Analyst View Article Online
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
PAPER
View Journal
Magnetic nanoparticles coated with different shells for biorecognition: high specific binding capacity Cite this: DOI: 10.1039/c3an01726e
Hayrettin Tumturk,* Ferat Sahin and Eylem Turan Modifying the surfaces of magnetic nanoparticles (MNPs) by the covalent attachment of biomolecules will enable their application as media for magnetically-assisted bioseparations. In this paper, we reported both the activity and specific binding capacity of ferritin antibodies on core–shell MNPs. The antibodies were covalently attached on silica-, silver- and polydopamine-coated MNPs by different methods. Anti-ferritin was
bound
onto
the
silica-
or
silver-coated
MNPs
by
conventional
methods
using
3-
aminopropyltriethoxysilane (APTES) or 11-mercaptoundecanoic acid (MUA), which was followed by activation of carboxyl groups by EDC/NHS. However with anti-ferritin immobilized onto the Fe3O4 nanoparticles modified with polydopamine, an in situ coating formed through the adhesive proteins. In addition, a great deal of anti-ferritin biomolecules covalently attached onto the MNPs. According to our results, the amounts of bound anti-ferritin onto the silica-, silver- and PDA-coated MNPs were 70, 75 and 95 mg anti-ferritin per mg MNP, respectively. In the experiments, polydopamine (PDA)-coated MNPs showed faster adsorption, more significant selectivity and a larger binding capacity than the others. Also, the equilibrium dissociation constants of the antigen–antibody complexes were determined on the antiferritin-immobilized MNPs. Silica-, PDA- and silver-coated MNPs had Kd values of 5.45 107, 2.12 107 and 3.91 108 mol L1, respectively. Based on these results, the affinity of the anti-ferritin for ferritin on the PDA-coated MNPs was approximately 10-fold higher than that on the silica- and silverReceived 10th September 2013 Accepted 4th December 2013
coated MNPs. In addition, among the anti-ferritin-immobilized silica-, silver- and PDA-coated MNPs, the PDA-coated MNPs showed the highest antigen selectivity values. As a result, anti-ferritin-immobilized
DOI: 10.1039/c3an01726e
PDA-coated MNPs represented a higher activity and stronger affinity for the specific antigen than the
www.rsc.org/analyst
others.
Introduction Because of their superparamagnetic properties and biocompatibility, silica-, metal- or polymer-coated magnetic nanoparticles have widely been the subject of several investigations such as cell separation, gene targeting, drug delivery, magnetic resonance imaging, and hyperthermia.1–5 Many of these applications demand the core–shell nanoparticles to be chemically stable, uniform in size and well dispersed in liquid medium. The antibodies attached to magnetic nanoparticles (MNPs) have been shown to be specic biological vectors.6 When silica-, metal- or polymer-coated MNPs are biofunctionalized with an antibody, they can be applied for highly sensitive immunoassays.7,8 As a result of the biofunctionalization of nanoparticles with antibodies, they will have new properties such as the specic and selective recognition ability of the antibodies to the antigens. Moreover, the advantages of antibody-immobilized nanoparticles are intercellular stability and cellular uptake. For using MNPs in these applications, the Gazi University, Faculty of Science, Department of Chemistry, 06500 Besevler, Ankara, Turkey. E-mail: [email protected]; Fax: +90 312 2213202; Tel: +90 312 2021129
This journal is © The Royal Society of Chemistry 2014
surface properties of the nanoparticles must be well known. One way of bringing specicity to nanoparticles is through the mimicking of biological interactions such as antigen–antibody or enzyme–substrate interactions, also called lock-and-key interactions.9 In this study, we have investigated the effects of different layers on the antigen–antibody interaction in the nanoparticle system. For this purpose, we have synthesized MNPs via a coprecipitation method and then these nanoparticles were coated with different shell materials such as, silica, polydopamine (PDA), and silver. Silica- and silver-coated MNPs were modied with organic molecules for easy antibody conjugation but PDAcoated nanoparticles were not modify with any organic molecules. Owing to the catechol and amine groups of the PDA structure, these nanoparticles exhibit reactivity towards the amine and thiol groups in antibodies. The specic and selective binding of anti-ferritin antibodies and ferritin protein was examined as a model antigen–antibody interaction. Anti-ferritin was covalently immobilized onto silica-, PDA- and silver-coated MNPs, and the performance of these immobilized nanoparticles in affinity separations was evaluated by using a plasma protein solution.
Analyst
View Article Online
Analyst
Experimental
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
Materials All chemicals were purchased from Sigma-Aldrich at the highest purity and were used without further purication. Chemicals used for the synthesis of magnetic and modied MNPs were iron(II) chloride tetrahydrate (FeCl2$4H2O), iron(III) chloride hexahydrate (FeCl3$6H2O), sodium hydroxide (NaOH), sodium citrate monobasic (C6H7O7$Na), tetraethyl orthosilicate (TEOS, C8H20O4Si), ammonium hydroxide solution (28 wt%, NH3 in H2O), (3-aminopropyl)triethoxysilane (APTES, C9H23NO3Si), dopamine hydrochloride (C8H11NO2$HCl), 11-mercaptoundecanoic acid (MUA, C11H22O2S), silver nitrate solution (2.5% (w/v) AgNO3 in H2O), polyvinylpyrrolidone K30 (PVP) and ethanol (C2H5OH). Ferritin, type I, from horse spleen; anti-ferritin, antibody produced in rabbit; myoglobin from equine heart (Mb), hemoglobin from mouse (Hb) and brinogen from human plasma (Fbg) biomolecules were obtained from Sigma. Ultra-pure water (Human Power I+ Scholar-UV) was used for all experiments.
Paper
the silica-coated MNPs in ethanol was 20 mg mL1. APTES was added into the ask and the mixture was stirred at 50 C for 24 h under the nitrogen ow. The weight ratio between APTES and the silica-coated core–shell nanoparticles 0.35 : 1. The resulting amino-functionalized nanoparticles were washed with ethanol and acetone repeatedly. Aer collection of the nanoparticles using permanent magnet, the product was dried in a vacuum oven. Synthesis of PDA-coated magnetic nanoparticles Fe3O4 nanoparticles were dispersed in 2 mg mL1 dopamine solution (10 mM, pH 8.5 Tris–HCl) under continuous stirring for 24 h at room temperature. Aer the self-polymerization of dopamine, the product was collected using a magnet and washed with ultra-pure water under stirring for 12 h before drying at 50 C under vacuum for 12 h before electroless plating. Synthesis of Ag-coated magnetic nanoparticles
1 M Fe2+ and 2 M Fe3+ chloride solutions were prepared by dissolving FeCl2$4H2O and FeCl3$6H2O in 25 mL ultra-pure water with a molar ratio of 1 : 2. Magnetic nanoparticles (Fe3O4) were synthesized by adding 1.5 M sodium hydroxide solution dropwise from a burette at 50 C under nitrogen ow (50 mL min1) with vigorous mechanical stirring. Black precipitates were formed immediately. Nanoparticles were washed with ethanol and ultra-pure water, repeatedly before nally being re-dispersed in ultra-pure water. During the washing steps, a permanent magnet was used to assist collection of the magnetic black precipitate. The obtained Fe3O4 nanoparticles were dried at 50 C under vacuum for 12 h.
Silver-coated MNPs were synthesized with minor modication by a reported method.10 The procedure is as follows. Silvercoated core–shell nanoparticles were prepared via the reduction of silver on the PDA-modied MNPs. Ammoniacal silver nitrate solution was prepared by adding ammonia dropwise into AgNO3 solution. PDA-modied Fe3O4 nanoparticles were added to the transparent ammoniacal silver nitrate solution (30.0 g L1, 100 mL) and le for 15 min. In order to reduce the [Ag(NH3)2]+ ions, 100 mL of 1.0 M hot glucose solution was added to the mixture. The weight ratio between AgNO3 and glucose was 2 : 1. Ammoniacal silver nitrate and glucose solutions included 0.02 wt% ethanol and 0.02 wt% PVP. Aer the electroless plating, the silver-coated MNPs were washed with ultra-pure water several times. Aer collection of the nanoparticles using a permanent magnet, the product was dried in a vacuum oven.
