Article pubs.acs.org/Langmuir
Development of a Novel Antifouling Platform for Biosensing Probe Immobilization from Methacryloyloxyethyl PhosphorylcholineContaining Copolymer Brushes Piyaporn Akkahat,†,‡ Suda Kiatkamjornwong,§ Shin-ichi Yusa,∥ Voravee P. Hoven,*,⊥ and Yasuhiko Iwasaki*,# †
Program in Petrochemistry, Faculty of Science, Chulalongkorn University, Phayathai Road, Pathumwan, Bangkok 10330, Thailand Center for Petroleum, Petrochemicals, and Advanced Materials, Chulalongkorn University, Phayathai Road, Pathumwan, Bangkok 10330, Thailand § Department of Imaging and Printing Technology, Faculty of Science, Chulalongkorn University, Phayathai Road, Pathumwan, Bangkok 10330, Thailand ∥ Department of Materials Science and Chemistry, Graduate School of Engineering, University of Hyogo, 2167 Shosha, Himeji-shi, Hyogo 671-2201, Japan ⊥ Organic Synthesis Research Unit, Department of Chemistry, Faculty of Science, Chulalongkorn University, Phayathai Road, Pathumwan, Bangkok 10330, Thailand # Faculty of Chemistry, Materials and Bioengineering, Kansai University, 3-3-35 Yamate-cho, Suita-shi, Osaka 564-8680, Japan ‡
S Supporting Information *
ABSTRACT: The immobilization of thiol-terminated poly[(methacrylic acid)-ran-(2-methacryloyloxyethyl phosphorylcholine)] (PMAMPC-SH) brushes on gold-coated surface plasmon resonance (SPR) chips was performed using the “grafting to” approach via self-assembly formation. The copolymer brushes provide both functionalizability and antifouling characteristics, desirable features mandatorily required for the development of an effective platform for probe immobilization in biosensing applications. The carboxyl groups from the methacrylic acid (MA) units were employed for attaching active biomolecules that can act as sensing probes for biospecific detection of target molecules, whereas the 2-methacryloyloxyethyl phosphorylcholine (MPC) units were introduced to suppress unwanted nonspecific adsorption. The detection efficiency of the biotin-immobilized PMAMPC brushes with the target molecule, avidin (AVD), was evaluated in blood plasma in comparison with the conventional 2D monolayer of 11-mercaptoundecanoic acid (MUA) and homopolymer brushes of poly(methacrylic acid) (PMA) also immobilized with biotin using the SPR technique. Copolymer brushes with 79 mol % MPC composition and a molecular weight of 49.3 kDa yielded the platform for probe immobilization with the best performance considering its high S/N ratio as compared with platforms based on MUA and PMA brushes. In addition, the detection limit for detecting AVD in blood plasma solution was found to be 1.5 nM (equivalent to 100 ng/mL). The results have demonstrated the potential for using these newly developed surface-attached PMAMPC brushes for probe immobilization and subsequent detection of designated target molecules in complex matrices such as blood plasma and clinical samples.
■
INTRODUCTION The ability to resist the nonspecific adsorption of nontargeted proteins is an important issue that must be taken into consideration when developing a sensor platform for detecting specific protein interactions, especially in complex samples such as blood plasma and clinical samples. Nonspecific adsorption usually leads to an undesirable outcome such as high background noise or a low signal-to-noise ratio (S/N).1,2 This situation often causes an adverse impact on biosensor efficiency. Thus, a well-performed sensor platform should not only allow for immobilization of bioactive species but should also resist © 2012 American Chemical Society
nonspecific interactions. One common approach to reducing background from nonspecific adsorption is to form a selfassembled monolayer (SAM) of functionalized alkanethiol with highly hydrophilic terminal groups such as oligo(ethylene glycol) (OEG)3−5 and phosphorylcholine (PC).6 However, this approach is not always effective, particularly in more complex samples.7,8 Received: October 28, 2011 Revised: February 7, 2012 Published: February 24, 2012 5872
dx.doi.org/10.1021/la204229t | Langmuir 2012, 28, 5872−5881
Langmuir
Article
continuously reported, most extensively by Ishihara and coworkers.16,32−35 They have found that the terpolymer of poly[2-methacryloyloxyethyl phosphorylcholine (MPC)-co-n-butyl methacrylate (BMA)-co-p-nitrophenyloxycarbonyl poly(ethylene glycol) methacrylate (MEONP)] or PMBN can physically adsorb on hydrophobic materials such as poly(methyl methacrylate) microchip35 and polystyrene microplate wells32,34 which can later be used to detect specific antigen by fluorescence technique based on ELISA assays and on poly(L-lactic acid) nanoparticles33 which should be applicable for high-affinity separation of proteins. The MEONP units having active ester entities of nitrophenyloxycarbonyl groups were used as binding sites for designated antibodies. The presence of MPC units apparently helped in reducing the nonspecific adsorption so that the analysis can be performed without the use of blocking reagents. Recently, they have also