Sci: Mater Med Table 1 Frequency of occurrence of aligned fiber from -5° to 5° Rotation speed (rpm)
0
1000
2000
3000
OF
5
5
7
39
(%)
indicating that 2,000 rpm as rate is not sufficient to impart an ordered deposition of electrospun fibers. Once set the best electrospinning parameters for fibers alignment, the mean fibers diameter has been calculated on selected SEM images. No significant difference has been found comparing to random PCL/gelatin fibers, i.e. 0.98 ± 0.40 lm. Several studies have demonstrated that the diameter of the fiber scaffold is related to the axonal diameter. As a function of the body site (i.e., vertebrae, peripheral nerve, etc.), there is a large variation in the size of axons, ranging between less than 1 lm and up to 50 lm (i.e., vertebrae, peripheral nerve) [26]. Fibers with sizes comparable to axon diameters are a prerequisite for nerve scaffold design. In this case, fiber diameter is properly on the same scale as the neurites for peripheral nerve applications. The topography of random and aligned PCL/gelatin fibers has been estimated by using TM-AFM, by a tip scanning perpendicular to the fiber axis to assess fiber profile. In the case of random fiber organization, the continuous overlapping of the fibers during the electrospinning process leads to larger z-scale range, i.e. 3.0 lm (Fig. 4a) comparing to the aligned mat, i.e. 0.5 lm (Fig. 4b). As shown in Figs. 3 and 4a, in fact, fibers form a mesh-like network with branches and crossing point among fibers, i.e., kinks, with average diameter falling in the micron range, according to qualitative SEM measurement. In Figs. 3 and 4b, a regular pattern of longitudinally aligned
Fig. 3 TM-AFM images of random (a) and aligned (b) PCL/gelatin fibers by top view. The intensity of the color reflects the height of the fibers and the topography section analysis is also shown
123
fibers is also reported. AFM angle measurements on 12 different aligned fibers show a degree of fiber angle of 4.6° in agreement with image analyses data. A biological validation with PC-12 and hMSC cells was performed in order to qualitatively identify the effect of the different pattern due to fibers spatial disposition on cells behavior. SEM pictures showed the spreading and cell attachment of cells after neuronal induction media for hMSC (Fig. 5, left column) and PC-12 (Fig. 5, right column). After 7 days in culture with NGF-rich medium, PC-12 and hMSC cells showed a good spreading and attachment to the fibers, inducing the formation of characteristic neurites with growth conical ends due to physical contact with electrospun fibers surface. Cells seeded on randomly oriented fibers show a polygonal shape and less organized growth patterns (Fig. 6, up-row), while cells seeded on the aligned electrospun fibers appear more elongated in a bipolar way, growing parallel to aligned fibers (Fig. 6, down-row). In particular, immunostaining images prove the preliminary differentiation of hMSC cells and PC-12 cultured onto random and aligned fibers scaffolds by detection of neurite marker GAP-43. After 7 days in neuronal induction medium, morphological appearance of PC-12 and hMSC confirms the correct advancement of differentiation process: cells sprout out neurites that extend and interact with the substratum with growth cones from them onto fibrous scaffolds. Moreover, neurites from hMSC and PC-12 cells grew along parallel directions to fiber axes, in the case of aligned fibers, thus recognizing the topological guidance dictated by the oriented fiber patterning. To quantify the cell elongation degree, the aspect ratio (Ar) of the cells (width/length) has been measured and
J Mater Sci: Mater Med Fig. 4 AFM 3-D view of random (a) and aligned (b) PCL/gelatin fibers on a glass substrate with a scan size of 20 9 20 lm2
Fig. 5 SEM images of hMSC (left) and PC-12 (right) cell morphology after 7 days culture with NGF addition on random (a, b) and aligned (c, d) fibers
summarized in Fig. 7. In the case of randomly organized fibers, the cytoskeletal morphology of hMSC and PC-12 cells do not show a significant elongation as confirmed by values of (0.53 ± 0.26) and (0.72 ± 0.23) respectively, while the Ar strongly decreases in the case of cells seeded onto aligned electrospun fibers, e.g. (0.10 ± 0.02) for hMSC and (0.23 ± 0.14) for PC-12. Moreover, neurites length also varies for both cells types depending on membranes topography: aligned fibers provide a direct guide for neurite outgrowth while randomly organized fibers could act as barrier to the linear extension of cells protrusion as a consequence of the large presence of kinks. As shown in Fig. 8, in the case of differentiated hMSC cells, neurites length varied from
