Article pubs.acs.org/JPCB
The Attachment Affinity of Hemoglobin toward Silver-Containing Bioactive Glass Functionalized with Glutaraldehyde C. Gruian,† A. Vulpoi,† E. Vanea,† B. Oprea,† H.-J. Steinhoff,‡ and S. Simon*,† †
Faculty of Physics & Institute of Interdisciplinary Research in Bio-Nano-Sciences, Babes-Bolyai University, Cluj-Napoca 400084, Romania ‡ Physics Department, University of Osnabrück, Osnabrück 49069, Germany ABSTRACT: Bioactive glasses belonging to the 56SiO2·(40 − x)CaO·4P2O5·xAg2O system, with x = 0, 2, and 8 mol %, were surface functionalized with the protein coupling agent glutaraldehyde (GA) and further evaluated in terms of hemoglobin affinity. The bare and GA-functionalized samples were investigated before and after protein attachment, by electron paramagnetic resonance (EPR) spectroscopy combined with spin-labeling procedure. Methanethiosulfonate spin label was used to explore the local environment of β-93 cysteine in horse hemoglobin, in terms of spin label side chain mobility. The EPR simulation methods were employed to quantify the rotational correlational times and fraction of the immobilized spin labels. The EPR absorption spectrum was further exploited to estimate the amount of hemoglobin loaded on the substrates. The surface elemental composition obtained by X-ray photoelectron spectroscopy revealed similar tendency in terms of surface coverage. Changes in surface architecture, that is, changes in surface morphology after protein coverage, were observed by scanning electron microscopy. It was concluded that GA improves the stability of protein attachment and induces polymerization of hemoglobin molecules.
■
INTRODUCTION In the last decades, significant progress has been reported in the field of biomaterials used for implants. Considerable attention has been directed to bioactive glasses (BGs) used as bone graft substitutes, because of their ability to form in vivo a bone-like apatite layer capable of promoting the osteointegration.1 Two important issues concerning the success of bone implant materials consist in the lack of any cytotoxic response and bioactive behavior. However, one of the most serious problems related to implant materials is the infection that can develop after the surgery treatment.2,3 Moreover, the infections that may appear are often resistant to antibiotics, the healing solution being the implant removal.1 The well-known antibacterial effect of silver has been exploited in wound healing and in biomedical applications to prevent infections. Silver-containing BGs can therefore be considered to be suitable implant materials that combine the bioactivity and biocompatibility of SiO2−CaO− P2O5-based systems with the antibacterial properties of silver. Recent studies performed in our group have shown that high amounts of silver (up to 8%) do not diminish the bioactivity of calcium−phospho−silicate BGs.4,5 One important aspect to consider in investigating the biocompatibility of a biomaterial is the response in terms of interaction with proteins. The protein-formed layer on a biomaterial surface provides the topographical and chemical cues to guide cells and increases the biomaterial ability to support and foster cells attachment.6 In this respect, one approach is to load inorganic substrates, for example, ceramics and porous glasses, with biomolecules, for example, proteins, to stimulate cell adhesion and proliferation onto the material surface.7 Surface © 2013 American Chemical Society
modification with a protein coupling agent, for example, glutaraldehyde (GA), represents a possible approach to provide accessible and chemical functional groups for protein immobilization8 and to attach proteins without losing their conformational functionality.9−11 A previous study performed on 45S5 Bioglass scaffolds has shown that the surface functionalization with 3-aminopropyltriethoxysilane (APTS) and GA favors the formation of hydroxyapatite upon immersion in simulated body fluid.12 To date, only a few studies investigated the effect of GA on protein immobilization onto BGs.13,14 Therefore, the present study aims to investigate the influence of GA in adsorption of spin-labeled hemoglobin, as model protein, onto sol−gel derived silvercontaining BGs of the 56SiO2·(40 − x)CaO·P2O5·xAg2O system (0 ≤ x ≤ 8 mol %). Since the amino acid sequence and the structure of human and horse hemoglobin display a large similarity,15,16 in this study the more common horse hemoglobin was considered. This protein is a globular metalloprotein found in blood cells that carry the oxygen from the lungs to the peripheral tissues. It consists of four polypeptide chains: two α (141 amino acids each) and two β chains (146 amino acids each). Each β chain contains one cysteine residue in position β-93 (Figure 1), which was considered in this study for spin labeling. The method of spin labeling in combination with EPR spectroscopy was predominantly used for studying the structure and dynamics of soluble proteins in solution or membrane Received: September 3, 2013 Revised: December 5, 2013 Published: December 5, 2013 16558