Synthesis of silica-coated magnetic nanoparticles
Synthesis of carboxylated Fe3O4 nanoparticles
80 mL ethanol, 20 mL ultra-pure water and 5 mL TEOS were added to a 250 mL three-neck ask in an ultrasonic water bath. 20 mg of Fe3O4 nanoparticles were added to the ask and treated with TEOS solution by ultrasonication for 20 min. Then 5 mL ammonia solution (10 wt%) was added dropwise to the Fe3O4 suspension under vigorous mechanical stirring at room temperature for 12 h. The resulting silica-coated Fe3O4 nanoparticles were thoroughly washed with ethanol to remove nonmagnetic by-products. Aer collection of the nanoparticles using a permanent magnet, the product was dried in a vacuum oven.
The as-synthesized carboxylated silver-coated core–shell nanoparticles were transferred to ethanol through ligand exchange using 11-mercaptoundecanoic acid (MUA) by following procedure: A 3 mM MUA solution was prepared by adding MUA into 20 mL ethanol. The silver-coated nanoparticles (5 mg) were dispersed in the MUA solution with stirring under nitrogen ow for 6 h. Aer the ligand exchange process, the carboxylated silver-coated Fe3O4 nanoparticles were washed with ethanol to remove unreacted MUA, and then the nanoparticles were washed in ultra-pure water several times, and collected by magnetic separation, which was followed by drying at 50 C under vacuum for 12 h.
Synthesis of magnetic nanoparticles
Synthesis of amino-functionalized silica-coated magnetic nanoparticles Amino-functionalized silica-coated MNPs were prepared by surface functionalization of silica-coated nanoparticles using APTES as the silylation agent. Silica-coated Fe3O4 nanoparticles and ethanol were added to a 50 mL three-neck ask and then dispersed for 30 min by ultrasonication. The concentration of
Analyst
Antibody attachment 1 mL of solution of the amino-functionalized silica-coated MNPs (5 mg mL1) was dispersed in PBS (pH 7.4) by sonication for 20 min. Anti-ferritin (1 mL of a 1 mg mL1) was also preactivated in PBS containing EDC (10 mg) and NHS (10 mg) for This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
Paper
10 min, before being added to the amino-functionalized silicacoated MNPs dispersion and incubated at room temperature for 2 h. 1 mL of solution of carboxylated silver-coated MNPs (5 mg mL1) was dispersed in PBS (pH 7.4) by sonication for 20 min. Anti-ferritin (1 mL of a 1 mg mL1) was also preactivated in PBS containing EDC (50 mg) and NHS (50 mg) for 10 min, before being added to the carboxylated silver-coated MNPs dispersion and incubated at room temperature for 2 h. 1 mL of solution of the PDA-coated MNPs (5 mg mL1) were dispersed in PBS (pH 7.4) by sonication for 20 min and then anti-ferritin (1 mL of a 1 mg mL1) was added to the dispersion without activation by EDC/NHS, before being incubated at room temperature for 2 h. The antibody-attached MNPs were washed several times with the buffer solution. The washing solution and remaining antibody solution were collected and then the adsorbed amount of antibody on the nanoparticles was determined by UV/Vis spectroscopy at 280 nm. Finally they were re-dispersed in PBS (pH 7.4) containing 1% (w/w) bovine serum albumin (BSA) to prevent non-specic protein adsorption. Aer re-dispersion in PBS the antibody-coated nanoparticles were stored at +4 C until further use (to prevent decomposition of the immobilized antibodies). Antibody–antigen reaction on different magnetic nanoparticles To observe the antibody–antigen reaction, 0.1 mg mL1 ferritin solution was prepared (in PBS, pH 7.4) and 1 mg anti-ferritinattached MNPs (silica, PDA and silver-coated) was added to this solution. The resulting solution was incubated for 1 h at room temperature. Aer washing steps, quantitative determination of the amount of reacted ferritin was calculated using the Lambert–Beer law by UV/Vis spectroscopy. The amount of reacted ferritin was determined using a standard ferritin calibration curve. The amount of ferritin bound to the anti-ferritinimmobilized nanoparticles was calculated using the following equation: Amount of bound ferritin ¼ (initial amount of ferritin) (remaining amount of ferritin)
Analyst
combined. Then, the adsorbed amount of antigens and different proteins were determined by UV-Vis spectroscopy at different wavelengths. The absorbances of ferritin, Fbg, Hb and Mb proteins were measured 280, 595, 406 and 409 nm, respectively. Quantitative determination of the amount reacted protein (antigen or protein) was used to calculate the Lambert– Beer law as explained above. The amount of adsorbed protein was determined using standard protein calibration curve. The amount of adsorbed protein on the anti-ferritin-immobilized nanoparticles was calculated using the following equation: Amount of adsorbed protein ¼ (initial amount of protein) (remaining amount of protein) Characterization Transmission electron microscopy (TEM) images were recorded using a JEOL 100CX transmission electron microscope with an acceleration voltage of 120 kV. The samples were prepared without staining by placing a drop of a dilute ethanolic dispersion of the MNPs on a copper grid coated with a carbon lm. XPS spectra were recorded using a SPECS ESCA spectrometer equipped with a Mg Ka X-ray source. Aer peak tting of the C1s spectra, all the spectra were calibrated in reference to the aliphatic C1s component at a binding energy of 285.0 eV. FTIR spectra were recorded using a Bruker IFS 66/S spectrophotometer. Raman spectra were recorded using a Bruker FRA 106/S spectrophotometer with a 532 nm laser source. Magnetic measurements of the samples were performed by a vibrating sample magnetometer (VSM) from Cryogenic Limited PPMS system. The samples were in the form of a powder and were placed in a Teon sample holder. Magnetic measurements were performed using a superconducting quantum interference device magnetometer with magnetic elds up to 7 T at 25 C. UV-Vis spectra were collected using a Shimadzu UV-2450 spectrophotometer at room temperature.
Results and discussion
Selectivity
Ferritin antibody attachment on different magnetic nanoparticles
The selectivity experiments of the antibody attachment nanoparticles were carried out by using Fbg, Mb, and Hb as reference proteins. The nanoparticles were dispersed different protein solutions containing ferritin/Fbg, ferritin/Hb and ferritin/Mb at equal concentration of 0.1 mg mL1 in PBS (pH 7.4). The antiferritin-immobilized MNPs were incubated with these protein solutions for 1 h at RT. The nanoparticles were taken out from the solution by permanent magnet and washed with pH 7.4 buffer solution with several times. UV/Vis spectroscopy was chosen to determine the adsorption capacities of the nanoparticles. The adsorption capacities of the antibody attachment nanoparticles were also calculated from absorbance by Lambert–Beer law. Aer washed and collected the nanoparticles, rinsing solutions and remain protein solutions were
To prepare antibody-attached MNPs by covalent binding, silica-, silver- and PDA-coated MNPs were synthesized rst of all. For the initial MNP surface modication step, reaction with silica, silver and PDA was veried using TEM, XPS and VSM analysis. Secondly, silica-coated MNPs were modied using APTES by a literature procedure11 and silver-coated MNPs were functionalized with MUA via the ligand exchange procedure described by Abad et al.12 for antibody conjugation. The characterization results are summarized in Fig. 1–3. Fig. 1 demonstrates the TEM images of bare and modied MNPs. The bare MNPs were not uniform in size and seen to exhibit strong agglomeration in Fig. 1a. Their average diameters were 8–10 nm. With the silica-coated MNPs, they are well dispersed and the typical core–shell structure with an average
This journal is © The Royal Society of Chemistry 2014
Analyst
View Article Online
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
Analyst
Paper
Fig. 3
Fig. 1 TEM images of (a) bare MNPs, (b) silica-coated MNPs, (c) PDAcoated MNPs, and (d) silver-coated MNPs.
diameter 89 nm is seen in Fig. 1b. A typical TEM image of PDAcoated MNPs prepared via the oxidative self-polymerization of dopamine is shown in Fig. 1c. In the TEM image of the PDAcoated MNPs, a continuous and thin PDA layer was clearly observed on the outer shell of the Fe3O4 core. We repeated calculation of the PDA layer thickness and the average value of the PDA layer was found to be between 6 and 12 nm. In Fig. 1d, silver-coated MNPs were uniform in size and showed no agglomeration. Their average diameter was 44 nm.