6.50 ± 1.69 to 28.10 ± 8.24 lm when seeded on random and aligned fibers respectively, while for PC12 cells, length is increased passing from 3.68 ± 2.19 to 24.95 ± 8.44 lm. In summary, all these results univocally indicate that the homogeneous distribution of fiber diameters at micrometric scale promote the creation of microtextured surfaces able to interact more effectively with cells, due to the higher surface volume ratio. The presence of the protein cue also provides specific signals to promote cell growth and functionality. Schnell et al. [27] just reported that a biochemical interaction between cells and collagen exposed on the surface of the nanofibers is mediated by the ECM glycoprotein fibronectin, which binds to collagen I and to
123
J Mater Sci: Mater Med Fig. 6 Confocal images of hMSC (a–c) and PC-12 (b– d) cells morphology after 7 day culture with NGF addition on random (up-row) and aligned (down-row) fibers
Fig. 7 Evaluation of hMSC and PC-12 cells elongation by measuring Ar as the ratio between cell width and length after 7 days in culture. Error bars represent mean ± SD. Asterisk denotes significant differences (P \ 0.05) as determined by Student’s t test
integrin receptors on cell membranes. Indeed, gelatin macromolecules, directly blended inside the fibers, guarantee more favorable conditions in cell adhesion, able to influence the response of cells’ physiological and biological functions, proliferation and differentiation [24]. Meanwhile, fully aligned fibers may indicate a preferential direction to neurite elongation with effects on the rate and the extent of neurite elongation. Previous studies have shown that highly aligned fibers with diameters in the 1–2 lm range have great potential in guiding in vitro neurite outgrowth from DRG explants in preferential ways [28]. However, other ones have also shown that neurite extension may tend to extend perpendicularly to the
123
aligned fibers due to the presence of defects or local misalignment of fibers [29]. Here, we verify that fully aligned fibers obtained by properly set the rotating rates during the electrospinning process allows to minimize topological defects (i.e., kinks) which may interfere to unidirectional outgrowth of neurites. According to previous studies [30] we verify that randomly oriented fibers characterized by the presence of more frequent kinks due to fiber crossing may potentially hinder the natural progress of neurite outgrowth. Wang et al. [31] have just investigated the ability of randomly organized fibers to divert neurite elongation by correlating the fiber density to Schwann cell response. They have demonstrated that the use of aligned
J Mater Sci: Mater Med
References
Fig. 8 Measurements of neurites length from hMSC and PC-12 cells after 7 days of culture. Error bars represent mean ± SD. Asterisk denotes significant differences (P \ 0.05) as determined by Student’s t test
fibers with higher packing degree increases Schwann cell attachment also supporting more neurite extension. Accordingly, we suggest that the fully aligned fibers with low formation of crossing kinks, may also support the attachment of nerve-like cells, further directing the alignment of neurites and avoiding misdirectional axonal growth.
4 Conclusion Fully aligned and kinkless fibers have significant potential to promote nerve cell recognition and to direct neurite outgrowth, being promising for nerve regeneration applications. The accurate control of electrospinning process condition allowed to improve the fiber alignment, minimizing the formation of topological defects which may alter the recognition of the surface patterning. This is extremely necessary to promote nerve regeneration more efficiently. Here, we have verified that the topographic features of fully aligned fibers are able to stimulate hMSC and PC-12 cells to form neurites and their elongation. Furthermore, peculiar topological signals offered by fully fiber alignment corroborated by the effect of bioactive signals—i.e. gelatin endowed with PCL fibers—concur to guide the biological response of nerve cells, thus resulting a really promising system for nerve regeneration. Acknowledgments This work has been financially supported by MERIT n. RBNE08HM7T and FIRB-NEWTON RBAP11BYNP. Scanning Electron Microscopy was supported by the Transmission and Scanning Electron Microscopy Labs (LAMEST) of the National Research Council.
