Article pubs.acs.org/Biomac
Interaction of Platelets with Poly(vinylidene fluoride-cohexafluoropropylene) Electrospun Surfaces Furqan Ahmed,† Namita Roy Choudhury,*,† Naba K. Dutta,† Susana Brito e Abreu,† Andrew Zannettino,‡ and Elizabeth Duncan‡ †
Ian Wark Research Institute, University of South Australia, Mawson Lakes Campus, South Australia, Australia Myeloma Research Laboratory, School of Medical Science, University of Adelaide, South Australia, Australia
‡
S Supporting Information *
ABSTRACT: Platelets are the major contributors in the process of thrombosis and in the failure of biomedical implants. A number of factors influence the platelet interaction with foreign surfaces such as surface morphology, surface chemistry, and adsorbed proteins. This study examined the effect of surface topography and chemistry of pristine and fibrinogen-adsorbed solvent cast (SC) and electrospun (ES) samples of poly(vinylidene fluoride-co-hexafluoropropylene) (PVDF-HFP) on platelet adhesion, activation, and aggregation. Qualitative and quantitative studies of fibrinogen adsorption were performed using time-of-flight secondary ion mass spectrometry (ToF-SIMS), while SEM, aggregometry, and liquid scintillation analyses were performed to evaluate platelet adhesion, aggregation, and serotonin release. While little or no platelet adhesion was observed on pristine ES surfaces, considerable adhesion, and measurable aggregation and serotonin release were observed on pristine SC surfaces. Notably, increased adhesion of platelets was observed following fibrinogen adsorption on SC surface with considerable aggregation and serotonin release compared with ES samples, where limited aggregation and platelet adhesion was observed. A further comparison of platelet adhesion, aggregation, and serotonin release was performed with plasma-adsorbed SC and ES surfaces. SC surfaces showed enhanced platelet adhesion, aggregation, and serotonin release compared to ES surfaces. This study shows that the morphology of samples plays a critical role on the biocompatibility of samples by altering the adsorption and adhesion of biomolecules and cells. The low level of adhesion, low aggregation, and serotonin release of platelets, even in the presence of fibrinogen and plasma-derived proteins, suggested that ES samples have the least thrombogenicity.
■
INTRODUCTION The success of blood contacting medical devices is influenced by the adsorption of nonspecific proteins from blood plasma that leads to the growth of endothelial cells, which is desirable for confluent blood flow. Alternately, there may be enhanced adhesion of platelets in response to protein adsorption, leading to thrombus formation and failure of the implant. Most current synthetic materials used as blood contacting medical devices suffer from poor blood compatibility due to abundant platelet adhesion and subsequent clot formation.1,2 The initial events following implantation are critical to the success of the graft. Upon implantation of a medical device, such as small diameter artificial vascular graft (SDAVG), the proteins from the blood plasma adsorb to the surface of device within seconds.3 Among these proteins, fibrinogen, fibronectin, vitronectin, and von Willebrand factors trigger platelet adhesion and aggregation by interacting with receptors on their surfaces.4 The types of plasma proteins and the degree of adsorption will play a critical role in platelet adhesion and is dependent on the nature of material, surface topography, chemistry, and resultant protein conformation. Therefore, the selection of material, scaffold morphology, and chemistry in artificial medical devices will © 2014 American Chemical Society
dictate their biocompatibility, their intrinsic platelet activating properties (hence, thrombus forming properties), and their ultimate success as a graft. Platelets are non-nucleated cell fragments that circulate in the blood in an inactive form and are activated in response to damage to endothelial cell lining or activation of coagulation cascade. Once they adhere to the endothelial lining defect or implant’s (like SDAVG) surface, they change their shape, release alpha granules, and aggregate.5 The circulating platelets interact with plasma proteins (fibrinogen, von willebrand factor, vitronectin, and fibronectin) adsorbed onto the damaged endothelial cell lining or the implant’s surface via GPIb-IXV (a platelet glycoprotein complex to von Willebrand factor exposed in the subendothelium) and αIIbβ3 integrin receptors. Further platelet to platelet interaction/aggregation is supported by fibrinogen RGD sequence, which facilitate more recruitment of platelets during the clot formation process.6 Received: October 17, 2013 Revised: January 28, 2014 Published: January 30, 2014 744