Fig. 2
Analyst
VSM curves of bare, silica, PDA, and silver-coated MNPs.
The XPS spectra of bare and silica-coated MNPs are shown in Fig 2a and 2b. According to Fig 2a, the peaks of Fe2p3/2 (710 eV), Fe2p1/2 (720 eV) and O1s (530 eV) were found in the bare ultrane MNPs, which corresponded to a characteristic region of the bare MNPs. From Fig 2b, the peaks of Si2p (103 eV) can be seen on the silica-coated MNPs, which originate from the silica coating, and the Fe2p peaks disappeared due to the presence of the thick silica layer. Also, the binding energy of the O1s peak was at approximately 531.7 eV but it was higher than for bare MNPs by 1.7 eV. The formation of the Fe–O–Si chemical bond was the main reason for that. In addition, the presence of the weak C1s (285 eV) signal was probably due to the adsorption of atmospheric contaminants on the surface of the silica-coated MNPs during XPS analysis. The XPS spectra of PDA-coated MNPs are shown in Fig. 2c. By comparison with Fig. 2a, it can be seen in Fig. 2c that new signals of C1s (285 eV) and N1s (400 eV) appeared. The N1s peak was attributed to the amine groups of the PDA structure;
XPS spectra of (a) bare MNPs, (b) silica-coated MNPs, (c) PDA-coated MNPs, and (d) silver-coated MNPs.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
Paper
the O1s (532 eV) peak corresponded to the catechol groups of the PDA structure. Fig. 2d shows the survey scan spectrum of the silver-coated MNPs. The strong signal of Ag3d (370 eV) in Fig. 2d indicates that a layer of silver has been deposited on the PDA-coated MNPs. As shown in the XPS spectrum of the silver-coated MNPs, the peaks at 369.8 eV (Ag3d5/2) and 374.4 eV (Ag3d3/2) attributable to the Ag0 species, further conrming the presence of silver in the metallic state. The VSM curves of bare, silica-, silver- and PDA-coated MNPs indicated that these nanoparticles demonstrated typical superparamagnetic behavior at room temperature. The magnetization saturation values of the bare and core–shell MNPs were 63.2, 22.1, 28.5 and 36.2 emu g1, respectively (Fig. 3). The bare MNPs showed a higher value of magnetization saturation than the silica-, PDA- and silver-coated MNPs. The decrease in saturation magnetization values also indicated that the silica, silver, and PDA coatings were achieved successfully. Accordingly, these MNPs can be readily separated from the solution with external magnetic eld due to their superparamagnetism and large saturation magnetization. Anti-ferritin was preactivated by EDC/NHS in solution before it was bonded onto the APTES-modied MNPs surfaces. For covalent immobilization on MUA-modied MNPs surfaces, carboxyl groups were activated using EDC/NHS, and antiferritin was conjugated onto these nanoparticles.13 On the other hand, ferritin antibodies were covalently attached on the PDA-coated MNPs which contain residual quinones through Michael addition and/or Schiff base formation. In the present case, similar reactions such as through the quinone groups of the PDA layer and the amino or thiol groups of the antibodies were used for the covalent immobilization of the antiferritin.14,15 Antibody conjugation was accomplished via reacting a constant concentration of silica-, silver- and PDA-coated MNPs with a xed concentration of anti-ferritin solution. The reactions between silica-, silver- and PDA-coated MNPs and ferritin antibodies were spontaneous and rapid, resulting in precipitations of anti-ferritin attachment MNPs aer mixing. Free antibody concentrations in solution during covalent attachment at room temperature showed little change aer 1 h, thus, covalent binding reaction was completed. Quantitative UV assays were conducted to determine the amount of ferritin antibody (as described in the Experimental section). Fig. 4 shows the amount of anti-ferritin attached on the different MNPs. High numbers of anti-ferritin molecules were covalently attached on the silica-, silver- and PDA-coated MNPs. Among the silica-, PDA- and silver-coated MNPs, the PDA-coated MNPs showed the highest value which was approximately 95 mg of anti-ferritin per mg of MNPs. In addition, the amounts of bound anti-ferritin onto the silica- and silver-coated MNPs were 70 and 75 mg anti-ferritin per mg MNPs, respectively. For modied MNPs coated with PDA, the formation of a Schiff base occurs between the active quinone groups on the nanoparticle surface and amine or thiol groups on the antibody surface. Consequently, many more anti-ferritin molecules were covalently attached on the PDA-coated MNPs. These results showed that the PDA-coated MNPs are more suitable than the other
This journal is © The Royal Society of Chemistry 2014
Analyst
Fig. 4 The amount of anti-ferritin attached on the different MNPs.
prepared nanoparticles for covalent attachment of high amounts of anti-ferritin. For verication of the protein layer on the nanoparticles, Raman spectroscopy analysis was carried out. Fig. 5 demonstrates the 532 nm-excitation Raman spectra of anti-ferritin attached on the different MNPs. The peaks located at 1661, 1620, 1555, and 1440 cm1 can be attributed to the amide I, aromatic ring band, amide II and C–H symmetric bending band, respectively. Apart from these peaks, the peaks in the 1250–1350 cm1 region can be attributed to the amide III bands. Covalent attachment of the anti-ferritin on the MNPs was also veried by FTIR analysis. Fig. 6 demonstrates the FTIR spectra of anti-ferritin attached on the different MNPs. The presence of the amide I and amide II bands around the 1650– 1500 cm1 region and also the asymmetric and symmetric bands of methylene groups at the 2800–2900 cm1 region also supported the presence of the protein layer on the MNPs. The Raman intensity and absorbance of spectrum increased with an increasing amount of protein on the MNPs. Moreover, these results agreed with UV-Vis results.
Raman spectra of (a) silica-coated, (b) silver-coated, and (c) PDA-coated MNPs (anti-ferritin unmodified MNPs were used as the baseline for all Raman spectra).
Fig. 5
Analyst
View Article Online
Analyst
Paper
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
reduced when the antibody was immobilized in a random manner. In addition, zero, one or two ferritin molecules can bind to an antibody. However, each ferritin molecule contains many identical antibody binding sites or epitopes. Due to its polyepitopic nature, the availability of antibody binding sites is less dependent on the conformation/orientation of the antiferritin molecule on the MNPs surface.17 On the basis of these results, the anti-ferritin-coated different MNPs can be used as affinity nanoparticles. Dissociation constants of antigen–antibody complexes on the different MNPs
FTIR spectra of (a) silica-coated, (b) silver-coated, and (c) PDAcoated MNPs (anti-ferritin unmodified MNPs were used as the baseline for all FTIR spectra).
Fig. 6
Antibody–antigen reaction on different magnetic nanoparticles The activities and functions of the antibodies attached on the silica-, silver- and PDA-coated MNPs were evaluated via investigation of the antibody–antigen reactions. Fig. 7 shows the amount of specically bound ferritin to the anti-ferritin-attached different MNPs as a function of the ferritin concentration. In all cases, the amount of ferritin increased linearly at low antigen concentrations and was followed by a plateau region beginning at 0.1 mg mL1 [ferritin]. At this point, saturation was reached at 74 mg ferritin per mg MNPs for the anti-ferritinattached PDA-coated MNPs. Similarly, for anti-ferritin-attached silica- and silver-coated MNPs, the saturation values were at approximately 41 and 48 mg ferritin per mg MNPs, respectively. The closely corresponding saturation points of these two graphs indicate that approximately 1 : 1 binding was observed between the attached anti-ferritin and ferritin on the PDA-coated MNPs. When the antibody is bound to the nanoparticle surface by primary amine groups in the Fc domain, both Fab regions in the antibody can bind to the target antigens.16 As in the case of the anti-ferritin-attached MNPs, the antigen binding capacity was
Fig. 7 Antibody–antigen reaction on the surface of different MNPs as a function of ferritin concentration.