dx.doi.org/10.1021/jp408830t | J. Phys. Chem. B 2013, 117, 16558−16564
The Journal of Physical Chemistry B
Article
deionized water, immersed for 1h in GA [CH2(CH2CHO)2] solution (1 mol/L) at room temperature, and finally washed again in deionized water.12 Spin Labeling. Lyophilized horse hemoglobin was purchased from Sigma-Aldrich, and a protein solution was prepared in phosphate buffer (0.01 M, pH 7.4). Afterward, the protein was incubated with 10 mM dithiothreitol (DTT) at 4 °C for 2 h, in order to reduce the cysteine thiol groups. DTT was removed by repeated dilution steps with sodium phosphate buffer, pH 7.4, using centrifugal filter units with 30 kDa molecular weight cutoff (Amicon/Millipore, Carringtwohill, Co. Cork, Ireland). For spin labeling the protein solutions were incubated for 16 h at 4 °C with 6-fold molar excess of (1-oxyl-2,2,5,5-tetramethylpyrroline3-methyl)methanethio-sulfonate spin label (MTS) (Toronto Research; Alexis Biochemicals). Unbound MTS was removed by repeated ultrafiltration as described above. Labeling efficiency (spin label per β-93 cysteine) was estimated to 40%. Protein Adsorption. The powder samples were incubated at 37 °C, for 4 h, in a solution of 150 μM (∼10 mg/mL) horse hemoglobin in phosphate buffer solution (PBS) (0.01 M, pH 7.4) with 10 mM NaCl. After incubation, all samples were washed three times with PBS to remove the protein molecules that were unbound and/or detached from the BG surface. In this order, after sample sedimentation, the entire volume of supernatant was exchanged with 1 mL of fresh PBS and afterward intensively pipetted to extract all the unbound protein molecules from the matrix. This procedure was repeated three times. After each washing step, the supernatant collected was examined by EPR spectroscopy to verify if it contained any protein. The last collected supernatant was shown to be proteinfree solution.The washed hemoglobin-loaded powders dispersed in PBS were subjected to a detailed investigation by EPR spectroscopy. Finally, to verify the stability of protein attachment, all samples were ultrasonicated for 45 min at room temperature and then washed again with buffer solution to remove the protein detached from the surface. Afterward, the samples were examined again by EPR spectroscopy. For SEM and XPS measurements, the wet samples collected before ultrasonication were lyophilized for 24 h in an Alpha 1−2 LD type freeze dryer at 217 K and 0.05 mbar. cw-EPR Measurements. X-band cw-EPR experiments were performed at room temperature, using a homemade EPR spectrometer equipped with a Bruker dielectric resonator. The microwave power was set to 1.0 mW; the B-field modulation amplitude was 0.15 mT. Glass capillaries of 0.9 mm inner diameter were filled with about 15 μL of samples (the EPR active volume of the sample tube was 10 μL). The integrated EPR absorption signal is directly proportional to the spin concentration in the sample and was used to calculate the amount of the protein attached on the BG. As reference spin probe, 2,2,6,6tetramethyl-1-piperidinyloxy (TEMPO) (100 μM) was used. The EPR spectra simulations were performed using the slowmotion approach.19 This program allows a nonlinear leastsquares (NLLS) fitting of a single nitroxide EPR spectrum with two components having different mobility and magnetic tensor parameters. The values of the magnetic g tensor for both components (denoted here the mobile component M and immobile component I) were fixed to the values which yielded best fit results for all spectra: gxx (I) = 2.00900; gxx (M) = 2.00850; gyy (I) = 2.00650; gyy (M) = 2.00670; and gzz (I) = gzz (M) = 2.00230. Similarly, the components Axx and Ayy of the hyperfine tensor for the two spectral components were fixed to the following values: Axx (I) = 6.00; Axx (M) = 6.30; Ayy (I) =
Figure 1. Detailed view of the spin-label binding site in horse hemoglobin (2ZLU from Protein Data Bank). The α chains are colored in green, and the β chains are colored in blue. The spin-label side chain bound at position β-93 is highlighted in stick representation (black). The tyrosine (β-145) (red) situated in a pocket close to the C-terminus of the β chain is able to reorient to allow the nitroxide ring to enter the pocket. The heme group (gray) harbors the iron atom (brown) in its center.
