Biofunctional Nanofibrous Substrate Comprising Immobilized

May 22, 2014 - This work aims to develop a biofunctional polymeric substrate capable of selectively binding growth factors (GFs) of interest from a po...
2 downloads 0 Views 5MB Size
Download PDF
Article pubs.acs.org/Biomac

Biofunctional Nanofibrous Substrate Comprising Immobilized Antibodies and Selective Binding of Autologous Growth Factors Catarina Oliveira,†,‡ Ana R. Costa-Pinto,†,‡ Rui L. Reis,†,‡ Albino Martins,†,‡ and Nuno M. Neves*,†,‡ †

3B’s Research Group − Biomaterials, Biodegradables and Biomimetics, Department of Polymer Engineering, University of Minho, Headquarters of the European Institute of Excellence on Tissue Engineering and Regenerative Medicine, AvePark, Zona Industrial da Gandra S. Cláudio do Barco, 4806-909 Caldas das Taipas, Guimarães, Portugal ‡ ICVS/3B’s, PT Government Associate Laboratory, 4710-057 Braga/Guimarães, Portugal S Supporting Information *

ABSTRACT: The immobilization of biomolecules at the surface of different biomedical devices has attracted enormous interest in order to enhance their biological functionality at the cellular level. This work aims to develop a biofunctional polymeric substrate capable of selectively binding growth factors (GFs) of interest from a pool of proteins present in a biological fluid: platelet lysate (PL). To achieve this goal, the surface of electrospun PCL nanofibers needs to be activated and functionalized to be able to insert chemical groups for the immobilization of antibodies. After determining the maximum immobilization capacity of each antibody, TGF-β1 (12 μg mL−1), bFGF (8 μg mL−1), and VEGF (4 μg mL−1), the next step was to confirm their bioavailability using recombinant proteins. The binding efficiency of PL-derived GFs was of 84−87% for TGF-β1, 55−64% for bFGF, and 50−59% for VEGF. Cellular assays confirmed the biological activity of the bound VEGF (both recombinant and PL-derived). Multiple antibodies (i.e., bFGF and VEGF) were also immobilized over the same structure in a mixed or side-by-side fashion. Using both autologous biological fluids and cells, it is possible to use this platform to implement very effective and personalized therapies that can be tailored to specific medical conditions.

1. INTRODUCTION The biofunctionalization of surfaces has gained special interest because of the need to optimize the biological integration of implantable medical devices.1 Of particular interest is the immobilization of antibodies, because of their highly specific interaction with their ligand molecules, to produce a more efficient and irreversible attachment of the target protein/ ligand. Protein immobilization requires that the substrate should provide a biocompatible and bioactive environment, as it should interact positively with the native structure of the proteins and biomolecules.2 Antibodies have been previously immobilized onto different substrates such as solid surfaces,3 microfluidic platforms,4 carboxyl surfaces,5 microarrays,6 ultraflat polystyrene surfaces,7 gold particles,8 chitosan microparticles,9 and chitosan surfaces.10 The ultimate goal of the antibodies’ immobilization at the substrate surface is to create specific sites for stimulating cell and protein adhesion.11 Different immobilization methods and strategies can be implemented to achieve the biofunctionalization of substrates such as physical adsorption, covalent immobilization, or antibody-binding proteins.12 Covalent immobilization is the most reported method because it leads to a strong and stable attachment. In this method, the presence of functional groups is required in both the substrate and the molecule to be © 2014 American Chemical Society

immobilized. Because of the use of coupling agents, modifications of the antigen binding site can occur, which may cause a loss of functionality of the antibodies. Therefore, there is a need to confirm that the biological activity of the corresponding bound antigen was not compromised. With this immobilization method, it is also possible to overcome the problems related to the adsorption method, where deadsorption and antibody denaturation often occur.11,12 Electrospun nanofiber meshes (NFMs) are very interesting polymeric substrates because of their high specific surface area, flexibility in surface functionality, and physical fibrous structure that mimics the morphology of the native extracellular matrix of most tissues.13,14 Other important properties of electrospun NFMs are their high levels of porosity and pore sizes that are smaller than the dimensions of most cells. These features are of high importance for various applications in tissue engineering and regenerative medicine because electrospun NFMs will be selectively permeable to soluble factors that hinder the permeation of cultured cells.15 Received: March 5, 2014 Revised: April 25, 2014 Published: May 22, 2014 2196