Analyst
The equilibrium dissociation constant (Kd) indicates the achievement of antigen–antibody complexes in bio-separation, diagnosis and detection systems. We calculated the equilibrium dissociation constants (Kd) of the antigen–antibody complexes using the Hill equation, which is appropriate for multivalent binding analysis.18,19 The equilibrium dissociation constant Kd is as follows: q¼
½Agn Kd þ ½Agn
where q is the fractional occupancy, that is, the ratio of the amount of bound antigen to the amount of surface-immobilized antibody, [Ag] is the antigen concentration, and n is the Hill coefficient. Rearranging this equation results in the Hill plot, which is derived as the following equation: log q ¼ n log[Ag] log Kd. To calculate the equilibrium dissociation constants of antigen– antibody complexes, the amounts of bound antigens to the antiferritin-immobilized MNPs were measured in various antigen concentrations (6.25 103 to 0.1 mg mL1). Linear curves were observed for all three substrates. Dissociation constants and Hill coefficients were calculated from these curves. The Hill plots obtained are displayed in Fig. 8. The Hill plots showed linear behaviours with moderately high R2 values for the three platforms. Kd values of immobilized antibodies to specic antigens range from 107 to 105 mol L1, however, their soluble counterparts showed approximately 1000-fold stronger binding affinities with Kd values from 1010 to 109 mol L1.20 Silica-, PDA- and silver-coated MNPs had Kd values of 5.45 107, 2.12 107 and 3.91 108 mol L1, respectively. According to these values, the loss of bioactivity is widely acknowledged and is thought to result from (i) denaturation of the three-dimensional structure of the proteins, and (ii) steric hindrance of the antigen-recognition sites.21 In addition, these results showed that the formation of the strongest antigen–antibody complex was on the anti-ferritin–PDA-coated MNPs. The affinity of the anti-ferritin to ferritin on the PDAcoated MNPs was approximately 10-fold higher than that on the silica- and silver-coated MNPs. The Hill coefficient, n, denes the cooperativity of antigen binding. If n > 1, i.e. the cooperativity is positive, when the antigen molecule binds to the specic antibody, the affinity of the antibody to other antigens increases. If n < 1, i.e. the
This journal is © The Royal Society of Chemistry 2014
View Article Online
Paper
Analyst
difference in the state of the antibody on the nanoparticle. Further, non-specic protein adsorption on the nanoparticle surface hardly occurred because of the immobilized antibody. In addition, small proteins were more easily adsorbed on the nanoparticle surfaces due to their molecular size.
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
Conclusions
Hill plots of the reaction between ferritin and anti-ferritinattached various MNPs.
Fig. 8
cooperativity is negative, when the antigen molecule binds to the specic antibody, the affinity of the antibody to other antigens decreases. If n ¼ 1, no cooperativity is observed and each binding site functions independently.22 The Hill coefficients of the silica-, silver- and PDA-coated MNPs were 1.16, 1.23 and 1.31, respectively. As a result, silica-, silver- and PDA-coated MNPs demonstrated positive cooperativity. These results may affect the amount of protein adsorption and the separation efficiency.
Because of their good dispersibility in an aqueous medium and a very high ratio of surface area to volume, modied MNPs have been widely used for the recognition and analysis of a specic biomolecule from a mixture under biological conditions. In order to increase the specic separation and recognition, antibodies were immobilized on MNPs coated with different functional layers such as silica, silver and PDA. The antibodyimmobilized PDA-coated MNPs showed high activity and affinity toward the specic antigens. In addition, it was determined that antibody-immobilized PDA-MNPs showed a high affinity and selectivity toward the complementary antigen in a protein–antigen mixture. According to the results of this study, antibody-immobilized PDA-coated MNPs could be good candidates for use in the bioaffinity recognition and separation of proteins. Moreover, aer extra modication steps silica- and silver-coated MNPs could be used for similar applications.
Acknowledgements This study was supported by Gazi University Scientic Research Projects Unit, Project no: 05/2010-76.
Affinity selection of target antigen with References different molecular weight proteins In order to determine the selectivity of antibody-coated MNPs to specic antigen, the binding capacity for variously sized proteins was measured. Table 1 demonstrated the amount of captured proteins per each immobilized antibody. Generally, anti-ferritin-immobilized silica-, silver- and PDAcoated MNPs could bind ferritin signicantly compared with other plasma proteins such as Fbg (340 kDa), Hb (65 kDa), and Mb (17 kDa). The selectivities of the anti-ferritin-immobilized silica-coated MNPs for ferritin toward Fbg, Hb, and Mb were 58, 34, and 19, respectively. In the same way, the antigen selectivities of the antibody-immobilized silver- and PDA-coated MNPs for Fbg, Hb, and Mb were 87, 52, and 27, and 178, 78 and 63, respectively. The PDA-coated MNP selectivity values were larger than those observed for the other MNPs. This is due to the
Table 1 Affinity selection of target antigen with different molecular weight proteins (each protein concentration: 0.1 mg mL1, time: 60 min at RT)
MNPs
Fbg
Hb
Mb
Silica-coated Ag-coated PDA-coated
58 87 178
34 52 78
19 27 63
This journal is © The Royal Society of Chemistry 2014
1 M. Kuhara, H. Takeyama, T. Tanaka and T. Matsunaga, Anal. Chem., 2004, 76, 6207. 2 S. Prijic and G. Sersa, Radiol. Oncol., 2011, 45, 1. 3 C. S. McBain, H. H. P. Yiu and J. Dobson, Int. J. Nanomed., 2008, 3, 169. 4 B. Gleich and J. Weizenecker, Nature, 2005, 435, 1214. 5 H. Pardoe, P. R. Clark, T. G. St. Pierre, P. Moroz and S. K. Jones, Magn. Reson. Imaging, 2003, 21, 483. 6 I. Garcia, J. Gallo, N. Genicio, D. Padro and S. Penades, Bioconjugate Chem., 2011, 22, 264. 7 T. Matsunaga, M. Kawasaki, X. Yu, N. Tsujimura and N. Nakamura, Anal. Chem., 1996, 68, 3551. 8 T. Tanaka and T. Matsunaga, Anal. Chem., 2000, 72, 3518. 9 M. Arruebo, M. Valladares and A. Gonzales-Fernandez, J. Nanomater., 2009, 2009, 1. 10 W. Wang, Y. Jiang, Y. Liao, M. Tian, H. Zou and L. Zhang, J. Colloid Interface Sci., 2011, 358, 567. 11 H. Wang, M. Peng, J. Zheng and P. Li, J. Colloid Interface Sci., 2008, 326, 151. 12 J. M. Abad, S. F. L. Mertens, M. Pita, V. M. Fernandez and D. J. Schiffrin, J. Am. Chem. Soc., 2005, 127, 5689. 13 S. Hong, D. Lee, H. Zhang, J. Q. Zhang, J. N. Resvick, A. Khademhosseini, M. R. King, R. Langer and J. M. Karp, Langmuir, 2007, 23, 12261.
Analyst
View Article Online
Analyst
18 S. T. Pathirana, J. Barbaree, B. A. Chin, M. G. Hartell, W. C. Neely and V. Vodyanoy, Biosens. Bioelectron., 2000, 15, 135. 19 H. Li, S. Chen and H. Zhao, Biophys. J., 1990, 58, 1313. 20 B. Friguet, A. Fedorov, A. A. Serganov, A. Navon and M. E. Goldberg, Anal. Biochem., 1993, 210, 344. 21 J. E. Butler, J. Immunol. Methods, 2000, 22, 4. 22 N. Tajima, M. Takai and K. Ishihara, Anal. Chem., 2011, 83, 1969.
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
14 H. Lee, S. M. Dellatore, W. M. Miller and P. B. Messersmith, Science, 2007, 318, 426. 15 H. Lee, J. Rho, K. Holmburg and P. B. Messersmith, Adv. Mater., 2009, 21, 431. 16 S. Puertas, P. Batalla, M. Moros, E. Polo, P. Pino, J. M. Guisan, V. Grazu and J. M. Fuente, ACS Nano, 2011, 5, 4521. 17 S. Allen, X. Chen, J. Davies, M. C. Davies, A. C. Dawkes, J. C. Edwards, C. J. Roberts, J. Seon, S. J. B. Tendler and P. M. Williams, Biochemistry, 1997, 36, 7457.