employed the surfacegrafted block copolymer brushes of poly(2-methacryloyloxyethyl phosphorylcholine-co-2-aminoethyl methacrylate), generated by SIP (grafting from) method, as a platform for binding antibodies having size variation (albumin, CRP, IgG, and fibrinogen) and determined the effect of antibody orientation on the efficiency of antigen detection. Of particular note, most of their studies were conducted in either phosphate buffer solutions16,32,34 or protein mixture solutions.33 In our research, the “grafting to” method was featured as an alternative method of preparing the sensor platform for biosensing applications. Thiol-terminated poly[(methacrylic acid)ran-(2-methacryloyloxyethyl phosphorylcholine)] (PMAMPCSH) synthesized by RAFT polymerization was grafted to the gold surface of an SPR sensor chip to yield a PMAMPCmodified gold-coated SPR chip. In the copolymer structure, the carboxyl groups from MA units were introduced as active sites for binding probes, whereas the MPC units were incorporated to function as hydrophilic entities that should help in suppressing nonspecific protein adsorption. In this investigation, biotin and avidin were used as the model for the sensing probe and target analyte, respectively. Parameters that might affect sensitivity and specificity in terms of the S/N ratio of the sensors for detecting target molecules in blood plasma were investigated by using the SPR technique. These issues are likely to be equally important in the development of sensor platforms for biosensing applications. It should be highlighted that SPR is a well-known optical-based method of which measurements are rapid and do not require specific labels. These should not only decrease the costs and time of detection but also reduces interference problems at the potential cost of the loss of an amplification stage. The fundamental knowledge gained from this research should be highly beneficial for the development of label-free biosensors, especially those having metal (i.e., gold, silver)-coated substrates and expand the applicability of the PMPC-containing polymers in the area that has not yet been much explored.
Recently, it was found that polymer brushes exhibit much lower protein adsorption not only from single-protein solutions but also from human blood plasma than does SAM having the same functional group.9,10 Furthermore, the polymer brushes, of which the number of binding sites for bioactive molecules can be controlled, effectively enhance the sensor response as a consequence of the high functional group density at the brush interface and thus provide more binding sites for bioactive species than do SAM-based platforms.11−19 Most strategies to establish functionalized polymer brushes onto surfaces are usually based on either “grafting to” or “grafting from” methods. The “grafting to” approach is generally accomplished by reacting a substrate with preformed end-functionalized polymer molecules. Among all controlled polymerization processes, reversible addition−fragmentation chain transfer (RAFT) polymerization has increasingly become an attractive route to producing the end-functionalized polymer brushes to be grafted on substrates. Upon aminolysis20,21 or reduction,22,23 the synthesized thiocarbonylthio end-capped polymers can be converted to thiol-terminated polymer chains that are readily available for attaching to gold surfaces,24 common substrates used in many bioanalytical devices.22,25 Previously, we have developed a sensing platform based on densely packed poly(acrylic acid) (PAA) brushes (0.21− 0.32 chain/nm2) prepared using a “grafting from” approach based on surface-initiated atom transfer radical polymerization (SI-ATRP) on an SPR sensor chip.11 The biotinylated PAA brushes showed a high specific binding with streptavidin (SA) and a low nonspecific adsorption of the negatively charged proteins (bovine serum albumin and fibrinogen) in comparison with a SAM of carboxyl-terminated alkanethiol as determined by SPR. However, the accessibility of the SA molecule to the immobilized biotin situated inside the polymer layer was quite limited, leading to low binding efficiency, particularly at high graft density of which the space between the polymer chains is narrow. The target molecules, especially those are large in size, are hindered from reaching the immobilized probes situated inside the inner layer of the polymer brushes. It is expected that this limitation due to the steric hindrance can be overcome by lower the grafting density of the closely packed, grafted polymer chains. In addition, being suffered from the nonspecific adsorption of the positively charged proteins, the surfacegrafted PAA brushes are not suitable for use in systems having positively charged components or complex samples. To overcome the problem caused by the limited accessibility of the target molecules to the immobilized probes, lowering the grafting density by using “grafting to” instead of “grafting from” seems to be a convenient solution. Despite its simplicity, the “grafting to” approach still suffers from unavoidable nonspecific protein adsorption, which is a consequence of the relatively low grafting density of the polymer brushes. In order to prevent the occurrence of such an undesirable event, highly hydrophilic polymers that possess nonfouling properties have been proven effective in this regard.26−30 In particular, poly[2-methacryloyloxyethyl phosphorylcholine] (PMPC), a biocompatible and antithrombogenic polymer, can be a potential candidate because it is not only capable of effectively reducing the nonspecific adsorption of plasma proteins and cells but was also found to accommodate a better environment for the preservation of immobilized proteins during prolonged storage so they maintained their activity and stability.31 The success of PMPC in preventing adsorption of nontarget components for detection of antigen−antibody specific interactions has been
■
EXPERIMENTAL SECTION
Materials. 2-Methacryloyloxyethyl phosphorylcholine (MPC) was purchased from NOF Corp. (Japan). Methacrylic acid (MA) was distilled under reduced pressure with added p-methoxyphenol (59 °C/ 13.5 mmHg). 11-Mercaptoundecanoic acid (HOOC(CH2)11SH or MUA), bovine serum albumin (BSA), avidin (AVD, from egg white), antibovine serum albumin (anti-BSA, developed in rabbit), and phosphate buffered saline (PBS) were purchased from Aldrich. Human platelet-poor plasma (PPP) was donated from a healthy volunteer. 1-(3-(Dimethylamino)propyl)-3-ethylcarbodiimide hydrochloride (EDC) and N-hydroxysuccinimide (NHS) were purchased from 5873
dx.doi.org/10.1021/la204229t | Langmuir 2012, 28, 5872−5881
Langmuir
Article
Fluka. (+)-Biotinyl-3,6,9-trioxaundecanediamine (biotin-NH2) was purchased from Thermo Scientific Co., Ltd. Gold-coated SPR chips were purchased from Moritex Co. (Kanagawa, Japan). Solutions were made with Milli-Q water purified by an ultrapure water system using a Millipak-40 filter unit (0.22 μm, Millipore) and Millipore Milli-Q system that involves reverse osmosis followed by ion exchange and filtration steps (18.2 MΩ). Synthesis of Thiol-Terminated (Co)polymer. MPC monomer (1.5 g, 5 mmol) was dissolved in 20 mL of mixed solvent (1:1, EtOH:PBS). After the MPC monomer was completely dissolved, MA monomer (0.43 g, 5 mmol), 4,4′-azobis(4-cyanovaleric acid) (ACVA, 18.78 mg, 0.067 mmol), and 4-cyanopentanoic acid dithiobenzoate (CPD, 55.87 mg, 0.2 mmol) were added to the monomer solution. The solution was bubbled with Ar gas for 30 min and then put in an oil bath at 70 °C for a set reaction time. The samples were withdrawn from the solution periodically to monitor the average molecular weight and molecular weight distribution (MWD) by GPC. The reaction was terminated by cooling the reaction mixture with an ice bath. The polymer solution was then purified by dialysis in DI water for 3 days and then lyophilized. In order to vary the composition and molecular weight of the copolymer, the mole ratio of the MA and MPC monomers in the feed and the ratio of monomer to CTA were varied. The composition of the copolymer was determined by 1H NMR. Homopolymers of PMPC and poly(methacrylic acid) (PMA) were prepared and purified using the same method as described above. The (co)polymers were characterized by 1H NMR and FT-IR. 0.5 g each of PMAMPC, PMPC, and PMA were dissolved in 5 mL of DI water. After the polymer completely dissolved in the water, a trace of tris(2-carboxyethyl)phosphine (TCEP) was added to the solution, and then 30 mol equiv of ethanolamine was added. The solution was stirred at ambient temperature for 1 h or until the polymer solution became colorless. The product was obtained after purification by dialysis in DI water for 3 days and lyophilization and then characterized by 1H NMR, FT-IR, GPC, and UV. Stepwise preparation of the thiol-terminated (co)polymers, PMAMPC-SH, PMPC-SH, and PMA-SH, is outlined in Figure 1a. Preparation of (Co)polymer-Modified Gold-Coated SPR Chip. An SPR chip was cleaned by oxygen plasma treatment for 2 min, immersed in Milli-Q water for 2 min, dried under a nitrogen stream, and finally immersed in a 0.2 mM of PMAMPC-SH solution in
PBS buffer (10 mM, pH 7.4) for 48 h. The substrates were then removed from the solution, rinsed by constant agitation in PBS buffer containing 1% SDS and Milli-Q water for 22 and 2 h, respectively, and dried under a nitrogen stream. The homopolymer brushes of PMPCSH and PMA-SH were also immobilized on the gold-coated SPR chip by using the same procedure as described above. The modified goldcoated SPR chip was characterized by XPS, water contact angle, ATRFTIR, and AFM. Contact Angle Measurements. The advancing and receding water contact angles were measured using a contact angle goniometer (First Ten Angstroms FTÅ125 goniometer), equipped with a Gilmont syringe and a 24-gauge flat-tipped needle. All measurements were performed in air at ambient temperature. A silhouette image of the droplet was projected onto a back screen from which the contact angle at the solid/liquid/air