1. Ingber DE. Tensegrity-based mechanosensing from macro to micro. Prog Biophys Mol Biol. 2008;97:163–79. 2. Wozniak MA, Chen CS. Mechanotransduction in development: a growing role for contractility. Nat Rev Mol Cell Biol. 2009;10: 34–43. 3. Ruiz SA, Chen CS. Emergence of patterned stem cell differentiation within multicellular structures. Stem Cells. 2008;26:2921–7. 4. Guaccio A, Guarino V, Alvarez-Perez MA, Cirillo V, Netti PA, Ambrosio L. Influence of electrospun fiber mesh size on hMSC oxygen metabolism in 3D collagen matrices: experimental and theoretical evidences. Biotechnol Bioeng. 2011;108(8):1965–76. 5. Stevens MM, George JH. Exploring and engineering the cell surface interface. Science. 2005;310:1135–8. 6. Li D, Ouyang G, McCann JT, Xia Y. Collecting electrospun nanofibers with patterned electrodes. Nano Lett. 2005;5:913–6. 7. Guarino V, Raucci MG, Ambrosio L. Micro/Nanotexturing and bioactivation strategies to design composite scaffolds and ECMlike analogues. Macromol Symp. 2013;331–332:65–70. 8. Agarwal S, Wendorff JH, Greiner A. Use of electrospinning technique for biomedical applications. Polymer. 2008;49(26): 5603–21. 9. Dayal P, Liu J, Kumar S, Kyu T. Experimental and theoretical investigations of porous structure formation in electrospun fibers. Macromolecules. 2007;40(21):7689–94. 10. Yang F, Murugan R, Wang S, Ramakrishna S. Electrospinning of nano/micro scale poly(L-lactic acid) aligned fibers and their potential in neural tissue engineering. Biomaterials. 2005;26:2603–10. 11. Chew SY, Mi R, Hoke A, Leong KW. The effect of the alignment of electrospun fibrous scaffolds on Schwann cell maturation. Biomaterials. 2008;29:653–61. 12. Ma Z, He W, Yong T, Ramakrishna S. Grafting of gelatin on electrospun poly(caprolactone) nanofibers to improve endothelial cell spreading and proliferation and to control cell orientation. Tissue Eng. 2005;11:1149–58. 13. Riboldi SA, Sadr N, Pigini L, Neuenschwander P, Simonet M, Mognol P, Sampaolesi M, Cossu G, Mantero S. Skeletal myogenesis on highly orientated microfibrous polyesterurethane scaffolds. J Biomed Mater Res A. 2008;84:1094–101. 14. Rockwood DN, Akins RE Jr, Parrag IC, Woodhouse KA. Culture on electrospun polyurethane scaffolds decreases atrial natriuretic peptide expression by cardiomyocytes in vitro. Biomaterials. 2008;29:4783–91. 15. Arras MLM, Grasl C, Bergmeister H, Schima H. Electrospinning of aligned fibers with adjustable orientation using auxiliary electrodes. Sci Technol Adv Mater. 2012;13:035008 (8 pp). 16. Prabhakaran MP, Vatankhah E, Ramakrishna S. Electrospun aligned PHBV/collagen nanofibers as substrates for nerve tissue engineering. Biotechnol Bioeng. 2013;110(10):2775–84. 17. Alvarez-Perez MA, Guarino V, Cirillo V, Ambrosio L. Influence of gelatin cues in PCL electrospun membranes on nerve outgrowth. Biomacromolecules. 2010;11:2238–46. 18. Ghasemi-Mobarakeh L, Prabhakaran MP, Morshed M, Nasr-Esfahani MH, Ramakrishna S. Electrospun poly(3-caprolactone)/ gelatin nanofibrous scaffolds for nerve tissue engineering. Biomaterials. 2008;29:4532–9. 19. Kim CH, Khil MS, Kim HY, Lee HU, Jahng KY. An improved hydrophilicity via electrospinning for enhanced cell attachment and proliferation. J Biomed Mater Res B. 2006;78B:283–90. 20. Li WJ, Cooper JA Jr, Mauck RL, Tuan RS. Fabrication and characterization of six electrospun poly(a-hydroxyester)-based nanofibrous scaffolds for tissue engineering applications. Acta Biomater. 2006;2:377–85.
123
J Mater Sci: Mater Med 21. Brandsch R, Bar G, Whangbo MH. On the factors affecting the contrast of height and phase images in tapping mode atomic force microscopy. Langmuir. 1997;13:6349–53. 22. Han N, Johnson JK, Bradley PA, Parikh KS, Lannutti JJ, Winter JO. Cell attachment to hydrogel–electrospun fiber mat composite materials. J Funct Biomater. 2012;3:497–513. 23. Anderson JM, Rodriguez A, Chang DT. Foreign body reaction to biomaterials. Semin Immunol. 2008;20:86–100. 24. Guarino V, Alvarez-Perez MA, Cirillo V, Ambrosio L. hMSC interaction with PCL and PCL/gelatin platforms: a comparative study on films and electrospun membranes. J Bioact Compat Polym. 2011;26(2):144–60. 25. Schneider T, Kohl B, Sauter T, Kratz K, Lendlein A, Ertel W, Schulze-Tanzil G. Influence of fiber orientation in electrospun polymer scaffolds on viability, adhesion and differentiation of articular chondrocytes. Clin Hemorheol Microcirc. 2012;52:325–36. 26. Rydmark M. Nodal axon diameter correlates linearly with intermodal axon diameter in spinal roots of the cat. Neurosci Lett. 1981;24:247–50.
123
27. Schnell E, Klinkhammer K, Balzer S, Brook G, Kleeb D, Dalton P, Mey J. Guidance of glial cell migration and axonal growth on electrospun nanofibers of poly-e-caprolactone and a collagen/ poly-e-caprolactone blend. Biomaterial. 2007;28:3012–25. 28. Wang HB, Mullins ME, Cregg JM, McCarthy CW, Gilbert RJ. Varying the diameter of aligned electrospun fibers alters neurite outgrowth and Schwann cell migration. Acta Biomater. 2010;6: 2970–8. 29. Xie J, MacEwan MR, Li X, Sakiyama-Elbert SE, Xia Y. Neurite outgrowth on nanofiber scaffolds with different orders, structures, and surface properties. ACS Nano. 2009;3:1151–9. 30. Corey JM, Lin DY, Mycek KB, Chen Q, Samuel S, Feldman EL, Martin DC. Aligned electrospun nanofibers specify the direction of dorsal root ganglia neurite growth. J Biomed Mater Res A. 2007;83:636–45. 31. Wang HB, Mullins ME, Cregg JM, Hurtado A, Oudega M, Trombley MT, Gilbert RJ. Creation of highly aligned electrospun poly-L-lactic acid fibers for nerve regeneration. J Neural Eng. 2009;6:016001.