Paper
Analyst
This journal is © The Royal Society of Chemistry 2014
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
PAPER
View Journal
Magnetic nanoparticles coated with different shells for biorecognition: high specific binding capacity Cite this: DOI: 10.1039/c3an01726e
Hayrettin Tumturk,* Ferat Sahin and Eylem Turan Modifying the surfaces of magnetic nanoparticles (MNPs) by the covalent attachment of biomolecules will enable their application as media for magnetically-assisted bioseparations. In this paper, we reported both the activity and specific binding capacity of ferritin antibodies on core–shell MNPs. The antibodies were covalently attached on silica-, silver- and polydopamine-coated MNPs by different methods. Anti-ferritin was
bound
onto
the
silica-
or
silver-coated
MNPs
by
conventional
methods
using
3-
aminopropyltriethoxysilane (APTES) or 11-mercaptoundecanoic acid (MUA), which was followed by activation of carboxyl groups by EDC/NHS. However with anti-ferritin immobilized onto the Fe3O4 nanoparticles modified with polydopamine, an in situ coating formed through the adhesive proteins. In addition, a great deal of anti-ferritin biomolecules covalently attached onto the MNPs. According to our results, the amounts of bound anti-ferritin onto the silica-, silver- and PDA-coated MNPs were 70, 75 and 95 mg anti-ferritin per mg MNP, respectively. In the experiments, polydopamine (PDA)-coated MNPs showed faster adsorption, more significant selectivity and a larger binding capacity than the others. Also, the equilibrium dissociation constants of the antigen–antibody complexes were determined on the antiferritin-immobilized MNPs. Silica-, PDA- and silver-coated MNPs had Kd values of 5.45 107, 2.12 107 and 3.91 108 mol L1, respectively. Based on these results, the affinity of the anti-ferritin for ferritin on the PDA-coated MNPs was approximately 10-fold higher than that on the silica- and silverReceived 10th September 2013 Accepted 4th December 2013
coated MNPs. In addition, among the anti-ferritin-immobilized silica-, silver- and PDA-coated MNPs, the PDA-coated MNPs showed the highest antigen selectivity values. As a result, anti-ferritin-immobilized
DOI: 10.1039/c3an01726e
PDA-coated MNPs represented a higher activity and stronger affinity for the specific antigen than the
www.rsc.org/analyst
others.
Introduction Because of their superparamagnetic properties and biocompatibility, silica-, metal- or polymer-coated magnetic nanoparticles have widely been the subject of several investigations such as cell separation, gene targeting, drug delivery, magnetic resonance imaging, and hyperthermia.1–5 Many of these applications demand the core–shell nanoparticles to be chemically stable, uniform in size and well dispersed in liquid medium. The antibodies attached to magnetic nanoparticles (MNPs) have been shown to be specic biological vectors.6 When silica-, metal- or polymer-coated MNPs are biofunctionalized with an antibody, they can be applied for highly sensitive immunoassays.7,8 As a result of the biofunctionalization of nanoparticles with antibodies, they will have new properties such as the specic and selective recognition ability of the antibodies to the antigens. Moreover, the advantages of antibody-immobilized nanoparticles are intercellular stability and cellular uptake. For using MNPs in these applications, the Gazi University, Faculty of Science, Department of Chemistry, 06500 Besevler, Ankara, Turkey. E-mail: [email protected]; Fax: +90 312 2213202; Tel: +90 312 2021129
This journal is © The Royal Society of Chemistry 2014
surface properties of the nanoparticles must be well known. One way of bringing specicity to nanoparticles is through the mimicking of biological interactions such as antigen–antibody or enzyme–substrate interactions, also called lock-and-key interactions.9 In this study, we have investigated the effects of different layers on the antigen–antibody interaction in the nanoparticle system. For this purpose, we have synthesized MNPs via a coprecipitation method and then these nanoparticles were coated with different shell materials such as, silica, polydopamine (PDA), and silver. Silica- and silver-coated MNPs were modied with organic molecules for easy antibody conjugation but PDAcoated nanoparticles were not modify with any organic molecules. Owing to the catechol and amine groups of the PDA structure, these nanoparticles exhibit reactivity towards the amine and thiol groups in antibodies. The specic and selective binding of anti-ferritin antibodies and ferritin protein was examined as a model antigen–antibody interaction. Anti-ferritin was covalently immobilized onto silica-, PDA- and silver-coated MNPs, and the performance of these immobilized nanoparticles in affinity separations was evaluated by using a plasma protein solution.
Analyst
View Article Online
Analyst
Experimental
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
Materials All chemicals were purchased from Sigma-Aldrich at the highest purity and were used without further purication. Chemicals used for the synthesis of magnetic and modied MNPs were iron(II) chloride tetrahydrate (FeCl2$4H2O), iron(III) chloride hexahydrate (FeCl3$6H2O), sodium hydroxide (NaOH), sodium citrate monobasic (C6H7O7$Na), tetraethyl orthosilicate (TEOS, C8H20O4Si), ammonium hydroxide solution (28 wt%, NH3 in H2O), (3-aminopropyl)triethoxysilane (APTES, C9H23NO3Si), dopamine hydrochloride (C8H11NO2$HCl), 11-mercaptoundecanoic acid (MUA, C11H22O2S), silver nitrate solution (2.5% (w/v) AgNO3 in H2O), polyvinylpyrrolidone K30 (PVP) and ethanol (C2H5OH). Ferritin, type I, from horse spleen; anti-ferritin, antibody produced in rabbit; myoglobin from equine heart (Mb), hemoglobin from mouse (Hb) and brinogen from human plasma (Fbg) biomolecules were obtained from Sigma. Ultra-pure water (Human Power I+ Scholar-UV) was used for all experiments.
Paper
the silica-coated MNPs in ethanol was 20 mg mL1. APTES was added into the ask and the mixture was stirred at 50 C for 24 h under the nitrogen ow. The weight ratio between APTES and the silica-coated core–shell nanoparticles 0.35 : 1. The resulting amino-functionalized nanoparticles were washed with ethanol and acetone repeatedly. Aer collection of the nanoparticles using permanent magnet, the product was dried in a vacuum oven. Synthesis of PDA-coated magnetic nanoparticles Fe3O4 nanoparticles were dispersed in 2 mg mL1 dopamine solution (10 mM, pH 8.5 Tris–HCl) under continuous stirring for 24 h at room temperature. Aer the self-polymerization of dopamine, the product was collected using a magnet and washed with ultra-pure water under stirring for 12 h before drying at 50 C under vacuum for 12 h before electroless plating. Synthesis of Ag-coated magnetic nanoparticles
1 M Fe2+ and 2 M Fe3+ chloride solutions were prepared by dissolving FeCl2$4H2O and FeCl3$6H2O in 25 mL ultra-pure water with a molar ratio of 1 : 2. Magnetic nanoparticles (Fe3O4) were synthesized by adding 1.5 M sodium hydroxide solution dropwise from a burette at 50 C under nitrogen ow (50 mL min1) with vigorous mechanical stirring. Black precipitates were formed immediately. Nanoparticles were washed with ethanol and ultra-pure water, repeatedly before nally being re-dispersed in ultra-pure water. During the washing steps, a permanent magnet was used to assist collection of the magnetic black precipitate. The obtained Fe3O4 nanoparticles were dried at 50 C under vacuum for 12 h.
Silver-coated MNPs were synthesized with minor modication by a reported method.10 The procedure is as follows. Silvercoated core–shell nanoparticles were prepared via the reduction of silver on the PDA-modied MNPs. Ammoniacal silver nitrate solution was prepared by adding ammonia dropwise into AgNO3 solution. PDA-modied Fe3O4 nanoparticles were added to the transparent ammoniacal silver nitrate solution (30.0 g L1, 100 mL) and le for 15 min. In order to reduce the [Ag(NH3)2]+ ions, 100 mL of 1.0 M hot glucose solution was added to the mixture. The weight ratio between AgNO3 and glucose was 2 : 1. Ammoniacal silver nitrate and glucose solutions included 0.02 wt% ethanol and 0.02 wt% PVP. Aer the electroless plating, the silver-coated MNPs were washed with ultra-pure water several times. Aer collection of the nanoparticles using a permanent magnet, the product was dried in a vacuum oven.