interface was determined. The dynamic advancing (θA) and receding contact angles (θR) were recorded while a small amount of water was either added to or withdrawn from the droplet. Data for each sample were taken from five different areas of the substrate, after which they were expressed as the arithmetic mean ± standard deviation (SD). Gel Permeation Chromatography (GPC). The molecular weight and the molecular weight distribution of the PMA, PMPC, and their copolymers were characterized with a Tosoh GPC system with an RI detector and size-exclusion columns, Shodex SB-804 HQ and SB-806 HQ. The column was eluted with distilled water containing 10 mM LiBr at a flow rate of 0.75 mL/min. Poly(ethylene glycol) (PEG, Tosoh) standards were used for generating a calibration curve. Atomic Force Microscopy (AFM). The morphology and thickness of the samples were analyzed by AFM recorded with a NanoScopeIV scanning probe microscope (Veeco). Measurements were performed in air using tapping mode. Silicon nitride tips with a resonance frequency of 267−295 kHz and a spring constant of 20− 80 N/m were used. X-ray Photoelectron Spectroscopy (XPS). The surface composition was characterized by XPS on a Scienta ESCA 200 spectrometer (Uppsala, Sweden) with Al Kα X-rays. All XPS data were collected at takeoff angles of 15° and 90°. Attenuated Total Reflection-Fourier Transform Infrared (ATR-FTIR) Spectroscopy. The structural information on the polymer brushes on the SPR sensor chip was obtained by a Nicolet
Figure 1. Schematic diagram showing (a) synthesis of the thiol-terminated (co)polymers and (b) PMAMPC-modified gold-coated SPR chip immobilized with biotin. 5874
dx.doi.org/10.1021/la204229t | Langmuir 2012, 28, 5872−5881
Langmuir
Article
6700 FT-IR spectrometer. Spectra in the infrared region (4000−650 cm−1) were collected with 64 scans at a spectral resolution of 4 cm−1. SPR Measurements. SPR measurements were conducted using a dual-channel surface plasmon resonance sensor system (SPR-670M; Moritex Inc., Kanagawa, Japan) at room temperature with the plane face of the prism coupled to the gold-coated glass via index-matching fluid. PBS (10 mM, pH 7.4) was used as the running buffer at a constant flow rate of 15 μL min−1 and held at a constant temperature of 30 °C. An automatic flow injection system was used to inject the test solutions. Measurement of the SPR angle shift was done under nonflow liquid conditions. For comparison, a gold-coated SPR chip bearing a monolayer of carboxyl-terminated thiol, 11-mercaptoundecanoic acid (MUA), was prepared by immersing a cleaned disk in an ethanolic solution of MUA (1 mM) at ambient temperature for 24 h. The disk was then rinsed thoroughly with ethanol and dried in a stream of nitrogen. The modified gold-coated SPR chip was first seated in the SPR cell before being stabilized with a running solution of PBS buffer (10 mM, pH 7.4) until the equilibrium SPR angle frequency was reached in buffer solution. The sensor was thereby ready for use. Protein Adsorption Study. Protein adsorption on a modified gold-coated SPR chip was tested with three proteins, BSA, AVD, and blood plasma in PBS buffer (10 mM, pH 7.4). After the modified-SPR chip was seated in the SPR cell and the baseline SPR response was recorded in running buffer, 0.1 mg/mL each of BSA and AVD solution was injected over the substrate. The chip was left for 10 min followed by a 5 min rinse with PBS. In the case of blood plasma, whole blood plasma carrying about 70 mg of proteins per milliliter of plasma was diluted by PBS buffer at a plasma to PBS ratio of 1:700 to obtain blood plasma protein with a concentration of 0.1 mg/mL or equivalent to 0.14% blood plasma in PBS. The blood plasma solution was allowed to flow over the substrate for 15 min, which was then washed with PBS buffer for 5 min. Immobilization of Biotin Probes onto Sensor Surface. The immobilization of biotin was done outside the SPR instrument. The carboxyl groups were first activated by immersing gold-coated SPR chips bearing PMAMPC brushes, PMA brushes, and MUA in an aqueous solution of 0.2 M EDC and 0.05 M NHS for 15 min. The substrates were then rinsed immediately with Milli-Q water before being immersed in a solution of biotin-NH2 (1 mg/mL) in PBS buffer for 2 h at ambient temperature and washed thoroughly with PBS solution. The chemical structure of the PMAMPC-modified goldcoated SPR chip immobilized with biotin is shown in Figure 1b. The unreacted activated carboxyl groups (having NHS groups attached) were deactivated by being immersed in PBS buffer (10 mM, pH 7.4) overnight. Specific Interactions of Immobilized Biotin Probes with Avidin in Complex Sample. Gold-coated SPR chips bearing PMAMPC brushes, PMA brushes, and MUA immobilized with biotin were first seated in the SPR cell before being rinsed with a running solution of PBS buffer. After a baseline SPR response was established, AVD (10 μg/mL equivalent to 0.15 μM) in blood plasma solution (0.1 mg/mL or 0.14% in PBS buffer) was applied and left for 10 min.