dx.doi.org/10.1021/bm500346s | Biomacromolecules 2014, 15, 2196−2205

Biomacromolecules

Article

by pH adjustment to 4.7, and the solution was mixed for 15 min at room temperature for antibody activation. Five different EDC/NHS ratios were tested (1:4, 1:2, 1:1, 2:1, and 4:1), and the optimized ratio was further assayed at four different concentrations (10 mM EDC + 40 mM NHS, 26 mM EDC + 104 mM NHS, 50 mM EDC + 100 mM NHS, and 100 mM EDC + 400 mM NHS). Under the optimized reaction conditions, in terms of the EDC/NHS ratio and respective concentrations, the final concentration of the linker in the antibody solution was determined at 1, 5, and 10% concentrations. 2.5. Optimization of Single Antibody Immobilization and Standard Curves Determination. Three different antibodies were immobilized (anti-TGF-β1, anti-bFGF, and anti-VEGF) at the surface of the activated and functionalized electrospun nanofiber meshes. A wide range of primary antibody concentrations (from 0 to 20 μg mL−1) was tested to determine the maximum immobilization capacity of the nanofibrous substrate. The electrospun NFMs were incubated with 200 μL of each primary antibody concentration overnight at 4 °C. NFMs were washed twice with PBS. The BSA solution was removed, the secondary antibody (1:200 in PBS) was applied, and the NFMs were incubated for 1 h at room temperature. In order to determine the degree of immobilization, an indirect method was used to quantify the fluorescence of unbound secondary antibody solution (n = 3 samples, read in triplicate). For the TGF-β1 and VEGF antibodies, Alexa Fluor 594 was used, and the reading parameters were the absorption at 590 nm and the emission at 617 nm. In the case of the anti-bFGF antibody, the selected secondary antibody was Alexa Fluor 488-conjugated, and the reading parameters were 495 nm for the adsorption and 519 nm for the emission. Negative control samples were also prepared, where all antibody immobilization steps were performed with the exception of the primary antibody incubation, which was substituted with PBS. 2.6. Mixed Immobilization of Antibodies. The VEGF and bFGF antibodies were mixed in the same PBS solution at the concentrations previously optimized, resulting in a final volume of 200 μL per mesh. The antibody mixture was incubated overnight at 4 °C, the samples were washed twice with 0.1 M PBS (5 min each), and a 3% BSA incubation step for 30 min at room temperature was performed to block all nonspecific sites. The BSA solution was removed, and the Alexa Fluor 594-conjugated secondary antibody (for anti-VEGF) was incubated with the samples for 1 h at rt. The excess secondary antibody solution was collected for further quantification (n = 3 samples, read in triplicate), as previously described, and the sample was washed twice. The same approach was carried out for the Alexa Fluor 488-conjugated secondary antibody (for anti-bFGF). Both secondary antibodies were prepared at a 1:200 dilution in PBS. A negative control sample was performed without the immobilization of the primary antibodies, although all other steps were done to allow the samples to be observed by laser scanning confocal microscopy. 2.7. Side-by-Side Immobilization of Antibodies. In order to obtain a substrate with two different antibodies immobilized in a sideby-side design, a compartmental water-tight chamber was developed capable of physically dividing a single 1 × 2 cm2 functionalized electrospun NFM into two distinct areas without allowing the different antibody solutions to mix. Two different antibody solutions (i.e., antiVEGF and anti-bFGF) were prepared at the concentrations described above and dropped over each area of the functionalized electrospun NFM. All of the antibody immobilization steps (washing, BSA blocking, and secondary antibody incubation) were performed as previously described for the single antibody immobilization. The quantification of unbound secondary antibody was also performed, and the samples were recovered to characterize the spatial distribution of the antibodies by laser scanning confocal microscopy. 2.8. Laser Scanning Confocal Microscopy. Laser scanning confocal microscopy was conducted in order to characterize the spatial distribution of the antibodies at the surface of the electrospun NFMs. The fluorescently labeled biological molecules were analyzed by selecting the appropriate wavelengths: excitation at 495 nm for Alexa Fluor 488 and 590 nm for Alexa Fluor 594 and emission at 570 nm for the red channel and 540 nm for the green channel. The single and multiple (either mixed or side-by-side) antibody-immobilized (i.e., TGF-β1, bFGF, and VEGF) samples were placed onto glass slides and

Proteins, especially growth factors (GFs), have an important role in the regulation of a variety of cellular processes as well as in the coordination of the healing process in different tissues.16 GF functions and purposes range from inducing vascularization and angiogenesis to regulating cell growth, proliferation, and differentiation. Therefore, the local or systemic administration of GFs may be a valuable therapeutic approach in the successful treatment of different chronic wounds.17 Clinically, GF-based strategies are applied in plastic surgery, bone and cartilage lesions, muscle injuries, ulcers, and skin regeneration.18 In particular, biological fluids, such as platelet lysate (PL), that consist of a cocktail of different GFs (e.g., bFGF, VEGF, TGFβ, PDGF-ββ, EGF, and IGF-I) provide a complex mixture of chemical signals to the cells at the injury site, which is often associated with a nonspecific action.19,20 The leading goal of the work presented herein is to develop a highly biofunctional nanofibrous substrate, taking advantage of the specific and efficient interactions between a specific antibody and its antigen. With this strategy, it is possible to selectively immobilize the GFs of interest (i.e., bFGF, TGF-β1, and VEGF) from a pool of highly concentrated GFs present in PL for the envisioned application. Furthermore, this biofunctionalization strategy also enables the simultaneous immobilization of multiple antibodies at a time, distributed in a mixed or in a side-by-side fashion over the same nanofibrous substrate, to elucidate the synergistic effect of the combined GFs.

2. MATERIALS AND METHODS 2.1. Materials. Polycaprolactone (PCL; Mw = 70 000−90 000 determined by GPC), chloroform, N,N-dimethylformamide (DMF), hexamethylenediamine (HMD), 1-ethyl-3-(3-(dimethylamino)propyl)carbodiimide (EDC), and hydroxysuccinimide (NHS) were purchased to Sigma-Aldrich and use as received. Mouse anti-human TGF-β1 monoclonal antibody was purchased from PeproTech Inc. (Rocky Hill, NJ, USA), rabbit anti-human bFGF oligoclonal antibody (clone 7HCLC, ABfinity recombinant) was purchased from Life Technologies (Carlsbad, CA, USA), and mouse anti-human VEGF (JH121) was purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA, USA). Regarding the secondary antibodies, Alexa Fluor 488 donkey anti-rabbit IgG (H + L) and Alexa Fluor 594 goat anti-mouse IgG (H + L) were purchased from Life Technologies (Carlsbad, CA, USA). The GFs, namely, recombinant human TGF-β1, recombinant basic-FGF, and recombinant human VEGF121, were all purchased from PeproTech Inc. (Rocky Hill, NJ, USA). 2.2. Electrospinning of Nanofiber Meshes. A 17% (w/v) PCL solution was prepared with an organic solvent mixture of chloroform and DMF at a 7:3 ratio as described elsewhere.21 The PCL solution was electrospun by applying a voltage of 13.6 kV, a needle tip to ground collector distance of 18 cm, and a flow rate of 1 mL h−1. After the complete processing of 1 mL of PCL solution, the nanofiber mesh (NFM) was allowed to dry for 1 day. This processed NFM was cut into samples of 1 × 1 cm2 for further assays. 2.3. Ultraviolet−Ozone Irradiation and Aminolysis. For the activation of the nanofibers surface, an ultraviolet−ozone (UV−ozone) cleaner system was used (ProCleaner 220, Bioforce Nanoscience). Both sides of the electrospun NFMs were exposed for 4 min to UV− ozone irradiation, as optimized previously. After this surface activation, amine groups (−NH2) were inserted at the surface of the electrospun nanofibers by immersion in a 1 M 1,6-hexanediamine (HMD) (SigmaAldrich) solution for 1 h at 37 °C. The amount of −NH2 (2.83 ± 0.11 nmol cm−2) was determined indirectly by quantifying the amount of free −SH groups according to a method using Ellman’s reagent.22 2.4. EDC/NHS Ratio and Concentrations Optimization. EDC/ NHS reagents were dissolved in 0.1 M 2-(N-morpholino)ethanesulfonic acid (MES) buffer with 0.9% (w/w) NaCl followed 2197