proteins reconstituted in their native membrane environment.17 Recent studies performed in our group reported that this method can be extended to proteins encapsulated18 or adsorbed on solid surfaces5,13 to investigate conformational changes induced at the level of the backbone fold. In the present study the mobility of spin labels (SL) was characterized based on the values obtained from EPR simulations for the SL rotational correlation times and for the fraction of immobilized SL. From this we determined the hemoglobin immobilization upon interaction with the BG surface. Moreover, from the intensity of the continuous wave (cw) EPR spectra, we could estimate the molar concentration of the adsorbed hemoglobin. The obtained values were further correlated with the results obtained by X-ray photoelectron spectroscopy (XPS) and specific surface area analysis, concerning the protein surface coverage. Images obtained by scanning electron microscopy (SEM) provided information about the protein distribution on the BG surfaces.
■
MATERIALS AND METHODS Bioactive Glass Preparation. Samples belonging to the system of 56SiO2·(40 − x)CaO·4P2O5·xAg2O, with x = 0, 2, and 8 mol %, were prepared via the sol−gel method, which has been reported elsewhere.4 Briefly, the samples were synthesized using tetraethyl orthosilicate (TEOS), calcium nitrate tetrahydrate (Ca(NO 3 ) 2 ·4H 2 O), ammonium phosphate dibasic ((NH4)2HPO4), and silver nitrate (AgNO3). The molar ratio of EtOH/TEOS was 1:1. The other precursors were previously dissolved in distilled water and added to the TEOS solution. The pH of the final solution was adjusted to 2 by using HNO3. The obtained gels were aged for 7 d at room temperature. The dried gels were heat treated at 580 °C for 30 min and ground/ powdered to particles of micrometric size. Prior to protein attachment, the BG particles were silanized with APTS and surface functionalized with GA. First, the BG was immersed for 4 h in an aqueous H2N(CH2)3Si(OC2H5)3 [APTS] solution (0.45 mol/L, pH adjusted to 8 by adding 1 M HCl) at 80 °C. After 4 h the sample was collected, washed in 16559
dx.doi.org/10.1021/jp408830t | J. Phys. Chem. B 2013, 117, 16558−16564
The Journal of Physical Chemistry B
Article
7.20; and Ayy (M) = 6.40. Azz was allowed to be optimized by the NLLS analysis. For the mobile component, we used the Brownian rotational diffusion in an isotropic solvent, with spherical symmetric rotational diffusion tensor; anisotropy of the rotational diffusion tensor was considered only for the immobile component, assuming that this component is more affected by the interaction with the BG substrate. Thereby, the isotropic component of the rotational diffusion tensor R1 was optimized by the fittings for both spectral components. In addition and only for the immobile spectral component, the axial component of the rotational diffusion tensor R2 was allowed to vary for best fitting, whereas the rhombic component log (R3) was fixed at 2.08 for all samples. Specific Surface Area Analysis. The surface area and average pore volume were determined by measuring nitrogen adsorption/desorption at 77 K with Qsurf Series M1 surface area analyzer, on the basis of the Brunauer, Emmett, and Teller (BET) method. For all samples, the measurements were performed under similar conditions using one point BET method with N2/ He gas mixtures (30:70%). XPS Measurements. The XPS measurements carried out to analyze the surface functionalization of BG microparticles with protein were performed using a SPECS PHOIBOS 150 MCD system equipped with monochromatic AlKα source (200 W, hυ = 1486.6 eV), a hemispherical analyzer, and a multichannel detector. The pressure in the analysis chamber during the measurements ranged from 10−9 to 10−10 mBar. Charge neutralization was used for all samples. The binding energy scale was charge referenced to the C 1s at 284.6 eV. Elemental compositions were determined from spectra acquired at a pass energy of 100 eV. High-resolution spectra were obtained using an analyzer pass energy of 30 eV. The curve-fitting analysis of N 1s was performed by using the Gaussian curve-fitting function in CasaXPS software. SEM Measurements. The SEM images were recorded using FEI Quanta 3D FEG 200/600 scanning electron microscope. The sample powders were covered with platinum (Pt) to amplify the secondary electron’s signal. This coating was performed in an Agar automatic sputter coater at standard atmospheric pressure (1013 mBar).
Figure 2. (a) The experimental EPR spectrum of an MTS spin label covalently bound to the sulfhydryl groups of horse hemoglobin in position β-93 is composed of two components that are shown in panel b. (b) The immobile I and mobile M components were calculated based on the slow motion approach (see Methods). The weight of each spectral component is given in percents. Fitting parameters are listed in Table 1.