Unbound AVD was removed by washing with PBS for 5 min. The amount of specific binding of AVD was quantified by the shift of the SPR response angle at the end-point of the washing step and after baseline subtraction. Nonspecific binding (binding in the absence of AVD in blood plasma solution) was also determined in order to quantify the specific binding of AVD in blood plasma in terms of the signal-to-noise (S/N) ratio. The signal-to-noise (S/N) ratio was calculated by the equation S/N = (SPR angle shift after exposure to blood plasma with AVD)
■
/(SPR angle shift after exposure to blood plasma without AVD)
(1)
RESULTS AND DISCUSSION Synthesis of Thiol-Terminated Copolymer. Synthesis of PMAMPC copolymer using RAFT polymerization was performed in the presence of 4-cyanopentanoic acid dithiobenzoate (CPD) and 4,4′-azobis(4-cyanovaleric acid) (ACVA) as chain transfer agent (CTA) and radical initiator, respectively. The composition of the MA and MPC units in the copolymer is designated as x and y, respectively. The copolymer identity is then written as PMAxMPCy. Table 1 gives an overview of the reaction conditions and the molecular weight information on copolymers prepared by RAFT polymerizations under different conditions. The ratio of CTA/initiator or [CTA]/[I] of 4/1 was fixed, while the ratio of monomer/CTA and copolymer ratio were varied. The data suggest that the copolymer composition determined by 1H NMR (Figure S1 in Supporting Information) closely resembles the copolymer ratio in the feed. In other words, the copolymer composition can be controlled by varying the mole fraction of monomer in the feed. The molecular weight (M̅ n) increased with an increase of monomer/CTA ratios and polymerization time. In addition, a controlled character of this polymerization can also be realized from the data shown in Figure 2. The semilogarithmic plot indicates that the polymerization is a first-order reaction with respect to the MPC comonomer. The linearity of the first-order plot of the MPC concentration implies that the concentration of polymer radical remains constant for the duration of the polymerization. The dithiobenzoate group at the chain end of PMAMPC can be converted to a thiol group by aminolysis using ethanolamine.20,21 The disappearance of aromatic proton peaks around 7.4−8.2 ppm in 1H NMR and UV absorbance of the dithiobenzoate group at 305 nm after aminolysis indicated that the dithiobenzoate group at the chain end of PMAMPC was removed by aminolysis yielding PMAMPC-SH having a terminal thiol group (Figures S1 and S2 in Supporting Information). The GPC trace of PMAMPC before and after hydrolysis,
Table 1. Summary of Reaction Conditions and Molecular Weight Information of PMAMPC Copolymers Synthesized by RAFT Polymerization MPC content (mol %) abbreviation
[MPC]/ [CTA]
in feed
in copolymer
time (h)
M̅ n (×103)
M̅ w (×103)
PDI
PMA25MPC75 PMA26MPC74 PMA21MPC79 PMA39MPC61 PMA39MPC61 PMA37MPC63 PMA55MPC45 PMA57MPC43 PMA66MPC34
50 50 200 50 50 200 50 50 200
70 70 70 50 50 50 30 30 30
75 74 79 61 61 63 45 43 34
2 8 8 2 8 8 2 8 8
10.8 24.7 49.8 12.0 25.9 54.5 6.60 29.3 49.5
14.4 30.1 61.5 14.9 31.5 67.0 11.3 35.4 61.3
1.33 1.21 1.24 1.24 1.21 1.23 1.69 1.21 1.24
5875
dx.doi.org/10.1021/la204229t | Langmuir 2012, 28, 5872−5881
Langmuir
Article
XPS, water contact angle measurements, and ATR-FTIR. Figure 3 shows the XPS spectra of the SPR chip that was adsorbed with PMA21MPC79 (49.8 kDa) and PMA21MPC79-SH (49.3 kDa) as compared with the substrate before adsorption (bare gold). Determined at a takeoff angle of 15°, phosphorus (P2p) and nitrogen (N1s) signals attributed to the phosphorylcholine group of the MPC units were only observed on the chip adsorbed with the thiol-terminated copolymer (PMA21MPC79-SH), indicating that the copolymer itself cannot be strongly bound to the gold-coated SPR chip unless modified with the thiol group. The signal from S2p of bound thiol at a binding energy of 161 eV (determined at a takeoff angle of 90°) was strongest on the substrate adsorbed with PMA21MPC79SH. The fact that there were peaks corresponding to S2p appearing on the unmodified disk and that was adsorbed with PMA21MPC79 implied that there might be a trace amount of sulfur-containing small molecule contaminants that cannot be completely removed after polymerization and aminolysis. Table 2 lists the wettability of the gold-coated SPR chips after being adsorbed with the copolymer, PMAMPC-SH, and
Figure 2. Percentage of conversion (●, ■) and semilogarithmic plots of MPC conversion (○, □) as a function of time: MPC/CTA ratio of 50 (●, ○) or 200 (■, □).