dx.doi.org/10.1021/bm500346s | Biomacromolecules 2014, 15, 2196−2205

Biomacromolecules

Article

analyzed by laser scanning confocal microcpy (FluoView 1000, Olympus, Germany) at 10× magnification. 2.9. Platelet Lysates: Preparation and Activation. Platelet rich plasma (PRP) was provided by the Instituto Português do Sangue, which certifies the biological product according to Portuguese law. The number of platelets was counted, and the sample volume was adjusted to 106 platelets/μL. At the 3B’s Research Group facilities, the collected PRP samples were subjected to three repeating temperature-shock cycles (i.e., frozen with liquid nitrogen at −196 °C and further heated at 37 °C), and the remaining platelets were eliminated by centrifugation. A pool of platelet lysates (PL) was stored at −20 °C until further use. At the time of each experiment, a PL solution was thawed and filtered with a 0.22 μm filter. 2.10. Fluorescence-Linked Immunosorbent Assay (FLISA). After completing all of the antibody immobilization steps previously described, 200 μL of the recombinant human protein solutions at different concentrations (ranging from 0 μg mL−1 to the concentration of each primary antibody) were incubated for 1 h at room temperature. The unbound recombinant human protein solutions were collected and stored at −20 °C until further quantification by ELISA. Two 0.1 M PBS washing steps (5 min each) were performed, and the biofunctionalized nanofibrous substrates were incubated overnight at 4 °C with the corresponding primary antibody. After removal of the excess primary antibody solutions, the biofunctionalized substrates were washed again with PBS, another BSA blocking step was performed, and the corresponding secondary antibody was incubated for 1 h at room temperature. The fluorescence of unbound secondary solutions was also read out in a microplate reader (Synergy HT, Bio-Tek). When the PL was used as the natural source of GFs, the same procedure was followed, although the recombinant human protein solution was substituted with 200 μL of PL. 2.11. Enzyme-Linked Immunosorbent Assay (ELISA). For the quantification of unbound GFs, human basic-FGF and VEGF development ELISA kits were purchased from PeproTech (Rocky Hill, NJ, USA), whereas the human TGF-β1 ELISA kit was purchased from Boster Biological Technology (Fremont, CA, USA). For the bFGF and VEGF development ELISA kits, the primary antibodies were first incubated overnight in a 96-well plate (Nunc-Immuno MicroWell 96-well solid plates, Sigma-Aldrich). All solutions were prepared according to the manufacturer’s protocol, and in the last step, 100 μL of 2,2′-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) (ABTS) liquid substrate was added to each sample, and each plate was read at 405 and 650 nm. The TGF-β1 ELISA is a ready-to-use kit, where the bottom wells were previously coated with the antibody. Both the standards and the samples were incubated, and the assay was conducted according to the manufacturer’s protocol. In the last step of the procedure, 100 μL of the 3,3′,5,5′-tetramethylbenzidine (TMB) stop solution was added to each well, and the absorbance at 450 nm was read (Synergy HT, Bio-TEK). 2.12. Biological Assays. 2.12.1. Cell Culture and Seeding. A human pulmonary microvascular endothelial cell line (HPMECST1.6R) was used to validate the developed biofunctionalized nanofibrous substrate. This cell line has been used to study the angiogenic process in vitro.23 HPMEC-ST1.6R cells were cultured in M199 medium (SigmaAldrich) supplemented with 20% FBS (Alfagene), 2 mM Glutamax (Life Technologies), pen/strep (100 U/100 μg mL −1 ; Life Technologies), heparin (50 μg mL−1; Sigma-Aldrich), endothelial cell growth supplement (ECGS, 25 μg mL−1; Becton Dickinson) and incubated at 37 °C in a humidified 5% CO2 atmosphere. HPMECST1.6R were used at passages 30−32. Medium was changed twice a week until cell reached 90% confluence. Then, cells were harvested and seeded onto the activated and functionalized electrospun NFMs. The electrospun PCL NFMs were sterilized by ethylene oxide at Pronefro Produtos Nefrológicos, S.A. (Porto, Portugal). For HPMECST1.6R cell seeding, anti-VEGF antibody was immobilized overnight, and after the BSA blocking step, human recombinant protein (VEGF = 4 μg/mL) and PL were then added. Cell seeding was performed by dropping a 50 μL cell suspension containing 50 000 cells per substrate, and the cells were left overnight. After cell attachment, culture medium

was added. Tissue culture polystyrene (TCPS) coverslips, untreated electrospun PCL NFMs, and NFMs in which nanofibers were subjected to surface activation, aminolysis, and primary antibody immobilization were used as controls. After 1, 3, and 7 days of culture, samples were collected for cell viability assay, DNA, total protein quantification, and VEGF quantification. 2.12.2. Cell Viability. The metabolic activity of HPMEC-ST1.6R cells seeded on TCPS, untreated electrospun PCL NFM, NFMs with primary antibody immobilization, and biofunctionalized nanofibrous substrates (recombinant and PL-derived) was determined by MTS assay (CellTiter 96 AQueous One Solution, Promega). Briefly, at days 1, 3, and 7, the culture medium was removed and the samples were rinsed with sterile PBS. A mixture of culture medium and MTS reagent (5:1 ratio) was added to each mesh as well as to the negative control comprising no cells or samples. All conditions were performed in triplicate and left to incubate for 3 h at 37 °C in a humidified 5% CO2 atmosphere. Thereafter, the absorbance of the MTS reaction medium from each sample was read in triplicate at 490 nm (Synergy HT, BioTEK). 2.12.3. Cell Proliferation. Cell proliferation was determined by using a fluorimetric dsDNA quantification kit (Quant-iT, PicoGreen, Molecular Probes, Invitrogen). The samples were collected at days 1, 3, and 7, washed twice with sterile PBS, and transferred into eppendorf tubes containing 1 mL of ultrapure water. These sample were frozen at −80 °C until further analysis. Prior to DNA quantification, the various specimens for each samples were thawed and sonicated for 15 min. DNA standards were prepared at concentrations ranging from 0 to 2 μg mL−1. To each well of an opaque 96-well plate (Falcon) were added 28.7 μL of sample or standard (n = 3), 71.3 μL of PicoGreen solution, and 100 μL of TE buffer. The plate was incubated for 10 min in the dark, and the fluorescence was measured in a microplate reader (Synergy HT, Bio-Tek) using an excitation wavelength of 480 nm and an emission wavelength of 528 nm. The DNA concentration of each sample was calculated using a standard curve relating the DNA concentration and the fluorescence intensity. 2.12.4. Total Protein Synthesis. Samples were collected and prepared for assaying, as described above (Section 2.12.3). For the quantification of total protein synthesis, a micro BCA protein assay kit (Pierce, Thermo Scientific) was used. The assay was performed according to the manufacturer’s instructions. Briefly, standards were prepared at various concentrations ranging from 0 to 40 μg mL−1 in ultrapure water. One hundred fifty microliters of both samples and standards was assayed in triplicate, and 150 μL of working reagent was further added to each 96-well plate. The plate was sealed and incubated for 2 h at 37 °C. The plate was left to cool to room temperature, and then the absorbance at 562 nm was measured in a microplate reader (Synergy HT, Bio-Tek). 2.13. Statistical Analysis. Statistical analysis was performed using GraphPad Prism Software. Differences between the conditions tested in the cellular assays were analyzed using a nonparametric test (Kruskal−Wallis test), and p < 0.05 was considered to be statistically significant. Data are presented as the mean ± standard deviation.