dx.doi.org/10.1021/bm4015396 | Biomacromolecules 2014, 15, 744−755
Biomacromolecules
Article
uniform and consistent fibres. For the process of electrospinning, the 10% polymer solution was placed in a 20 mL glass syringe fitted with a 18 gauge diameter needle. PVDF-HFP nanofibers were fabricated at 13 kV applied voltage using a high voltage power supply. A syringe pump was used to feed polymer solution into the needle tip and the feeding rate of the syringe pump was fixed at 0.15 mL/h. The polymer solution formed a Taylor cone at the tip of the needle by the combined force of gravity and electrostatic charge. A jet formed from the Taylor cone was collected on a grounded copper plate in the form of fibers. The spun nanofibers were dried at room temperature overnight. ES samples were cut into 1 cm circular and 2 × 1 cm rectangular pieces. Fibrinogen Adsorption. A 5 g pack of type 1-S, 65−85% protein (Fibrinogen) was purchased from Sigma Aldrich and dissolved in 5 mL of PBS. The physiological serum concentration of fibrinogen of 2 mg/ mL was chosen for the adsorption studies on SC and ES samples. SC and ES samples were immersed for 2 h in 2 mg/mL of fibrinogen solution at room temperature and rinsed three times with Milli-Q water to remove the nonadsorbed fibrinogen molecules. The pristine samples were immersed in PBS without fibrinogen for 2 h. Characterization of PVDF-HFP Films by ToF-SIMS. Time-offlight secondary ion mass spectrometry (ToF-SIMS) analysis of pristine and fibrinogen-adsorbed PVDF-HFP solvent cast and electrospun surfaces was performed using the PHI TRIFT V nanoToF instrument (Physical Electronics Inc., Chanhassen, MN, U.S.A.). The analysis was conducted using a pulsed liquid metal Au+ primary ion beam operating at 30 kV set for high mass resolution (bunched mode). Positive and negative secondary ion mass spectra were acquired for areas of 100 × 100 μm2 for 1 min, with the ion dose less than 1011 ions/cm2, below the static limit. Dual charge neutralization was used during the analysis using an electron flood gun and Ar+ ions. Raw data files of the analysis were acquired and the spectral data was processed offline using the Win Cadence N software (Physical Electronics Inc., Chanhassen, MN, U.S.A.). The positive spectra were calibrated using the peaks CH3+, C2H5+, and C3H7+ and the negative spectra were calibrated using the CH−, C2H−, and Cl−. The data was normalized to the total ion yield in the spectrum. Multivariate statistical analysis was performed to interpret ToF-SIMS data using principal component (PC) analysis. The data was analyzed using the STATISTICA software package.25 Platelet Harvesting. For platelet adhesion, activation, and aggregation studies, human blood was collected from three healthy donors on three different days by venous phlebotomy. The donors had not taken any medication, including aspirin, for at least 10 days prior to blood collection. Blood was collected with a 19-G butterfly infusion set (at IMVS Hemostasis Laboratory, Royal Adelaide Hospital, Adelaide, Australia) directly into a 30 mL syringe and transferred to heparinized centrifuge tubes. Platelet rich plasma (PRP) was prepared by centrifugation of the blood at 140 g for 8 min at room temperature followed by careful removal of the upper platelet-rich layer (supernatant). Platelet Adhesion. For aggregation and activation studies, blood was collected into acid-citrate dextrose anticoagulant. The platelet adhesion study was performed according to previously described methods.26 The 1 cm diameter discs of SC and ES samples were sterilized with ethanol. Heparinized blood was centrifuged at 180 g for 10 min, and the supernatant was transferred to a fresh tube and constituted the platelet-rich plasma (PRP). The remaining pellet was retained as a source of platelet poor plasma (PPP) for further usage. The PRP was centrifuged at 1258 g for 15 min and the supernatant was discarded. The platelet pellet was resuspended in 10 mL of Tyrode’s (6.4 pH) wash buffer containing 200 μL of apyrase to avoid platelet activation during sample processing. The platelet pellet was washed by centrifuging the sample for 15 min at 1258 g, after which the supernatant was discarded. The washed pellets were resuspended in Tyrodes buffer pH 7.35. Finally, the platelets were counted with the help of Coulter counter and stored in a water bath at 37 °C until required. For the platelet adhesion studies, washed platelets (count of 3.7 × 108 platelets/mL) were spread onto 1 cm pristine, fibrinogen, and