Synthesis of silica-coated magnetic nanoparticles
Synthesis of carboxylated Fe3O4 nanoparticles
80 mL ethanol, 20 mL ultra-pure water and 5 mL TEOS were added to a 250 mL three-neck ask in an ultrasonic water bath. 20 mg of Fe3O4 nanoparticles were added to the ask and treated with TEOS solution by ultrasonication for 20 min. Then 5 mL ammonia solution (10 wt%) was added dropwise to the Fe3O4 suspension under vigorous mechanical stirring at room temperature for 12 h. The resulting silica-coated Fe3O4 nanoparticles were thoroughly washed with ethanol to remove nonmagnetic by-products. Aer collection of the nanoparticles using a permanent magnet, the product was dried in a vacuum oven.
The as-synthesized carboxylated silver-coated core–shell nanoparticles were transferred to ethanol through ligand exchange using 11-mercaptoundecanoic acid (MUA) by following procedure: A 3 mM MUA solution was prepared by adding MUA into 20 mL ethanol. The silver-coated nanoparticles (5 mg) were dispersed in the MUA solution with stirring under nitrogen ow for 6 h. Aer the ligand exchange process, the carboxylated silver-coated Fe3O4 nanoparticles were washed with ethanol to remove unreacted MUA, and then the nanoparticles were washed in ultra-pure water several times, and collected by magnetic separation, which was followed by drying at 50 C under vacuum for 12 h.
Synthesis of magnetic nanoparticles
Synthesis of amino-functionalized silica-coated magnetic nanoparticles Amino-functionalized silica-coated MNPs were prepared by surface functionalization of silica-coated nanoparticles using APTES as the silylation agent. Silica-coated Fe3O4 nanoparticles and ethanol were added to a 50 mL three-neck ask and then dispersed for 30 min by ultrasonication. The concentration of
Analyst
Antibody attachment 1 mL of solution of the amino-functionalized silica-coated MNPs (5 mg mL1) was dispersed in PBS (pH 7.4) by sonication for 20 min. Anti-ferritin (1 mL of a 1 mg mL1) was also preactivated in PBS containing EDC (10 mg) and NHS (10 mg) for This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
Paper
10 min, before being added to the amino-functionalized silicacoated MNPs dispersion and incubated at room temperature for 2 h. 1 mL of solution of carboxylated silver-coated MNPs (5 mg mL1) was dispersed in PBS (pH 7.4) by sonication for 20 min. Anti-ferritin (1 mL of a 1 mg mL1) was also preactivated in PBS containing EDC (50 mg) and NHS (50 mg) for 10 min, before being added to the carboxylated silver-coated MNPs dispersion and incubated at room temperature for 2 h. 1 mL of solution of the PDA-coated MNPs (5 mg mL1) were dispersed in PBS (pH 7.4) by sonication for 20 min and then anti-ferritin (1 mL of a 1 mg mL1) was added to the dispersion without activation by EDC/NHS, before being incubated at room temperature for 2 h. The antibody-attached MNPs were washed several times with the buffer solution. The washing solution and remaining antibody solution were collected and then the adsorbed amount of antibody on the nanoparticles was determined by UV/Vis spectroscopy at 280 nm. Finally they were re-dispersed in PBS (pH 7.4) containing 1% (w/w) bovine serum albumin (BSA) to prevent non-specic protein adsorption. Aer re-dispersion in PBS the antibody-coated nanoparticles were stored at +4 C until further use (to prevent decomposition of the immobilized antibodies). Antibody–antigen reaction on different magnetic nanoparticles To observe the antibody–antigen reaction, 0.1 mg mL1 ferritin solution was prepared (in PBS, pH 7.4) and 1 mg anti-ferritinattached MNPs (silica, PDA and silver-coated) was added to this solution. The resulting solution was incubated for 1 h at room temperature. Aer washing steps, quantitative determination of the amount of reacted ferritin was calculated using the Lambert–Beer law by UV/Vis spectroscopy. The amount of reacted ferritin was determined using a standard ferritin calibration curve. The amount of ferritin bound to the anti-ferritinimmobilized nanoparticles was calculated using the following equation: Amount of bound ferritin ¼ (initial amount of ferritin) (remaining amount of ferritin)
Analyst
combined. Then, the adsorbed amount of antigens and different proteins were determined by UV-Vis spectroscopy at different wavelengths. The absorbances of ferritin, Fbg, Hb and Mb proteins were measured 280, 595, 406 and 409 nm, respectively. Quantitative determination of the amount reacted protein (antigen or protein) was used to calculate the Lambert– Beer law as explained above. The amount of adsorbed protein was determined using standard protein calibration curve. The amount of adsorbed protein on the anti-ferritin-immobilized nanoparticles was calculated using the following equation: Amount of adsorbed protein ¼ (initial amount of protein) (remaining amount of protein) Characterization Transmission electron microscopy (TEM) images were recorded using a JEOL 100CX transmission electron microscope with an acceleration voltage of 120 kV. The samples were prepared without staining by placing a drop of a dilute ethanolic dispersion of the MNPs on a copper grid coated with a carbon lm. XPS spectra were recorded using a SPECS ESCA spectrometer equipped with a Mg Ka X-ray source. Aer peak tting of the C1s spectra, all the spectra were calibrated in reference to the aliphatic C1s component at a binding energy of 285.0 eV. FTIR spectra were recorded using a Bruker IFS 66/S spectrophotometer. Raman spectra were recorded using a Bruker FRA 106/S spectrophotometer with a 532 nm laser source. Magnetic measurements of the samples were performed by a vibrating sample magnetometer (VSM) from Cryogenic Limited PPMS system. The samples were in the form of a powder and were placed in a Teon sample holder. Magnetic measurements were performed using a superconducting quantum interference device magnetometer with magnetic elds up to 7 T at 25 C. UV-Vis spectra were collected using a Shimadzu UV-2450 spectrophotometer at room temperature.
Results and discussion
Selectivity
Ferritin antibody attachment on different magnetic nanoparticles
The selectivity experiments of the antibody attachment nanoparticles were carried out by using Fbg, Mb, and Hb as reference proteins. The nanoparticles were dispersed different protein solutions containing ferritin/Fbg, ferritin/Hb and ferritin/Mb at equal concentration of 0.1 mg mL1 in PBS (pH 7.4). The antiferritin-immobilized MNPs were incubated with these protein solutions for 1 h at RT. The nanoparticles were taken out from the solution by permanent magnet and washed with pH 7.4 buffer solution with several times. UV/Vis spectroscopy was chosen to determine the adsorption capacities of the nanoparticles. The adsorption capacities of the antibody attachment nanoparticles were also calculated from absorbance by Lambert–Beer law. Aer washed and collected the nanoparticles, rinsing solutions and remain protein solutions were
To prepare antibody-attached MNPs by covalent binding, silica-, silver- and PDA-coated MNPs were synthesized rst of all. For the initial MNP surface modication step, reaction with silica, silver and PDA was veried using TEM, XPS and VSM analysis. Secondly, silica-coated MNPs were modied using APTES by a literature procedure11 and silver-coated MNPs were functionalized with MUA via the ligand exchange procedure described by Abad et al.12 for antibody conjugation. The characterization results are summarized in Fig. 1–3. Fig. 1 demonstrates the TEM images of bare and modied MNPs. The bare MNPs were not uniform in size and seen to exhibit strong agglomeration in Fig. 1a. Their average diameters were 8–10 nm. With the silica-coated MNPs, they are well dispersed and the typical core–shell structure with an average
This journal is © The Royal Society of Chemistry 2014
Analyst
View Article Online
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
Analyst
Paper
Fig. 3
Fig. 1 TEM images of (a) bare MNPs, (b) silica-coated MNPs, (c) PDAcoated MNPs, and (d) silver-coated MNPs.
diameter 89 nm is seen in Fig. 1b. A typical TEM image of PDAcoated MNPs prepared via the oxidative self-polymerization of dopamine is shown in Fig. 1c. In the TEM image of the PDAcoated MNPs, a continuous and thin PDA layer was clearly observed on the outer shell of the Fe3O4 core. We repeated calculation of the PDA layer thickness and the average value of the PDA layer was found to be between 6 and 12 nm. In Fig. 1d, silver-coated MNPs were uniform in size and showed no agglomeration. Their average diameter was 44 nm.