3. RESULTS AND DISCUSSION 3.1. Optimization of Single Antibody Immobilization at the Nanofibrous Substrate. In the present study, a defined antibody was immobilized at the surface of electrospun nanofibers by a covalent bond mediated by a coupling agent; in this particular case, the EDC/NHS mixture. Prior to the immobilization step, the electrospun PCL NFMs needed to be chemically modified by the insertion of amine groups that could react specifically with the carboxyl groups of the antibody. It is know that EDC alone is able to increase the immobilization efficiency of biomolecules.9,24 However, with the addition of NHS, a two-step reaction occurrs, and the presence of NHS forms semistable amines, enhancing the immobilization efficiency of the antibodies at the surface of a substrate.25 In order to ensure that the nanofibrous substrate is being used at 2198

dx.doi.org/10.1021/bm500346s | Biomacromolecules 2014, 15, 2196−2205

Biomacromolecules

Article

Figure 1. Maximum immobilization capacity of a single antibody at the surface of activated and functionalized electrospun nanofibers: (a) immobilization of anti-TGF-β1, (b) immobilization of anti-bFGF, and (c) immobilization of anti-VEGF.

Figure 2. Spatial distribution of immobilized primary antibodies at the surface of activated and functionalized electrospun nanofibers. Primary antibodies were immobilized at the previously optimized concentrations: (a) 12 μg mL−1 of anti-TGF-β1, (b) 8 μg mL−1 of anti-bFGF, and (c) 4 μg mL−1 of anti-VEGF. In the case of the TGF-β1 and VEGF antibodies, the Alexa Fluor 594-conjugated secondary antibody was used, whereas the Alexa Fluor 488-conjugated secondary antibody was used for the bFGF antibody. The negative control samples (d−f) were not incubated with the primary antibody, altough all other incubation steps were performed.

its maximum immobilization capacity, different parameters concerning these two coupling reagents were optimized. The ratio was set at 1:4 ratio (Figure S1a), the optimal concentration was 50 mM EDC + 200 mM NHS (Figure S1b), and the concentration of EDC/NHS mixture in the final antibody solution was defined as 1% (v/v) (Figure S1c).

The antibodies against growth factors TGF-β1, bFGF, and VEGF were immobilized at the surface of activated and functionalized electrospun nanofibers in a wide range of concentrations (0−20 μg mL −1) to determine the maximum immobilization capacity of the nanofibrous substrate for each antibody. To achieve this goal, an indirect quantification method was used, based on the measurement of unbound 2199

dx.doi.org/10.1021/bm500346s | Biomacromolecules 2014, 15, 2196−2205

Biomacromolecules

Article

Figure 3. Capability of the biofunctionalized nanofibrous substrate to bind different concentrations of the recombinant protein: (a) TGF-β1, (b) bFGF, and (c) VEGF.

secondary antibody fluorescence after its incubation with the immobilized primary antibody. As observed in Figure 1, higher amounts of immobilized primary antibody correspond to a lower fluorescence signal of the free secondary antibody. When the fluorescence signal reaches a plateau, the nanofibrous substrate presents the maximum concentration of immobilized primary antibody, reaching the saturation point of the system. With the anti-TGF-β1 antibody, the maximum concentration of immobilized primary antibody is 12 μg mL −1. In the case of anti-bFGF antibody, the maximum capacity of the nanofibrous substrate is 8 μg mL −1, whereas with the anti-VEGF antibody its concentration reaches a value of 4 μg mL −1. As can be clearly seen, the different antibodies have different densities over the same activated and functionalized nanofibrous substrate. This observation may be related to the different sizes of the primary antibodies used. According to manufacturer’s data, the size of the anti-VEGF antibody is approximately 22 kDa, and the anti-bFGF antibody is 17 kDaA. In the case of the anti-TGF-β1 antibody, no information is given by the company, although according to other manufacturers, the molecular weight of anti-TGF-β1 antibody should be around 12−14 kDa. The immobilization data correlates with the antibodies size, because the more surface area occupied by the antibody, the lower its concentration at the nanofiber surface. Therefore, further experiments were performed with the concentrations of the primary antibodies that lead to the maximum immobilization capacity of the nanofibrous substrate. Antibodies were immobilized on different substrates, as referred to before, in a similar fashion as that of common ELISA methods, which also involve the immobilization of specific antibodies at the bottom of the well to quantify the amount of specific antigen existing in a sample. From the datasheet of the ELISAs, referred in the Materials and Methods section, it is possible to conclude that the values for anti-bFGF and anti-VEGF immobilization are 1 and 0.5 μg mL −1, respectively. The ELISA tests are conducted in 96 well-plates having approximately 0.32 cm2 of surface area per well. In the case of the developed biofunctionalized nanofibrous substrate, the apparent surface area is about three times higher (1 cm2), and despite having the same ratio between them, the immobilization capacity is eight times (8 μg mL−1 for anti-bFGF and 4 μg mL−1 for anti-VEGF) higher than that in the standard ELISAs. After determining the maximum antibody concentration immobilized at the surface of activated and functionalized electrospun nanofibers, a standard curve was determined for each antibody. A linear regression standard curve revealed an R2 value above 0.98 for each antibody (Figure S2).