Fibrinogen is one of the most important plasma proteins that mediates adhesion of platelets. During blood plasma interaction with materials, fibrinogen is also adsorbed in higher concentration with other plasma proteins.7,8 Chinn et al.9 observed fibrinogen-deficient, plasma-adsorbed polyetherurethane did not promote adhesion of platelets when compared with a fibrinogen-rich, plasma-adsorbed polyetherurethane surface, demonstrating the role played by fibrinogen on platelet adhesion. Fibrinogen is composed of two sets of three polypeptide chains of Aα, Bβ, and γ, which are joined together via the E domain. The γA chain of the C-terminus of fibrinogen plays a crucial role in the adhesion of platelets to the fibrinogen receptor.10 Besides the conformation of fibrinogen, biomaterial surface chemistry and surface topography are also the major players to define the role of fibrinogen−platelet adhesion.11 The conformational changes of adsorbed fibrinogen on a polymer surface have been studied using the ATR-FTIR flow cell method,12 circular dichroism (CD),13 quartz crystal microbalance with dissipation monitoring (QCM-D),14 and time of flight secondary ion mass spectrometry (ToF-SIMS).15 A recent study by Sprague and colleagues showed a direct relationship between the amounts of fibrinogen adsorbed on biomaterial surfaces with several key responses of the body. Specifically, they demonstrated that platelet and monocyte binding increased with increasing fibrinogen adsorption. Moreover, migration of endothelial cells decreased on surfaces with more adsorbed fibrinogen.16 Notably, apart from the total amount adsorbed, other studies have reported that the conformation of fibrinogen plays a significant role in determining platelet adhesion.17,18 The highly surface sensitive analytical technique time of flight secondary ion mass spectrometry (ToF-SIMS) has been used successfully to study protein adsorption and conformation.19 The principal component analysis (PCA) of secondary ion peaks can successfully provide the information relating to conformation/orientation and denaturation of proteins.20 In order to fabricate an artificial small diameter vascular graft (SDVG), which has the properties of being nonthrombogenic, elastic, resistant to infection, durable and resilient material, solvent cast (SC), and electrospun (ES) surfaces of PVDF-HFP were fabricated. PVDF-HFP has been widely used for filtration membranes,21 wound dressings, antibacterial membranes,22 medical devices,23 and tissue engineering. To understand the blood compatibility/thrombogenecity of PVDF-HFP surfaces, the role of adsorbed fibrinogen on platelet adhesion was investigated in this study. Furthermore, conformational change of fibrinogen adsorbed onto PVDF-HFP surfaces was investigated by ToF-SIMS.
■
MATERIALS AND METHODS
PVDF-HFP, with an inherent viscosity of 2300−2700 Pa and ∼400000 molecular weight, was purchased from Sigma Aldrich, Australia. The polymer was dissolved in a 70/30 ratio of N,N-dimethylacetamide (DMAc) and acetone (from Sigma Aldrich, Australia) at 10% (w/v) concentration and left for overnight at room temperature mixing with a magnetic stirrer bar. Preparation of PVDF-HFP Solvent Cast (SC) Films. Uniform thin films of the PVDF-HFP were prepared by casting of PVDF-HFP solution in a glass petri dish. The samples were kept in a controlled 65 °C environment to dry for 2 days. After drying, the films were cut to 1 cm circular size and 2 × 1 cm rectangular films for adhesion and aggregation assay. Fabrication of Electrospun (ES) Films. An optimized PVDFHFP copolymer24 at a concentration of 10% w/v was found to form 745
dx.doi.org/10.1021/bm4015396 | Biomacromolecules 2014, 15, 744−755
Biomacromolecules
Article