Fig. 2
Analyst
VSM curves of bare, silica, PDA, and silver-coated MNPs.
The XPS spectra of bare and silica-coated MNPs are shown in Fig 2a and 2b. According to Fig 2a, the peaks of Fe2p3/2 (710 eV), Fe2p1/2 (720 eV) and O1s (530 eV) were found in the bare ultrane MNPs, which corresponded to a characteristic region of the bare MNPs. From Fig 2b, the peaks of Si2p (103 eV) can be seen on the silica-coated MNPs, which originate from the silica coating, and the Fe2p peaks disappeared due to the presence of the thick silica layer. Also, the binding energy of the O1s peak was at approximately 531.7 eV but it was higher than for bare MNPs by 1.7 eV. The formation of the Fe–O–Si chemical bond was the main reason for that. In addition, the presence of the weak C1s (285 eV) signal was probably due to the adsorption of atmospheric contaminants on the surface of the silica-coated MNPs during XPS analysis. The XPS spectra of PDA-coated MNPs are shown in Fig. 2c. By comparison with Fig. 2a, it can be seen in Fig. 2c that new signals of C1s (285 eV) and N1s (400 eV) appeared. The N1s peak was attributed to the amine groups of the PDA structure;
XPS spectra of (a) bare MNPs, (b) silica-coated MNPs, (c) PDA-coated MNPs, and (d) silver-coated MNPs.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
Paper
the O1s (532 eV) peak corresponded to the catechol groups of the PDA structure. Fig. 2d shows the survey scan spectrum of the silver-coated MNPs. The strong signal of Ag3d (370 eV) in Fig. 2d indicates that a layer of silver has been deposited on the PDA-coated MNPs. As shown in the XPS spectrum of the silver-coated MNPs, the peaks at 369.8 eV (Ag3d5/2) and 374.4 eV (Ag3d3/2) attributable to the Ag0 species, further conrming the presence of silver in the metallic state. The VSM curves of bare, silica-, silver- and PDA-coated MNPs indicated that these nanoparticles demonstrated typical superparamagnetic behavior at room temperature. The magnetization saturation values of the bare and core–shell MNPs were 63.2, 22.1, 28.5 and 36.2 emu g1, respectively (Fig. 3). The bare MNPs showed a higher value of magnetization saturation than the silica-, PDA- and silver-coated MNPs. The decrease in saturation magnetization values also indicated that the silica, silver, and PDA coatings were achieved successfully. Accordingly, these MNPs can be readily separated from the solution with external magnetic eld due to their superparamagnetism and large saturation magnetization. Anti-ferritin was preactivated by EDC/NHS in solution before it was bonded onto the APTES-modied MNPs surfaces. For covalent immobilization on MUA-modied MNPs surfaces, carboxyl groups were activated using EDC/NHS, and antiferritin was conjugated onto these nanoparticles.13 On the other hand, ferritin antibodies were covalently attached on the PDA-coated MNPs which contain residual quinones through Michael addition and/or Schiff base formation. In the present case, similar reactions such as through the quinone groups of the PDA layer and the amino or thiol groups of the antibodies were used for the covalent immobilization of the antiferritin.14,15 Antibody conjugation was accomplished via reacting a constant concentration of silica-, silver- and PDA-coated MNPs with a xed concentration of anti-ferritin solution. The reactions between silica-, silver- and PDA-coated MNPs and ferritin antibodies were spontaneous and rapid, resulting in precipitations of anti-ferritin attachment MNPs aer mixing. Free antibody concentrations in solution during covalent attachment at room temperature showed little change aer 1 h, thus, covalent binding reaction was completed. Quantitative UV assays were conducted to determine the amount of ferritin antibody (as described in the Experimental section). Fig. 4 shows the amount of anti-ferritin attached on the different MNPs. High numbers of anti-ferritin molecules were covalently attached on the silica-, silver- and PDA-coated MNPs. Among the silica-, PDA- and silver-coated MNPs, the PDA-coated MNPs showed the highest value which was approximately 95 mg of anti-ferritin per mg of MNPs. In addition, the amounts of bound anti-ferritin onto the silica- and silver-coated MNPs were 70 and 75 mg anti-ferritin per mg MNPs, respectively. For modied MNPs coated with PDA, the formation of a Schiff base occurs between the active quinone groups on the nanoparticle surface and amine or thiol groups on the antibody surface. Consequently, many more anti-ferritin molecules were covalently attached on the PDA-coated MNPs. These results showed that the PDA-coated MNPs are more suitable than the other
This journal is © The Royal Society of Chemistry 2014
Analyst
Fig. 4 The amount of anti-ferritin attached on the different MNPs.
prepared nanoparticles for covalent attachment of high amounts of anti-ferritin. For verication of the protein layer on the nanoparticles, Raman spectroscopy analysis was carried out. Fig. 5 demonstrates the 532 nm-excitation Raman spectra of anti-ferritin attached on the different MNPs. The peaks located at 1661, 1620, 1555, and 1440 cm1 can be attributed to the amide I, aromatic ring band, amide II and C–H symmetric bending band, respectively. Apart from these peaks, the peaks in the 1250–1350 cm1 region can be attributed to the amide III bands. Covalent attachment of the anti-ferritin on the MNPs was also veried by FTIR analysis. Fig. 6 demonstrates the FTIR spectra of anti-ferritin attached on the different MNPs. The presence of the amide I and amide II bands around the 1650– 1500 cm1 region and also the asymmetric and symmetric bands of methylene groups at the 2800–2900 cm1 region also supported the presence of the protein layer on the MNPs. The Raman intensity and absorbance of spectrum increased with an increasing amount of protein on the MNPs. Moreover, these results agreed with UV-Vis results.
Raman spectra of (a) silica-coated, (b) silver-coated, and (c) PDA-coated MNPs (anti-ferritin unmodified MNPs were used as the baseline for all Raman spectra).
Fig. 5
Analyst
View Article Online
Analyst
Paper
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
reduced when the antibody was immobilized in a random manner. In addition, zero, one or two ferritin molecules can bind to an antibody. However, each ferritin molecule contains many identical antibody binding sites or epitopes. Due to its polyepitopic nature, the availability of antibody binding sites is less dependent on the conformation/orientation of the antiferritin molecule on the MNPs surface.17 On the basis of these results, the anti-ferritin-coated different MNPs can be used as affinity nanoparticles. Dissociation constants of antigen–antibody complexes on the different MNPs
FTIR spectra of (a) silica-coated, (b) silver-coated, and (c) PDAcoated MNPs (anti-ferritin unmodified MNPs were used as the baseline for all FTIR spectra).
Fig. 6
Antibody–antigen reaction on different magnetic nanoparticles The activities and functions of the antibodies attached on the silica-, silver- and PDA-coated MNPs were evaluated via investigation of the antibody–antigen reactions. Fig. 7 shows the amount of specically bound ferritin to the anti-ferritin-attached different MNPs as a function of the ferritin concentration. In all cases, the amount of ferritin increased linearly at low antigen concentrations and was followed by a plateau region beginning at 0.1 mg mL1 [ferritin]. At this point, saturation was reached at 74 mg ferritin per mg MNPs for the anti-ferritinattached PDA-coated MNPs. Similarly, for anti-ferritin-attached silica- and silver-coated MNPs, the saturation values were at approximately 41 and 48 mg ferritin per mg MNPs, respectively. The closely corresponding saturation points of these two graphs indicate that approximately 1 : 1 binding was observed between the attached anti-ferritin and ferritin on the PDA-coated MNPs. When the antibody is bound to the nanoparticle surface by primary amine groups in the Fc domain, both Fab regions in the antibody can bind to the target antigens.16 As in the case of the anti-ferritin-attached MNPs, the antigen binding capacity was
Fig. 7 Antibody–antigen reaction on the surface of different MNPs as a function of ferritin concentration.