The spatial distribution of the TGF-β1, bFGF, and VEGF antibodies at the surface of the activated and functionalized electrospun nanofibers is shown in Figure 2. All of the immobilized antibodies seem to be uniformly distributed through the nanofibers surface, resembling the random meshlike arrangement of the electrospun NFM structure. The TGFβ1 antibody seems to have a more intense and densely distributed fluorescence than the other immobilized antibodies, probably because of its higher concentration (12 μg mL−1). To ensure that the secondary antibody binds only to the immobilized primary antibody, we defined a control experiment in which all steps were performed except for the incubation with the primary antibody. These conditions were analyzed for fluorescence (Figure 2d−f), and no fluorescence was detected. Because no fluorescence signal was observed, Alexa Fluor 594and Alexa Fluor 488-conjugated secondary antibodies were not immobilized at the surface of activated and functionalized electrospun nanofibers, confirming the specific binding between these two secondary antibodies and their corresponding immobilized primary antibodies. 3.2. Growth Factors Binding Capacity to the Biofunctionalized Nanofibrous Substrate. 3.2.1. The Quantification of Bound Recombinant Proteins. After confirming the specific immobilization of the TGF-β1, bFGF, and VEGF antibodies and determining the corresponding standard curves, the binding capacity of the biofunctionalized nanofibrous substrates was assessed. Namely, we characterized the total amount of each growth factor that can bind to a functionalized mesh. For this, two different GF sources were tested: (i) recombinant proteins, to evaluate the maximum binding capacity of the biofunctionalized nanofibrous substrate, and (ii) PL-derived GFs, to assess the selective binding capacity of the biofunctionalized nanofibrous substrate. In fact, when using recombinant protein, we know that the only protein competing with the primary antibody is the one being tested. When PLs are used, there is a complex mixture of proteins competing for the antibodies, thus demonstrating the specificity of the bound proteins. Following the immobilization of each antibody at the surface of the activated and functionalized electrospun nanofibers, the corresponding recombinant protein was added at different concentrations, varying from 0 μg mL−1 to values higher than the concentration of the previously immobilized primary antibody (Figure 1). However, for all three antibodies in this study, the biofunctionalized nanofibrous substrate starts to reach its maximum GF binding capacity near the higher concentration of the primary antibody (i.e., 12 μg mL−1 for TGF-β1, 8 μg mL−1 for bFGF, and 4 μg mL−1 for VEGF). The results were also confirmed and quantified with commercially 2200

dx.doi.org/10.1021/bm500346s | Biomacromolecules 2014, 15, 2196−2205

Biomacromolecules

Article

immobilized antibody. Despite the concentrations of the GFs in the different samples, the binding of the GFs stayed in the same range for the two donor samples, showing the consistency of the method. Despite the order of magnitude differences in the concentration of GFs present in the PL (ranging from picograms per milliliter for VEGF to nanograms per milliliter for TGF-β1 and bFGF), these concentrations are much lower than the ones determined for the maximum binding capacity of the biofunctionalized nanofibrous substrate where recombinant proteins were used at a micrograms per milliliter concentration. The inability of the biofunctionalized nanofibrous substrate to immobilize 100% of the GFs present in the PL can be related to the fact that this biological fluid is highly rich in different GFs and proteins that can compete with the binding sites of each immobilized antibody. Another technical aspect that can explain the binding of GFs derived from the PL being in the range of 50−87% is the detection limits of the ELISAs used, which do not enable the GFs to be detected at very low concentrations (16 pg mL−1 for VEGF, 63 pg mL−1 for bFGF, and 15.6 pg mL−1 for TGF-β1). 3.3. VEGF Biological Activity. To confirm that the covalent immobilization method does not compromise the bioavailability of the antigen binding site of the antibodies and the behavior of the bound growth factors, we assessed the bioactivity of bound VEGF. VEGF was selected because VEGF is a less concentrated factor in PL and therefore is bound less to the biofunctionalized nanofibrous substrate. VEGF has been described to induce vascularization and angiogenesis; in this sense, human pulmonary microvascular endothelial cells (HPMEC-ST1.6R cell line) were seeded onto the biofunctionalized electrospun nanofibrous substrates. Different biological assays were conducted to assess the endothelial cells viability and proliferation and total protein synthesis as well as to quantify the intracellular synthesis of VEGF (Figure 4). The endothelial cells were cultured at the surface of five different substrate conditions: (i) untreated electrospun PCL NFM (NFM), (ii) NFM with primary antibody (NFM_Ab1), (iii) electrospun NFM with bound recombinant VEGF (NFM + VEGFRec), (iv) electrospun NFM with PL-derived VEGF (NFM + VEGFPL), and (v) NFM with primary anti-VEGF (NFM_Ab2). In conditions i and ii, the medium was supplemented with endothelial cell growth supplement (ECGS), a mixture of growth factors aimed at stimulating the growth of human and animal vascular endothelial cells.35,36 In conditions iii−v, the endothelial cells were cultured in basal medium because we wanted to evaluate the actions of the immobilized form of VEGF, so the medium was not supplemented by other GFs. Our data confirms the biological activity of the bound VEGF, because in all assays performed (cell proliferation, cell viability, total protein synthesis, and intracellular VEGF), significant differences between NFM + VEGFRec and NFM + VEGFPL compared to NFM_Ab2 (this condition only differs from the previous one by not having an immobilized protein) were always observed. This observation undoubtedly demonstrates that bound VEGF (recombinant or PL-derived) indeed makes a difference in the biofunctionalized nanofibrous substrate. Because there were no significant differences between NFM + VEGFPL and NFM + VEGFRec, even at a much higher concentration of recombinant VEGF (4 μg mL−1), it is possible to conclude that the amount of protein derived from PL samples (0.251 μg mL−1) is enough to induce cell proliferation.