After the I and M are attached to the BG substrates, the ratio between the two spectral components weights even more in favor of the strongly immobilized component. The fraction of strongly immobilized spin probes f(I) and the line width of the I component are significantly increased (Figure 3), reflecting the
■
RESULTS AND DISCUSSION The cw-EPR spectrum of the MTS spin-labeled horse hemoglobin in solution is shown in Figure 2a. The EPR spectrum is composed of two spectral components, associated with two populations of spin labels with different motional restrictions. The strongly immobilized component will be further labeled as “the immobile component” I and originates from spin labels entrapped in the so-called tyrosine pocket20 (Figure 1). The more mobile component, denoted here as “the mobile component” M, arises from spin labels that are oriented toward the surface of the protein molecule, allowing thus relatively free reorientational motion of the nitroxide ring.20 The two spectral components that overlap in the experimental spectrum were separated by spectrum simulations, as depicted in Figure 2b. The immobile component I is found to represent the larger fraction (59.7%) of the spectrum. Moreover, information on the polarity of the nitroxide surroundings can be obtained from the hyperfine splitting of the mobile component M.17 An enhanced solvent polarity leads to an increased hyperfine splitting component Azz of the nitroxide, because of a shift of the spin density toward the nitrogen atom. In addition, hydrogen bonds formed to the oxygen atom of the nitroxide radical influence gxx.21
Figure 3. Experimental (solid lines) and simulated (small open circles) EPR spectra of horse hemoglobin spin-labeled at position β-93, recorded in solution (a) and in the adsorbed state immediately after immersion on the pristine (gray) and GA-functionalized (black) BGs with different silver content. (b) 0% Ag. (c) 2% Ag. (d) 8% Ag. The numbers in the left column represent the fraction of the immobile component.
immobilization of the whole protein molecule on the BG substrates.13 Previous studies have shown that the silver content in the BG matrix also influences hemoglobin immobilization, as a consequence of interaction between the unlabeled thiol groups in the protein and the Ag+ ions.5,22 This hypothesis is also sustained by the fractions of immobilized spin labels and the rotational 16560
dx.doi.org/10.1021/jp408830t | J. Phys. Chem. B 2013, 117, 16558−16564
The Journal of Physical Chemistry B
Article
Table 1. The Values of Hyperfine Tensor Component Azz and the Logarithm of the Rotational Diffusion Tensor (R) Componentsa BG/BG−GA x (mol %) 0
2
8
sample →
component
horse hemoglobin
BG
BG−GA
BG
BG−GA
BG
BG−GA
Azz (mT)
I M I M I M I M I M
36.04 39.68 7.02 8.13 0.85 0 2.08 0 15.7 1.24
36.44 39.60 6.84 8.13 0.66 0 2.08 0 23.8 1.25
37.24 39.66 6.83 8.07 0.89 0 2.08 0 24.8 1.5
36.94 39.40 6.75 8.01 0.66 0 2.08 0 29.7 1.61
37.28 39.61 6.82 8.07 0.63 0 2.08 0 25.1 1.5
37.28 38.93 6.74 7.99 0.69 0 2.08 0 30.4 1.72
37.34 39.54 6.80 8.03 0.71 0 2.08 0 26.6 1.6
R1 R2 R3 τC (ns)
Values obtained from EPR spectra simulations for horse hemoglobin in solution, attached to 56SiO2·(40 − x)CaO·4P2O5·xAg2O BG or on functionalized BG (BG−GA) substrates with different silver content. The values are presented for the immobile I and the mobile M spectral components. The rotational correlation time τc is calculated from the diffusion rates (DR) according to τc = 1/(6 × DR), where DR was derived from the component R1 of the rotational diffusion tensor.16 Errors are estimated to be below 2% for Azz obtained from the fittings. a
Table 2. Molar Concentration (in μM) of Hemoglobin Attached on Pristine (BG) and Functionalized (BG−GA) Bioactive Glass with Different Silver Concentrationsa
a
Errors due to experimental settings and uncertainties in baseline subtraction are estimated to be ±5%.