plasma-adsorbed SC and ES samples for 30 min at 37 °C. After 30 min of incubation, SC and ES samples were four times washed with PBS to remove non-adherent cells followed by fixation with 4% glutaraldehyde for 30 min. Samples were postfixed with 2% osmium tetraoxide for 30 min after washing with PBS for 5 min. The samples were dehydrated with a series of 70, 90, and 100% ethanol for 10 min per alcohol concentration. After dehydration, the samples were incubated in 50:50 mixture of hexamethyldisilazane (HMZ) and 100% ethanol for 10 min, followed by incubation in pure HMZ for 10 min and left to dry overnight. After drying, the cellular constructs were sputter coated with platinum and observed under the SEM at an accelerating voltage of 10 kV. Platelet counts were determined on 10 randomly selected SEM images captured from different locations using Image J software. Morphological Characterization of ES and SC Samples by SEM. Philips XL30 field emission gun scanning electron microscopy (FEGSEM) with Oxford CT1500HF Cryo stage was used to characterize the morphology of cells on SC and ES samples with polystyrene as reference. To minimize the charging effect, platinum was deposited after cell fixation on each sample by sputtering and examined at an accelerating voltage of 10 kV. The diameter and porosity of the electrospun fiber mats was optimized as previously described.24 Platelet Aggregation. The aggregation and serotonin release assays were performed simultaneously on the same day with the same donor’s blood. Human blood was harvested from three different donors on three different days and three experimental runs were performed for each donor’s blood sample. Hydroxytryptamine creatinine sulfate (serotonin), 5-[1, 2-3H (N)]-, 1 mCi (37 MBq) was purchased from Perkin-Elmer, Australia. A total of 10 mL of PRP was incubated with 200 μL of diluted serotonin at 37 °C for 45 min, followed by washing. Platelet poor plasma (PPP) was separated from the remaining blood at 2600 g for 15 min to compare aggregation between plasma-adsorbed and fibrinogen-adsorbed SC and ES samples. For the aggregation study, an AggRAM platelet aggregrometer (Helena Laboratories, Beaumont, TX, U.S.A.) was used to determine the % aggregation. Data was analyzed in the form of % aggregation with baseline correction, with the slope of the aggregation curve recorded for each test. The platelet aggregation was assessed on pristine (SC0 and ES0), fibrinogen-adsorbed (SCFg and ESFg), and plasma-adsorbed (SCPl and ESPl) sample sleeves inserted into the silicone cuvette containing magnetic fleas. Collagen (10 μg/mL), arachidonic acid (AA, 1.65 μM), and thrombin receptor activating peptide (TRAP, 30 μM) agonists were used as positive controls (without SC and ES surfaces and with agonist). As shown in Figure 1, a small gap between the magnetic flea and the bottom of the SC and ES sleeves allowed the transmission of light by the photocell during centrifugation. Serotonin-labeled platelets (540 μL) were added to the silicon cuvettes containing SC0, ES0, SCFg, and ESFg samples. There were a maximum of eight channels in the instrument to read aggregation in the cuvette at a time. Each channel was started with a gap of 10 s, followed by the addition of 60 μL of Tyrode’s buffer in sample’s cuvette and 60 μL of agonists in positive control (without sample and with an agonist) tubes. Light transmission was measured for 30 min during the process of aggregation then 160 μL of EDTA was added to stop further reaction. Serotonin Release Assay. Following completion of the aggregation assay, the serotonin-labeled platelet solution was transferred from the silicon cuvette to an ependorff tube for centrifugation at 18407 g for 1 min to enable measurement of radiolabeled-serotonin released from the platelets during the activation process. A total of 100 μL of the supernatant was dissolved in 2 mL of Microscint 20 scintillation fluid (purchased from PerkinElmer, Australia). The samples were then counted using a TRI-CARB 2900 TR liquid scintillation analyzer and Quanta Smart software (Perkin-Elmer). Tritiated (3H)-serotonin (without platelet) was also evaluated in the liquid scintillation counter as a control. The total 3H serotonin counts were then used to calculate the % release of serotonin from the adherent cells using the following equation.
Figure 1. Placement of sleeve in silicon cuvette for aggregation study. %release =
release value − background × 100 total − background
(1)
where “release value” is the count per minute in region A (CPMA). A background sample was prepared separately by adding 540 μL of 3H serotonin-labeled platelet, 60 μL of Tyrode’s buffer, 160 μL of EDTA without incubating in silicon cuvette for aggregation, and then centrifuged at 18407 g for 1 min. A reference total serotonin count for sample was prepared, which is related to the same recipe of background sample but without centrifugation to get total count of radioactive serotonin. Data Analysis. Statistical comparisons from different cell counts were performed using Microsoft Excel and data were presented as mean ± standard deviation (SD). Differences were considered statistically significant with