available ELISAs because the maximum loading capacity corresponding to the above-mentioned values, reached almost 100% of the loading efficiency for all three GFs. As observed in Figure 3, an increase in the amount of recombinant protein leads to a decrease in the fluorescence signal, meaning that less secondary antibody is unbound. It is important to note that each GF bound to the biofunctionalized nanofibrous substrate has its own slope or rate. A higher slope is observed for TGFβ1 (40.833), followed that for bFGF (27.259); the lowest slope is for VEGF (21.771). A possible explanation for this observation relies on the fact that higher concentrations of primary antibodies are occupied by the corresponding recombinant protein faster, leading to a rapid decrease of the fluorescence signal of the unbound secondary antibodies. TGFβ1 has been previously reported to be immobilized in different substrates like gelatin26 and magnetic beads,27 whereas bFGF has been immobilized into different platforms like PEG hydrogel,28 PLGA films,29 and PLLA + collagen scaffolds.30 VEGF has been covalently immobilized into different substrates such as collagen scaffolds,31,32 PLGA,33 and hydrogels.34 From all of this literature data, it is possible to conclude that our nanofibrous substrate enables a higher concentration of GFs to be immobilized (on the order of micrograms per milliliter), whereas most of the other systems reported values that are on the order of nanograms per milliliter, reflecting the positive effect of the increased surface area of the electrospun nanofibers over the bound GFs. 3.2.2. Quantification of Bound PL-Derived Growth Factors. We quantified the amount of each GF of interest in the PL by ELISA. Table 1 shows the range of concentrations Table 1. Quantification of the Growth Factors of Interest Derived from the Human PL Samples concentration in PL samples (ng mL−1)

% binding

growth factors

donor 1

donor 2

donor 1

donor 2

TGF-β1 bFGF VEGF

4.2 8.6 0.0949

11.05 102.5 0.4263

83.92 ± 2.68 54.78 ± 4.75 49.52 ± 3.05

86.85 ± 3.26 63.97 ± 3.48 58.85 ± 4.02

obtained from two different human samples. Comparing this data to other values reported in the literature, TGF-β1 (169.9 ± 84.5 ng mL−1) is about 15 times higher than the ones obtained with our own samples, whereas the values of VEGF are within the range of other reported values (0.076 to 0.854 μg mL−1).19 Despite being described as one of the most abundant GFs of PRP samples, we found no data reporting the concentration of bFGF in the literature. The differences among the quantified GFs and their variability are related to the differences between donors, leading to different concentrations of the GFs of interest. After determining the recombinant human GF binding capacity of the activated and functionalized nanofibrous substrate, it was tested for the selective binding of GFs derived from PL samples. ELISAs were performed to determine the amount of bound autologous GFs for two different donors. For PL-derived TGF-β1, the binding efficiency (∼84−87%) was not as high as in the case of the recombinant protein. For bFGF, only around 55−64% of PL-derived protein was bound to the nanofibrous substrate-immobilized primary antibody. The same trend was observed for VEGF, where about 50−58% of PL-derived VEFG was bound by the corresponding 2201

dx.doi.org/10.1021/bm500346s | Biomacromolecules 2014, 15, 2196−2205

Biomacromolecules

Article

Figure 4. Biochemical performance of the endothelial cell line cultured on an unmodified electrospun PCL NFM, NFM with immobilized VEGF antibody (NFM_Ab1), NFM with immobilized VEGF antibody (NFM_Ab2), NFM with bound recombinant VEGF (NFM + VEGFRec), and PLderived VEGF (NFM + VEGFPL) (the first two conditions used media supplemented with ECGS; the last three conditions did not): (a) cell proliferation, (b) cell viability, (c) total protein synthesis, and (d) intracellular VEGF synthesis. Statistical analysis was performed for the five different conditions to compare each time point (days 1, 3, and 7). Data was considered statistically different if p < 0.05. * indicates significant differences when compared to NFM; +, when compared to NFM_Ab1 supplemented media, x, when compared to NFM_Ab2; #, when compared to NFM_VEGF; and &, when compared to NFM_PL.

sustained influence over the cells compared to the transient effect of soluble VEGF. GF immobilization on the substrates also promotes a local regulation and control over cellular activity, as expressed by the intracellular VEGF overexpression. Furthermore, immobilized GFs can provide extended signaling because the ligand will not be internalized as a ligand/receptor complex. Covalent attachment of angiogenic growth factors to biomaterial scaffolds is an advantageous strategy for the development of polymeric matrixes with enhanced angiogenic capabilities.37 3.4. Immobilization of Multiple Antibodies in Different Spatial Configurations. The methodology reported herein aims at immobilizing more than one antibody at the

Additionally, an excess of immobilized recombinant protein did not have adverse effects on the endothelial cells. Our controls do not present significant differences among them, showing that the chemical treatment does not affect the behavior of the cells. Another important aspect of these cellular experiments is the observation of significant differences between the bound VEGF and the conditions where the GFs are given in their soluble form (NFM and NFM_Ab1), demonstrating their bioactivity over time. This same trend has already been reported and shows some evidence that the immobilized form can be beneficial for endothelial cell metabolic activity and protein synthesis.31 This is probably due to the fact that immobilized VEGF can provide a more controlled and 2202

dx.doi.org/10.1021/bm500346s | Biomacromolecules 2014, 15, 2196−2205

Biomacromolecules

Article

substrate in a mixed design, the corresponding secondary antibodies were used, and their fluorescence was observed using a laser scanning confocal microscope. Figures 6a,b represents the Alexa Fluor 488- and Alexa Fluor 594-conjugated fluorescent antibodies bound to the anti-bFGF

surface of a single activated and functionalized NFM. Two different immobilization designs are presented: one with a mixed distribution of defined antibodies (i.e., anti-VEGF and anti-bFGF) and another with side-by-side localization of these distinct antibodies in different areas of the same nanofibrous substrate. With the immobilization of multiple antibodies at the surface of the same nanofibrous substrate, it is expected that this will be developed into a highly efficient system applied to cell biology, tissue engineering, and regenerative medicine strategies 3.4.1. Mixed Immobilization of Two Different GFs. The purpose of the mixed immobilization is to have two different but complementary antibodies at the surface of the same nanofibrous substrate, specifically the anti-bFGF and antiVEGF antibodies. To implement this strategy, the antibody concentrations optimized before (i.e., 8 μg mL−1 for bFGF and 4 μg mL−1 for VEGF) were used, and the antibodies were incubated with the substrate simultaneously. Figure 5 expresses the amount of the initial antibody concentration that was immobilized; these results were