functionalized samples, supporting the assumption that the protein coupling agent diminishes the influence of silver in hemoglobin adsorption. All the above results are in good agreement with the quantitative adsorption analysis. The intensities of the cw-EPR spectra reveal a higher amount of hemoglobin attached on the pristine samples comparing the GA-functionalized ones and also an increase of protein concentration with silver content (Table 2). Although functionalization with GA results in a lower quantity of adsorbed hemoglobin, the larger amount of protein remaining on the samples after ultrasonication revealed an enhanced stability of protein attachment (Table 2). GA offers more stable and specific sites for hemoglobin attachment on a BG substrate, so protein binding and distribution might be more organized on the GA-functionalized samples. This statement is further confirmed by the specific surface area analysis performed on the BG samples before protein attachment; although the functionalization with GA increases the specific surface area of all samples (Table 3), the amount of bound protein decreases. Hence, we conclude that the protein layer assembled on this type of substrate does not cover the surface randomly but binds only to specific sites on the surface.
correlation times obtained from simulations of the cw-EPR spectra (Figure 3 and Table 1). The fraction of immobilized spin labels increases from 59% (hemoglobin in PBS solution) to 73%, 84%, and 91% for hemoglobin attached on the system with 0%, 2%, and 8% silver content, respectively (Table 1). The rotational correlation times increased also upon hemoglobin attachment, the highest value being obtained when the protein is attached on the substrate with 8% silver content (Table 1). In addition, the decrease in polarity observed for M with increasing silver concentration in the case of pristine samples (see the Azz values in Table 1) could suggest that few rotamers of the spin label have access to the solvent phase, so the protein layer is more crowded on the BG with higher silver amount. Although the differences between the Azz values are minor and close to the error limit, the observed restriction of the mobility is strong evidence for the interaction of the spin labels with neighboring protein atoms or the material surface and, accordingly, the nitroxides will experience a reduced accessibility for water. The silver effect is less pronounced for hemoglobin attached on the GA functionalized BG substrates (Figure 3, black lines), suggesting that GA conceals the silver ions that are disposed on the BG surface, and thus the main interaction takes place between the protein and the GA layer from the material surface. Likewise, the Azz values are nearly constant for all 3 GA 16561
dx.doi.org/10.1021/jp408830t | J. Phys. Chem. B 2013, 117, 16558−16564
The Journal of Physical Chemistry B
Article
the protein surface. Furthermore, as can be observed in Figure 4 (lower row), significant differences appear in the distribution of the protein polymer layer on substrates with different silver content. As the silver content increases, the protein assemblies are shorter and more uniformly distributed. To complement the results obtained by EPR spectroscopy, all samples were investigated by XPS spectroscopy, which is one of the best-suited techniques for studying the surface of proteinfunctionalized materials.27 The XPS elemental analysis based on the survey spectra further support the obtained results (Table 4). The XPS elemental composition of the pristine samples reveals the presence of silicon, calcium, phosphorus, and oxygen. Carbon appears also, this element being detected on all samples exposed to the atmosphere28 and used as reference for charge compensation and binding energy scale calibration to C 1s at 284.6 eV. After functionalization with APTS/GA, higher concentrations of carbon and nitrogen were detected, as these elements enter the composition of APTS and GA, while the contribution of calcium and phosphorus is smaller (Table 4). This new elemental distribution confirms that the surface was successfully functionalized. Furthermore, the elemental composition on the surface of the protein-coated substrates, summarized in Table 4, reveals that all substrates are covered by a protein layer that increases with the silver amount in the sample. The bare substrates covered with protein show the presence of sulfur in small amounts. This element is contained in hemoglobin as it has two thiol groups in each monomer. In addition iron is visible in the sample containing 8% silver. These two elements prove that a higher amount of hemoglobin is attached to the pristine samples than it is on the GA-functionalized ones, supporting thus the trend found for the concentrations determined from cw-EPR spectra (Table 2). The amount of adsorbed protein was quantified following the evolution of the C 1s and N 1s photoelectron peaks. The amount of carbon and nitrogen significantly increases for all samples upon immersion in protein solution (Table 4). For all substrates functionalized with GA, the XPS data show the