shown in Figure S3 of the Supporting Information, also demonstrates that the molecular weight characteristic remained almost unchanged with a unimodal distribution. No bimodal distribution was observed, implying that PMAMPC-S-S-PMAMPC, which might occur as a result of disulfide coupling, did not form. These results suggest that aminolysis did not cause any degradation and/or coupling. Preparation of (Co)polymer-Modified Gold-Coated SPR Chip. The thiol-terminated polymer chain promotes the rapid covalent attachment to the gold surface by the gold− sulfur (Au−S) bonds with high affinity. This process, a so-called “grafting to” method, provides the formation of polymer brushes on the gold surface.21,22,27,28 A freshly cleaned goldcoated SPR chip was immersed in the solution of thiol-terminated polymer, and the physically adsorbed copolymer was removed by rinsing with copious amounts of surfactant. The presence of PMAMPC brushes on the SPR chip was verified by
Table 2. Water Contact Angles of Surface-Modified and Gold-Coated SPR Disks sample bare gold PMA21MPC79 PMA21MPC79-SH PMPC-SH PMA-SH
M̅ n (×103) 49.8 49.3 5.45 21.3
θA (deg) 100 70 26 22 66
± ± ± ± ±
5.3 4.8 2.1 4.4 5.3
θR (deg) 80 30 13 10 27
± ± ± ± ±
7.1 5.7 2.7 2.3 8.5
the homopolymer counterparts, PMA-SH and PMPC-SH. Extremely hydrophilic characteristic of the surface-grafted PMPC can be realized from its low advancing (θA) and receding (θR) water contact angles. The high composition of
Figure 3. XPS spectra of gold-coated SPR chips: (a) bare gold and after being adsorbed with (b) PMA21MPC79 (49.8 kDa) and (c) PMA21MPC79SH (49.3 kDa). 5876
dx.doi.org/10.1021/la204229t | Langmuir 2012, 28, 5872−5881
Langmuir
Article
topographically smooth and completely covered with the copolymer. The thickness of the polymer film (t) can be obtained from cross-section analysis and used for calculating graft density (σ) using the equation
PMPC (79%) in the copolymer made the surface-grafted PMAMPC very hydrophilic as well, especially when compared with the surface-grafted PMA. The lower water contact angle of the gold-coated SPR chip after exposure to the non-thiolated PMA21MPC79 suggested that there might be nonspecific adsorption. This observation is in accordance with the XPS data previously shown. It was also found that the water contact angles (both θA and θR) decreased with increasing M̅ n. This coincides with the fact that the copolymer with high M̅ n contained a greater number of hydrophilic MPC units on the surface. A similar explanation can be applied in the case where the water contact angles were lower as the greater content of MPC units was incorporated in the copolymers (Figures S4 and S5 in Supporting Information). Analysis by ATR-FTIR also confirmed the adsorption of PMAMPC-SH on gold-coated SPR chips. The characteristic absorption bands of both MPC units (CO (ester) = 1728 cm−1, OP−O−asym = 1264 cm−1, OP−O−sym = 1094 cm−1, −N+(CH3)3 = 975 cm−1, and C−O−P = 810 cm−1) and MA units (CO (acid) = 1728 cm−1, and C−O = 1181 cm−1) appeared in the spectra shown in Figure 4. In addition, the increment of C−O stretching of the MA unit was observed when the content of the MA unit in the copolymer was raised (see Figure 4c−e for comparison).