Figure 6. Spatial distribution of randomly immobilized primary antibodies at the surface of a single activated and functionalized nanofibrous substrates. The bFGF and VEGF antibodies were simultaneously immobilized in the same mesh at the previously optimized concentrations. (a) Alexa Fluor 448-conjugated antibody was used as the secondary antibody for anti-bFGF; (b) Alexa Fluor 594-conjugated antibody was used as the secondary antibody for antiVEGF; (c) spatial distribution of the two primary antibodies (merge view); and (d) activated and functionalized nanofibrous substrates without primary antibody immobilization. Figure 5. (a) Quantification of randomly immobilized bFGF and VEGF antibodies. (b) Relative quantification of randomly bound GFs (i.e., VEGF and bFGF) derived from PL.

and anti-VEGF antibodies at the surface of the activated and functionalized nanofibrous substrate. The antibodies are uniformly distributed over the functionalized nanofibrous substrate. However, the green fluorescence is slightly more intense than the that for the anti-VEGF antibody immobilization. This may be related to the higher concentration of the immobilized anti-bFGF antibody used, which can lead to a higher intensity of Alexa Fluor 488-conjugated antibody fluorescence. Figure 6c shows a merged image of the two different channels/fluorescences (green and red), corresponding to the VEGF and bFGF immobilized antibodies distributions. The merging of the two pictures yields a significant amount of yellow spots, which demonstrate the colocalization of both antibodies in a mixed fashion over the same nanofibrous substrate. Lastly, Figure 6d presents the negative control of the experiment, where the incubation step with the primary antibodies solution was not conducted to ensure that the secondary antibodies are bound only to the corresponding primary antibodies. 3.4.2. Side-by-Side Immobilization of Two Distinct Antibodies. Using the side-by-side immobilization of two antibodies, we intend to demonstrate the possibility of having two distinct GFs selectively immobilized from biological fluids, bound side by side, having in mind their functional roles over two distinct cell types spatially juxtaposed in physiological environments. In order to achieve this purpose, a compartmental water-tight device was designed to enable the creation of two distinct areas in the same nanofibrous substrate. Each

obtained by applying the determined standard curves for the single antibody immobilization strategy (Figure S2). In the case of the anti-bFGF antibody, around 63% of the initial antibody concentration was immobilized, whereas around 72% of the initial concentration was immobilized in the case of the antiVEGF antibody. These immobilization efficiencies were above the expected outcomes because the two antibodies, although used at different concentrations, are competing for the same amount of NH2 available at the surface of an activated and functionalized nanofibrous substrate. The antibodies can be immobilized until the system reaches its maximum capacity at which it does not have more chemical groups available for antibody immobilization. Therefore, in the mixed distribution of two antibodies, it is reasonable that the immobilization efficiency is not as high as that for the single immobilization strategy. Human PL samples were also incubated with the mixed immobilization nanofibrous substrate to evaluate if this system was able to bind the two growth factors of interest selectively and simultaneously. As observed in Figure 5b, the binding efficiency was about 64% for bFGF and 65% for VEGF. These binding efficiencies were slightly above the values obtained for the single binding strategy, which can be explained by the fact that a sample from a different donor was used. To evaluate the spatial distribution and to confirm that both antibodies were indeed immobilized at the same nanofibrous 2203

dx.doi.org/10.1021/bm500346s | Biomacromolecules 2014, 15, 2196−2205

Biomacromolecules

Article

Figure 7. Top: Schematic representation of the compartments in the water-tight device that allows the simultaneous immobilization of two different antibodies in two areas of a single mesh. Bottom: Laser scanning confocal microcopy image demonstrating the patterned antibodies immobilization over the same activated and functionalized nanofibrous substrate.

area of the activated and functionalized nanofibrous substrate was incubated with the defined primary antibody and then with the corresponding secondary antibody, leading to the patterned configuration and distribution presented in Figure 7. The reddish fluorescent area of the nanofibrous substrate corresponds to the anti-VEGF antibody immobilization, whereas the green area reports to the anti-bFGF antibody immobilization. The black area corresponds to the bar that separates the same nanofibrous substrate into the two areas. It is possible to detect local green spots or red dots, which can be the result of some diffusion of the antibody solutions through the activated and functionalized nanofibrous substrate. With this system, it will be possible to seed and culture two different cell types over the two areas of the biofunctionalized nanofibrous substrate, where a defined antibody and its corresponding GF are immobilized. With this strategy, it will be possible to obtain tailored and advanced coculture systems, allowing for the study of cell−cell interactions in vitro in the presence of specific GFs.

substrate-bound GFs were confirmed by biochemical data. The biological data suggests that this substrate offers unique possibilities to study basic cell biology as well as uses in the tissue engineering and regenerative medicine fields because it is possible to specifically bind different GFs of interest at the surface of the nanofibrous substrate. Ultimately, the developed biofunctionalized polymeric fibrous substrate can be used for the autologous regeneration of tissues, where both cells and bioactive factors are from the same patient, and its use in personalized therapy is envisioned.

4. CONCLUSIONS A covalent immobilization method was successfully implemented in nanofibrous substrates, presenting different efficiencies depending on the antibody of interest. After antibody immobilization in different patterns and designs, the biofunctionalized nanofibrous substrates enabled the binding of the corresponding growth factors as well as the selecting of a specific GF from a complex biological fluid (i.e., PLs) comprising a pool of different GFs and proteins. The bioactivity of the bound growth factors was confirmed by cell culture assays, and the beneficial outcomes of the nanofibrous

Corresponding Author

■

ASSOCIATED CONTENT

S Supporting Information *

EDC/NHS ratio and concentrations optimization, and standard curves for single antibody immobilization at the surface of activated and functionalized nanofiber meshes. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors would like to thank the Maxbone (PTDC/SAUENB/115179/2009/FCOMP-01-0124-FEDER-015729) and Osteography (PTDC/EME-MFE/2008) projects as well as QREN (project “RL1-ABMR-NORTE-01-0124-FEDER000016” cofinanced by the North Portugal Regional Opera2204

dx.doi.org/10.1021/bm500346s | Biomacromolecules 2014, 15, 2196−2205

Biomacromolecules

Article

(30) Ma, Z.; Gao, C.; Gong, Y.; Shen, J. Biomaterials 2005, 26, 1253− 1259. (31) Shen, Y. H.; Shoichet, M. S.; Radisic, M. Acta Biomater. 2008, 4, 477−489. (32) Miyagi, Y.; Chiu, L. L. Y.; Cimini, M.; Weisel, R. D.; Radisic, M.; Li, R.-K. Biomaterials 2011, 32, 1280−1290. (33) Sharon, J. L.; Puleo, D. A. Acta Biomater. 2008, 4, 1016−1023. (34) To, Y.; Hasuda, H.; Terai, H.; Kitajima, T. J. Biomed. Mater. Res., Part A 2005, 74, 659−665. (35) Folkman, J.; Haudenschild, C. Nature 1980, 288, 551−556. (36) Maciag, T.; Cerundolo, J.; Ilsley, S.; Kelley, P. R.; Forand, R. Proc. Natl. Acad. Sci. U.S.A. 1979, 76, 5674−5678. (37) Chiu, L. L. Y.; Weisel, R. D.; Li, R.; Radisic, M. J. Tissue Eng. Regener. Med. 2011, 5, 69−84.

tional Programme (ON.2, O Novo Norte) under the NSRF through the ERDF) for financing this research work.