Table 3. The Specific Surface Area Values Determined by the Brunauer, Emmet, and Teller (BET) Method for the 56SiO2· (40 − x)CaO·4P2O5·xAg2O Bioactive Glasses with Different Silver Content, before (BG) and after (BG−GA) Functionalization with Glutaraldehyde specific surface area (m2/g) sample
x=0
x=2
x=8
BG BG−GA
44 63
28 30
6 9
On the other hand, hemoglobin covers every accessible space on the surface of the pristine samples. Protein molecules having weaker attachment points can be easily detached during ultrasonication,13 as hemoglobin adsorption implies both reversible and irreversible adsorption operating in parallel.23 A closer look at the quantitative adsorption values obtained after the ultrasonication procedure reveals that the silver content influences also the protein binding stability. As the silver concentration increases, the stability of the binding decreases, and therefore less protein remains on the pristine samples with silver content (after ultrasonication 43%, 33%, and 31% of the initially adsorbed protein remained attached on the pristine BG with 0, 2, and 8% silver content, respectively). This behavior can also be explained considering that protein agglomeration around the silver ions5 leads to weaker bonds of protein molecules to the BG substrate. The SEM images show a more regular distribution of the protein layer on the substrates functionalized with GA than on the pristine samples (Figure 4). In fact, the regular protein assemblies observed on the GA-functionalized samples reveal another interesting aspect: hemoglobin polymerization due to the presence of GA. These repetitive structures have about 50 nm widths, which could correspond to clusters of 8−10 protein molecules. This phenomenon was also observed in other studies that have shown the cross-linking of hemoglobin in solution24−26 or on BGs surfaces13 in the presence of GA, because of interaction of the protein coupling agent with the lysines from
Figure 4. Surface morphology of protein-loaded bioactive glass substrates with different silver content (shown in the bottom of each image), according to SEM images. (upper row) Pristine samples. (lower row) GA-functionalized samples (scale bars = 500 nm). The clusters of protein molecules formed on GA-functionalized samples are highlighted by white squares; for each such sample, a zoom into defined structural motifs is presented in the inset. 16562
dx.doi.org/10.1021/jp408830t | J. Phys. Chem. B 2013, 117, 16558−16564
The Journal of Physical Chemistry B
Article
Table 4. Elemental Composition Obtained from XPS Analysis of the Bioactive Glass (BG) and GA-Functionalized Bioactive Glass (BG−GA) with Different Silver Content, before and after Hemoglobin (Hgb) Adsorption elemental composition (atom %) sample
x (mol %)
Si
Ca
P
BG BG Hgb BG−GA BG−GA Hgb BG BG Hgb BG−GA BG−GA Hgb BG BG Hgb BG−GA BG−GA Hgb
0
21.8 13.5 16.8 10.2 25.2 12.3 18.7 9.9 21.7 15.5 14.6 13.6
13.5 12.6 7.8 10.6 12.2 9 5.8 7.2 9.7 4.3 5.4 4.1
1.9 6.7 2.2 3.8 2.3 4.8 1.3 3.2 4.2 2.3 3 1.5
2
8
presence of nitrogen before protein adsorption. In this case, the N 1s photoelectron peak appears at a binding energy of 398.7 eV (Figure 5) and corresponds to nitrogen from the amino group in APTS. However, upon protein adsorption, the high resolution N 1s photoelectron peaks comprise two components centered at
Ag
O
C
1.1 1.1 1 0.8 5.8 3.1 3 2.7
49.9 49.3 41.5 40.9 52.2 43.4 39.2 35.3 44.3 37.8 34.7 31.3
12.9 16 29.5 31.6 7 26 32.1 37.5 14.3 27.9 34.2 39.1
N
S
1.2 2.2 2.9
0.7
2.6 1.9 6.1
0.8
5 5.1 7.7
1.4
Fe
2.6
399.3 and 397.2 eV (Figure 5), characteristic of N−CH2 bonds that are typical for nitrogen in proteins.29−31 This assessment is supported also by the decrease in oxygen, silver, and silicon concentration on the surface, which is another proof of protein coverage (Table 4). All XPS data confirm that for all samples the amount of adsorbed protein follows the same pattern evidenced by the cwEPR spectra, namely, it increases with the increase of silver content, the influence of silver being less obvious for the samples functionalized with GA (Table 4).