σ=
t ρNA 1 = M̅ n AX
(2)
where ρ is the mass density (1.3 g/cm for PMAMPC), M̅ n is the molecular weight of the free polymer, and NA is Avogadro’s number. Although the thickness of the PMAMPC brushes increased with increasing M̅ n, the graft density was inversely proportional to M̅ n (Figure 5). This is not surprising 3
Figure 5. Thickness and graft density of surface-grafted PMAMPC brushes having varied molecular weights.
considering that the grafting to method usually suffers from the entropic barrier due to crowding of the initial grafting polymer chains that prevent further insertion of the polymer. Such limitation becomes even more problematic for polymers having high molecular weight. Nonetheless, the graft density of surface-grafted PMAMPC brushes ranged from 0.13 ± 0.02 to 0.27 ± 0.07 chains/nm2 for M̅ n from 49.3 to 12.1 kDa, respectively. These values still fall within an extended brush regime (graft density >0.08 chains/nm2),36 suggesting that PMAMPC brushes with reasonable graft density can be formed despite their preparation being based on the “grafting to” method. According to our calculation, the distance between the grafted chains (d) which can be determined from (graft density)−1/2 is 3.2 nm for the lowest graft density of ∼0.13.37 In principle, polymer brushes can be introduced as thin films of end-grafted polymer molecules (under nonsolvated condition) when the following condition is satisfied: d < (r2)1/2, where r is designated as end-to-end distance.38,39 Outside these conditions the grafted layers are considered in the “mushroom regime. If the polymer chains are in the stretched conformation, r2 should be equivalent to degree of polymerization (N) which is 200 in this case. The fact that the distance between grafted chains (3.2 nm) is less than (200)1/2 indicates that the grafted polymer chains adopt extended conformation. In the case of PMAMPC brushes having a different copolymer composition, the graft density was assumed the same given that their M̅ n values are comparable. Protein Adsorption Test. The protein adsorption of PMAMPC brushes was investigated against BSA (60 kDa), AVD (66 kDa), and human blood plasma as compared with the substrates of bare gold and MUA. In PBS solution of pH 7.4, BSA and AVD with pI of 4.8 and 10.5 should bear negative and positive charges, respectively. Blood plasma representing a complex matrix is a yellow liquid component of blood
Figure 4. ATR-FTIR spectra of gold-coated SPR chips: (a) bare gold and after being adsorbed with (b) PMA21MPC79 (49.8 kDa), (c) PMA21MPC79-SH (49.3 kDa), (d) PMA37MPC63-SH (53.7 kDa), and (e) PMA66MPC34-SH (49.1 kDa).
Additional evidence of the success in surface grafting of PMAMPC brushes on gold-coated SPR chips can also be seen in an AFM image (Figure S6 in Supporting Information). The roughness of ∼2.7 nm suggested that the gold surface should be 5877
dx.doi.org/10.1021/la204229t | Langmuir 2012, 28, 5872−5881
Langmuir
Article
the specific binding of the sensor in terms of the signal-to-noise (S/ N) ratio. From the viewpoint of biosensing efficiency, especially in the situation that nonspecific components concurrently exist with the target analyte, the selectivity, in principle, can be judged from the ability to detect the target analyte as opposed to the nonspecific components being present in the sample. Here in this research, the SPR angle shift due to the adsorption of nontarget components in the blood plasma is considered as background or noise (N) of the detection, whereas the SPR angle shift due to the adsorption of target analyte (AVD in this case) in the blood plasma is therefore considered as the signal of detection (S). The efficiency of the developed sensor platform can then be expressed in terms of S/N ratio which was calculated by eq 1. This similar strategy of calculation has been previously described by others.14 The biotin-immobilized MUA and PMA (21.3 kDa) brushes showed a greater binding quantity of AVD in blood plasma as compared with the PMA21MPC79 (49.3 kDa) brushes (Figure 7a).
containing hundreds of dissolved proteins such as albumin, fibrinogen, and globulin for about 7% (70 mg of proteins per milliliter of plasma) and other components such as water, inorganic ions, and organic substances. The results illustrated in Figure 6 indicate that the surfacegrafted PMPC and PMAMPC brushes on a gold-coated SPR
Figure 6. SPR angle shift corresponding to the amount of adsorbed proteins: BSA and AVD (0.1 mg/mL) in PBS solution (10 mM, pH 7.4) and 0.14% blood plasma (0.1 mg/mL) in PBS on surfacemodified SPR chip.