■

REFERENCES

(1) Hildebrand, H. F.; Blanchemain, N.; Mayer, G.; Zhang, Y. M.; Melnyk, O.; Morcellet, M.; Martel, B. Key Eng. Mater. 2005, 288, 47− 50. (2) Giacomelli, C. E., Valenti, L. E., Carot, L. Biomolecules and Solid Substrate Interaction: Key Factors in Developing Biofunctional Surfaces; Taylor & Francis: Boca Raton, FL, 2007; pp 1−16. (3) Dixit, C. K.; Vashist, S. K.; MacCraith, B. D.; O’Kennedy, R. Nat. Protoc. 2011, 6, 439−445. (4) Feyssa, B.; Liedert, C.; Kivimaki, L.; Johansson, L.-S.; Jantunen, H.; Hakalahti, L. PLoS One 2013, 8, 68918. (5) Pei, Z.; Anderson, H.; Myrskog, A.; Dunér, G.; Ingemarsson, B.; Aastrup, T. Anal. Biochem. 2010, 398, 161−168. (6) Peluso, P.; Wilson, D. S.; Do, D.; Tran, H.; Venkatasubbaiah, M.; Quincy, D.; Heidecker, B.; Poindexter, K.; Tolani, N.; Phelan, M.; Witte, K.; Jung, L. S.; Wagner, P.; Nock, S. Anal. Biochem. 2003, 312, 113−124. (7) Qian, W.; Yao, D.; Yu, F.; Xu, B.; Zhou, R.; Bao, X.; Lu, Z. Clin. Chem. 2000, 46, 1456−1463. (8) Karyakin, A. A.; Presnova, G. V.; Rubtsova, M. Y.; Egorov, A. M. Anal. Chem. 2000, 72, 3805−3811. (9) Custódio, C. A.; Santo, V. E.; Oliveira, M. B.; Gomes, M. E.; Reis, R. L.; Mano, J. F. Adv. Funct. Mater. 2013, 24, 1391−1400. (10) Custódio, C. A.; Frias, A. M.; del Campo, A.; Reis, R. L.; Mano, J. F. Biointerphases 2012, 7, 65. (11) Dixit C. K. Ph.D. Thesis, Dublin City University, November 2011. (12) Jung, Y.; Jeong, J. Y.; Chung, B. H. Analyst 2008, 133, 697−701. (13) Zhan, J.; Singh, A.; Zhang, Z.; Huang, L.; Elisseeff, J. H. Biomatter 2012, 2, 202−212. (14) Pham, Q. P.; Sharma, U.; Mikos, A. G. Tissue engineering 2006, 12, 1197−211. (15) Martins, A.; Silva, M. A.; Reis, R. L.; Neves, N. M. Patent PCT/ IB2013/056218, July 27, 2012. (16) Barrientos, S.; Stojadinovic, O.; Golinko, M. S.; Brem, H.; Tomic-Canic, M. Wound Repair Regener. 2008, 16, 585−601. (17) Krishnamoorthy, L.; Morris, H. L.; Harding, K. G. J. Wound Care 2001, 10, 173−178. (18) Grazul-Bilska, A. T.; Johnson, M. L.; Bilski, J. J.; Redmer, D. A.; Reynolds, L. P.; Abdullah, A.; Abdullah, K. M. Drugs Today 2003, 39, 787−800. (19) Alsousou, J.; Thompson, M.; Hulley, P.; Noble, A.; Willett, K. J. Bone Jt. Surg., Br. Vol. 2009, 91, 987−996. (20) Anitua, E.; Sánchez, M.; Prado, R.; Orive, G. J. Biomed. Mater. Res., Part A 2011, 97, 536. (21) Martins, A.; Pinho, E. D.; Faria, S.; Pashkuleva, I.; Marques, A. P.; Reis, R. L.; Neves, N. M. Small 2009, 5, 1195−1206. (22) Monteiro, N. S.; Martins, A.; Pires, R. A.; Faria, S.; Fonseca, N. A.; Moreira, J. N.; Reis, R. L.; Neves, N. M. Biomaterials Science 2014, DOI: 10.1039/c4bm00069b. (23) Krump-Konvalinkova, V.; Bittinger, F.; Unger, R. E.; Peters, K.; Lehr, H.-A.; Kirkpatrick, C. J. Lab. Invest. 2001, 81, 1717−1727. (24) Sam, S.; Touahir, L.; Salvador, J. A.; Allongue, P.; Chazalviel, J.N.; Laemmel, C.; Djebbar, S. Langmuir 2010, 26, 809−814. (25) Staros, J. V.; Wright, R. W.; Swingle, D. M. Anal. Biochem. 1986, 156, 220−222. (26) Chou, C. H.; Cheng, W. T.; Lin, C. C.; Chang, C. H.; Tsai, C. C.; Lin, F. H. Appl. Biomater. 2006, 77, 338−348. (27) Motoyama, M.; Deie, M.; Kanaya, A.; Nishimori, M.; Miyamoto, A.; Yanada, S.; Adachi, N.; Ochi, M. J. Biomed. Mater. Res., Part A 2010, 92, 196−204. (28) DeLong, S. A.; Moon, J. J.; West, J. L. Biomaterials 2005, 26, 3227−3234. (29) Shen, H.; Hu, X.; Bei, J.; Wang, S. Biomaterials 2008, 29, 2388− 2399. 2205

dx.doi.org/10.1021/bm500346s | Biomacromolecules 2014, 15, 2196−2205