■
CONCLUSIONS The interaction of silver-containing BGs with hemoglobin was clearly evidenced by EPR and XPS spectroscopy. The results show that silver enhances the amount of adsorbed protein, although the stability of the protein binding is not increased. This behavior may be considered a disadvantage in case of BGs designed for implant devices and might be overcome by the use of a protein-coupling agent. All results obtained in this study indicate that GA improves the stability of the protein attachment, although it generally reduces the amount of the adsorbed protein. Even if the BG surface is covered by GA, silver has an influence on protein adsorption, and more protein is attached on the substrate with higher silver content. Apparently, the adsorption process does not occur randomly on the substrate functionalized with GA, but the protein binds to specific sites on the substrate surface. Interestingly, GA also leads to the cross-linking of hemoglobin, as revealed by SEM analysis.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Tel.: +40-264-405300. Fax: +40-264-591906. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The research was accomplished in the framework of PNII Idei PCCE-129/2008 project granted by the Romanian National University Research Council. Support by the DAAD programme “Ostpartnerschaften” is gratefully acknowledged. B. Oprea acknowledges financial support from the Sectorial Operational Program for Human Resources Development 2007−2013, cofinanced by the European Social Fund, under the Project
Figure 5. Deconvolution of XPS N 1s high resolution spectra of GAfunctionalized bioactive glass with different silver content x. (a) Before and (b) after immersion in protein solution (x = 0). (c) Before and (d) after immersion in protein solution (x = 2). (e) Before and (f) after immersion in protein solution (x = 8). 16563
dx.doi.org/10.1021/jp408830t | J. Phys. Chem. B 2013, 117, 16558−16564
The Journal of Physical Chemistry B
Article
Number POSDRU/107/1.5/S/76841 with the title “Modern Doctoral Studies: Internationalization and Interdisciplinary”.
■
Polarity and Proticity Effects in Spin-Labeled Proteins Studied by Multifrequency EPR. Appl. Magn. Reson. 2010, 37, 391−403. (22) Rogers, K. S. Variable Sulfhydryl Activity toward Silver Nitrate by Reduced Glutathione and Alcohol, Glutamate and Lactate Dehydrogenases. Biochim. Biophys. Acta 1972, 263, 309−314. (23) Everest, M. A.; Black, V. M.; Haehlen, A. S.; Haveman, G. A.; Kliewer, C. J.; Neill, H. A. Hemoglobin Adsorption to Silica Monitored with Polarization-Dependent Evanescent-Wave Cavity Ring-Down Spectroscopy. J. Phys. Chem. B 2006, 110 (39), 19461−19468. (24) Bobofchak, K.; Tarasov, E.; Olsen, K. W. Effect of Cross-Linker Length on the Stability of Hemoglobin. Biochim. Biophys. Acta 2008, 1784 (10), 1410−1414. (25) Guillochon, D.; Vijayakshmi, M. W.; Thiam-Sow, A.; Thomas, D. Effect of Glutaraldehyde on Hemoglobin: Functional Aspects and Mössbauer Parameter. Biochem. Cell. Biol. 1986, 64, 29−37. (26) MacDonald, S. L.; Pepper, D. S. Hemoglobin Polymerization. Methods Enzymol. 1994, 231, 287−308. (27) Vanea, E.; Simon, V. XPS Study of Protein Adsorption onto Nanocrystalline Aluminosilicate Microparticles. Appl. Surf. Sci. 2011, 257, 2346−2352. (28) Michel, R.; Pasche, S.; Textor, M.; Castner, D. G. Influence of PEG Architecture on Protein Adsorption and Conformation. Langmuir 2005, 21, 12327−12332. (29) Serro, A. P.; Gispert, M. P.; Martins, M. C. L.; Brogueira, P.; Colaco, R.; Saramago, B. Adsorption of Albumin on Prosthetic Materials: Implication for Tribological Behavior. J. Biomed. Mater. Res., Part A 2006, 78A, 581−589. (30) Vanea, E.; Magyari, K.; Simon, V. Protein Attachment on Aluminosilicates Surface Studied by XPS and FTIR Spectroscopy. J. Optoelectron. Adv. Mater. 2010, 12, 1206−1212. (31) Vanea, E.; Tamasan, M.; Albon, C.; Simon, V. Synthesis and Characterization of a New Composite Aluminosilicate Bioceramic. J. Non-Cryst. Solids 2011, 357, 3791−3796.