chip can completely resist nonspecific adsorption of BSA (negatively charged protein) and AVD (positively charged protein) as well as the multicomponent proteins in blood plasma as opposed to that of bare gold and MUA surfaces. On the other hand, surface-grafted PMA brushes could only prevent the adsorption of a negatively charged protein, BSA, but suffered extensively from the nonspecific adsorption of a positively charged protein, AVD, and the multicomponent proteins in blood plasma. The electrostatic binding with the positively charged protein of the PMA having pKa of 6.9−7.040 was similar to what has been previously observed in the system of surface-grafted PMA brushes.41 These results strongly indicated that the MPC units in the PMAMPC copolymer brushes play an important role in preventing nonspecific protein adsorption. This is in excellent agreement with many reports previously published on the fact that PMPC exhibits an outstanding resistance to nonspecific interactions with plasma proteins and cells.42−44 In particular, the random copolymer brushes of MPC and GMA (glycidyl methacrylate) also showed that the introduction of MPC units in the polymer brushes is effective in reducing the nonspecific adsorption of proteins.45 Specific Interactions of Immobilized Biotin with Avidin in Blood Plasma Solution. Biotin was chosen as a model of a sensing probe that can bind with AVD with high specificity and affinity, KD ≈ 10−15 M.46 (+)-Biotinyl-3,6,9trioxaundecanediamine (biotin-NH2) was immobilized on the PMAMPC copolymer brushes via an amide bond formation between the amino group of biotin-NH2 and the carboxyl group of the MA unit in the PMAMPC copolymer brushes.11 Immobilization was performed outside the SPR instrument. In order to investigate the possibility of using a PMAMPC sensor platform for detecting the target molecule in the complex sample, SPR analysis was conducted in 0.15 μM AVD solution (equivalent to 10 μg/mL) in diluted blood plasma (0.14% or 0.1 mg/mL). Nonspecific binding (binding in the absence of AVD in blood plasma) was also determined in order to quantify
Figure 7. SPR angle shift corresponding to the binding of components from 0.14% blood plasma and AVD 0.15 μM in 0.14% blood plasma and S/N ratio of AVD binding in blood plasma on different sensor platforms: (a) and on the sensor platforms based on PMA21MPC79 brushes having varied M̅ n (b) and MA:MPC content in copolymer (c). 5878
dx.doi.org/10.1021/la204229t | Langmuir 2012, 28, 5872−5881
Langmuir
Article
However, the nonspecific binding of blood plasma was also extremely high. Although the sensor platform based on the PMA21MPC79 (49.3 kDa) brushes gave the lowest AVD binding response, the nonspecific binding response was greatly suppressed. It was apparent that the sensor platform based on these copolymer brushes promoted specific binding over nonspecific binding, yielding the highest S/N ratio of AVD binding in blood. Enhancement of the S/N ratio was about 8.1- and 4.1fold as compared with the MUA and PMA brushes, respectively. To achieve the highest performance of PMAMPC brushes as a sensor platform in terms of specific binding, the effect of copolymer chain length as well as the proper balance between MA units and MPC units (copolymer composition) should be considered. Figure 7b shows that the signal of the AVD binding increased with an increase in M̅ n or thickness, supposedly due to an increase in sensing probe density, while the amount of nonspecific binding decreased. However, in the case of PMAMPC with low M̅ n (12.1 kDa), the strong influence of nonspecific adsorption was observed, showing large amounts of both specific and nonspecific adsorption. Evidently, the PMAMPC brushes with M̅ n of 49.3 kDa provided the highest S/N ratio followed by those with M̅ n of 25.1 and 12.1 kDa. These results indicate that the longer polymer chain not only promoted higher specific binding but also suppressed nonspecific binding. Increasing the MA content (from 21 to 37%) in the copolymer can help elevate the amount of AVD binding (Figure 7c). However, that also promotes the nonspecific binding of blood plasma. This can be described as a result of the proportional reduction of MPC content, which led to the deteriorated ability to prevent protein adsorption of the PMAMPC brushes. Further increasing the MA content to 66% did not increase the level of AVD binding, as seen in the case of PMA66MPC34. Presumably, this might be due to steric hindrance of a densely packed sensing probe resulting in a low S/N ratio. These results also imply that the amount of MA units in PMAMPC brushes of only about 20% was effective enough to provide a high efficiency sensor platform for detecting AVD in blood plasma solution. Limits of Detection (LODs). To determine the lowest detectable concentration of AVD in blood plasma, surfaceattached PMA21MPC79 (49.3 kDa) brushes, the sensor platform that gave the best performance, were used to detect AVD having different concentrations ranging from 0.15 to 150 nM in 0.14% blood plasma solution. The results illustrated in Figure 8 suggest that the binding signal increased as a function of AVD concentration and that the binding did not saturate even at a high AVD concentration of 150 nM. The lowest concentration of AVD in 0.14% blood plasma or the detection limit of this sensor platform that can distinguish between the signals of specific and nonspecific binding (S/N ≥ 2) was 1.5 nM (equivalent to 100 ng/mL) while the detection limit of the sensor platform obtained from MUA, a conventional sensor platform, was 150 nM (Figure S7 in Supporting Information). Although this figure is not considered superior given that lower LODs (