Analyst
The equilibrium dissociation constant (Kd) indicates the achievement of antigen–antibody complexes in bio-separation, diagnosis and detection systems. We calculated the equilibrium dissociation constants (Kd) of the antigen–antibody complexes using the Hill equation, which is appropriate for multivalent binding analysis.18,19 The equilibrium dissociation constant Kd is as follows: q¼
½Agn Kd þ ½Agn
where q is the fractional occupancy, that is, the ratio of the amount of bound antigen to the amount of surface-immobilized antibody, [Ag] is the antigen concentration, and n is the Hill coefficient. Rearranging this equation results in the Hill plot, which is derived as the following equation: log q ¼ n log[Ag] log Kd. To calculate the equilibrium dissociation constants of antigen– antibody complexes, the amounts of bound antigens to the antiferritin-immobilized MNPs were measured in various antigen concentrations (6.25 103 to 0.1 mg mL1). Linear curves were observed for all three substrates. Dissociation constants and Hill coefficients were calculated from these curves. The Hill plots obtained are displayed in Fig. 8. The Hill plots showed linear behaviours with moderately high R2 values for the three platforms. Kd values of immobilized antibodies to specic antigens range from 107 to 105 mol L1, however, their soluble counterparts showed approximately 1000-fold stronger binding affinities with Kd values from 1010 to 109 mol L1.20 Silica-, PDA- and silver-coated MNPs had Kd values of 5.45 107, 2.12 107 and 3.91 108 mol L1, respectively. According to these values, the loss of bioactivity is widely acknowledged and is thought to result from (i) denaturation of the three-dimensional structure of the proteins, and (ii) steric hindrance of the antigen-recognition sites.21 In addition, these results showed that the formation of the strongest antigen–antibody complex was on the anti-ferritin–PDA-coated MNPs. The affinity of the anti-ferritin to ferritin on the PDAcoated MNPs was approximately 10-fold higher than that on the silica- and silver-coated MNPs. The Hill coefficient, n, denes the cooperativity of antigen binding. If n > 1, i.e. the cooperativity is positive, when the antigen molecule binds to the specic antibody, the affinity of the antibody to other antigens increases. If n < 1, i.e. the
This journal is © The Royal Society of Chemistry 2014
View Article Online
Paper
Analyst
difference in the state of the antibody on the nanoparticle. Further, non-specic protein adsorption on the nanoparticle surface hardly occurred because of the immobilized antibody. In addition, small proteins were more easily adsorbed on the nanoparticle surfaces due to their molecular size.
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
Conclusions
Hill plots of the reaction between ferritin and anti-ferritinattached various MNPs.
Fig. 8
cooperativity is negative, when the antigen molecule binds to the specic antibody, the affinity of the antibody to other antigens decreases. If n ¼ 1, no cooperativity is observed and each binding site functions independently.22 The Hill coefficients of the silica-, silver- and PDA-coated MNPs were 1.16, 1.23 and 1.31, respectively. As a result, silica-, silver- and PDA-coated MNPs demonstrated positive cooperativity. These results may affect the amount of protein adsorption and the separation efficiency.
Because of their good dispersibility in an aqueous medium and a very high ratio of surface area to volume, modied MNPs have been widely used for the recognition and analysis of a specic biomolecule from a mixture under biological conditions. In order to increase the specic separation and recognition, antibodies were immobilized on MNPs coated with different functional layers such as silica, silver and PDA. The antibodyimmobilized PDA-coated MNPs showed high activity and affinity toward the specic antigens. In addition, it was determined that antibody-immobilized PDA-MNPs showed a high affinity and selectivity toward the complementary antigen in a protein–antigen mixture. According to the results of this study, antibody-immobilized PDA-coated MNPs could be good candidates for use in the bioaffinity recognition and separation of proteins. Moreover, aer extra modication steps silica- and silver-coated MNPs could be used for similar applications.
Acknowledgements This study was supported by Gazi University Scientic Research Projects Unit, Project no: 05/2010-76.
Affinity selection of target antigen with References different molecular weight proteins In order to determine the selectivity of antibody-coated MNPs to specic antigen, the binding capacity for variously sized proteins was measured. Table 1 demonstrated the amount of captured proteins per each immobilized antibody. Generally, anti-ferritin-immobilized silica-, silver- and PDAcoated MNPs could bind ferritin signicantly compared with other plasma proteins such as Fbg (340 kDa), Hb (65 kDa), and Mb (17 kDa). The selectivities of the anti-ferritin-immobilized silica-coated MNPs for ferritin toward Fbg, Hb, and Mb were 58, 34, and 19, respectively. In the same way, the antigen selectivities of the antibody-immobilized silver- and PDA-coated MNPs for Fbg, Hb, and Mb were 87, 52, and 27, and 178, 78 and 63, respectively. The PDA-coated MNP selectivity values were larger than those observed for the other MNPs. This is due to the
Table 1 Affinity selection of target antigen with different molecular weight proteins (each protein concentration: 0.1 mg mL1, time: 60 min at RT)
MNPs
Fbg
Hb
Mb
Silica-coated Ag-coated PDA-coated
58 87 178
34 52 78
19 27 63
This journal is © The Royal Society of Chemistry 2014
1 M. Kuhara, H. Takeyama, T. Tanaka and T. Matsunaga, Anal. Chem., 2004, 76, 6207. 2 S. Prijic and G. Sersa, Radiol. Oncol., 2011, 45, 1. 3 C. S. McBain, H. H. P. Yiu and J. Dobson, Int. J. Nanomed., 2008, 3, 169. 4 B. Gleich and J. Weizenecker, Nature, 2005, 435, 1214. 5 H. Pardoe, P. R. Clark, T. G. St. Pierre, P. Moroz and S. K. Jones, Magn. Reson. Imaging, 2003, 21, 483. 6 I. Garcia, J. Gallo, N. Genicio, D. Padro and S. Penades, Bioconjugate Chem., 2011, 22, 264. 7 T. Matsunaga, M. Kawasaki, X. Yu, N. Tsujimura and N. Nakamura, Anal. Chem., 1996, 68, 3551. 8 T. Tanaka and T. Matsunaga, Anal. Chem., 2000, 72, 3518. 9 M. Arruebo, M. Valladares and A. Gonzales-Fernandez, J. Nanomater., 2009, 2009, 1. 10 W. Wang, Y. Jiang, Y. Liao, M. Tian, H. Zou and L. Zhang, J. Colloid Interface Sci., 2011, 358, 567. 11 H. Wang, M. Peng, J. Zheng and P. Li, J. Colloid Interface Sci., 2008, 326, 151. 12 J. M. Abad, S. F. L. Mertens, M. Pita, V. M. Fernandez and D. J. Schiffrin, J. Am. Chem. Soc., 2005, 127, 5689. 13 S. Hong, D. Lee, H. Zhang, J. Q. Zhang, J. N. Resvick, A. Khademhosseini, M. R. King, R. Langer and J. M. Karp, Langmuir, 2007, 23, 12261.
Analyst
View Article Online
Analyst
18 S. T. Pathirana, J. Barbaree, B. A. Chin, M. G. Hartell, W. C. Neely and V. Vodyanoy, Biosens. Bioelectron., 2000, 15, 135. 19 H. Li, S. Chen and H. Zhao, Biophys. J., 1990, 58, 1313. 20 B. Friguet, A. Fedorov, A. A. Serganov, A. Navon and M. E. Goldberg, Anal. Biochem., 1993, 210, 344. 21 J. E. Butler, J. Immunol. Methods, 2000, 22, 4. 22 N. Tajima, M. Takai and K. Ishihara, Anal. Chem., 2011, 83, 1969.
Published on 04 December 2013. Downloaded by Aston University on 10/01/2014 11:37:59.
14 H. Lee, S. M. Dellatore, W. M. Miller and P. B. Messersmith, Science, 2007, 318, 426. 15 H. Lee, J. Rho, K. Holmburg and P. B. Messersmith, Adv. Mater., 2009, 21, 431. 16 S. Puertas, P. Batalla, M. Moros, E. Polo, P. Pino, J. M. Guisan, V. Grazu and J. M. Fuente, ACS Nano, 2011, 5, 4521. 17 S. Allen, X. Chen, J. Davies, M. C. Davies, A. C. Dawkes, J. C. Edwards, C. J. Roberts, J. Seon, S. J. B. Tendler and P. M. Williams, Biochemistry, 1997, 36, 7457.
Paper
Analyst
This journal is © The Royal Society of Chemistry 2014