REFERENCES
(1) Balagna, C.; Vitale-Brovarone, C.; Miola, M.; Verne, E. Antibacterial Effect of Silver Doped 3D-GlassCeramic Scaffolds for Bone Grafting. J. Biomater. Appl. 2011, 25 (6), 595−617. (2) Green, G. W.; Wenzel, R. P. Postoperative Wound Infections: A Controlled Study of the Increased Duration of Hospital Stay and Direct Cost of Hospitalization. Ann. Surg. 1977, 185, 264−268. (3) Whitehouse, J. D.; Friedman, N. D.; Kirkland, K. B.; Richardson, W. J.; Sexton, D. J. The Impact of Surgical-Site Infections following Orthopaedic Surgery at a Community Hospital and a University Hospital, Adverse Quality of Life, Excess Length of Stay, and Extra Cost. Infect. Control Hosp. Epidemiol. 2002, 23, 183−185. (4) Vulpoi, A.; Baia, L.; Simon, S.; Simon, V. Silver Effect on the Structure of SiO2CaOP2O5 Ternary System. Mater. Sci. Eng., C 2011, 32, 178−183. (5) Vulpoi, A.; Gruian, C.; Vanea, E.; Baia, L.; Simon, S.; Steinhoff, H. J.; Göller, G.; Simon, V. Bioactivity and Protein Attachment onto Bioactive Glasses Containing Silver Nanoparticles. J. Biomed. Mater. Res., A 2012, 100 (5), 1179−1186. (6) Koening, A. L.; Grainger, D. W. Method of Tissue Engineering; Academic Press: San Diego, CA, 2002. (7) Mathieu, H. J. Bioengineered Material Surfaces for Medical Applications. Surf. Interface Anal. 2001, 32, 3−9. (8) Yan-Peng, Jiao; Fu-Zhai, Cui. Surface Modification of Polyester Biomaterials for Tissue Engineering. Biomed. Mater. 2007, 2, R24−R37. (9) Heule, M.; Rezwan, K.; Cavalli, L.; Gauckler, L. J. A Miniaturized Enzyme Reactor Based on Hierarchically Shaped Porous Ceramic Microstruts. Adv. Mater. 2003, 15, 1191−1194. (10) Weetall, H. H. Methods in Enzymology; Academic Press: New York, U.S.A., 1976. (11) Williams, R. A.; Blanch, H. W. Covalent Immobilization of Protein Monolayers for Biosensor Applications. Biosens. Bioelectron. 1994, 9, 159−167. (12) Chen, Q. Z.; Rezwan, K.; Armitage, D.; Nazhat, S. N.; Boccaccini, A. R. The Surface Functionalization of 45S5 Bioglass-Based Glass Ceramic Scaffolds and its Impact on Bioactivity. J. Mater. Sci.: Mater. Med. 2006, 17, 979−987. (13) Gruian, C.; Vulpoi, A.; Steinhoff, H. J.; Simon, S. Structural Changes of Methemoglobin after Adsorption on Bioactive Glass, as a Function of Surface Functionalization and Salt Concentration. J. Mol. Struct. 2012, 1015, 20−26. (14) Gruian, C.; Vanea, E.; Simon, S.; Simon, V. FTIR and XPS Studies of Protein Adsorption onto Functionalized Bioactive Glass. Biochim. Biophys. Acta 2012, 1824, 873−881. (15) Baldwin, J. M. The Structure of Human Carbonmonoxy Haemoglobin at 27 Å Resolution. J. Mol. Biol. 1980, 136, 103−128. (16) Kan, H.-I.; Chen, I.-Y.; Zulfajri, M.; Wang, C. C. Subunit Disassembly Pathway of Human Hemoglobin Revealing the SiteSpecific Role of Its Cysteine Residues. J. Phys. Chem. B 2013, DOI: 10.1021/jp402292b. (17) Klare, J. P.; Steinhoff, H. J. Spin Labeling EPR. Photosynth. Res. 2009, 102, 377−390. (18) Vanea, E.; Gruian, C.; Rickert, C.; Steinhoff, H.-J.; Simon, V. Structure and Dynamics of Spin-Labeled Insulin Entrapped in a Silica Matrix by the Sol−Gel Method. Biomacromolecules 2013, 14 (8), 2582− 2592. (19) Budil, D. E.; Lee, S.; Saxena, S.; Freed, J. H. NonlinearLeastsquares Analysis of Slow-Motional EPR Spectra in One and Two Dimensions Using a Modified LevenbergMarquardt Algorithm. J. Magn. Reson., Ser. A 1996, 120, 155−189. (20) Moffat, J. K. Spin-Labelled Haemoglobins: A Structural Interpretation of Electron Paramagnetic Resonance Spectra based on X-ray Analysis. J. Mol. Biol. 1971, 55, 135−146. (21) Bordignon, E.; Brutlach, H.; Urban, L.; Hideg, K.; Savitsky, A.; Schnegg, A.; Gast, P.; Engelhard, M.; Groenen, E. J. J.; Möbius, K.; Steinhoff, H. J. Heterogeneity in the Nitroxide Micro-Environment: 16564
dx.doi.org/10.1021/jp408830t | J. Phys. Chem. B 2013, 117, 16558−16564