Binding Kinetics of Biomolecule Interaction at Ultralow Concentrations

Apr 5, 2011 - Department of Optics and Photonics, National Central University, ... Biomedical Engineering Research Center, Chang Gung University, ...
1 downloads 0 Views 2MB Size
ARTICLE pubs.acs.org/ac

Binding Kinetics of Biomolecule Interaction at Ultralow Concentrations Based on Gold Nanoparticle Enhancement Li-Chen Su,†,‡ Ying-Feng Chang,§ Chien Chou,*,†,‡,§,|| Ja-an Annie Ho,^ Ying-Chang Li,† Li-Dek Chou,‡ and Cheng-Chung Lee† †

Department of Optics and Photonics, National Central University, Taoyuan, Taiwan, 320 Graduate Institute of Electro-Optical Engineering, Chang Gung University, Taoyuan, Taiwan, 333 § Institute of Biophotonics, National Yang Ming University, Taipei, Taiwan, 112 Biomedical Engineering Research Center, Chang Gung University, Taoyuan, Taiwan, 333 ^ Department of Biochemical Science and Technology, National Taiwan University, Taipei, Taiwan, 106

)



bS Supporting Information ABSTRACT: Measuring the kinetic constants of protein protein interactions at ultralow concentrations becomes critical in characterizing biospecific affinity, and exploring the feasibility of clinical diagnosis with respect to detection sensitivity, efficiency and accuracy. In this study, we propose a method that can calculate the binding constants of proteinprotein interactions in sandwich assays at ultralow concentrations at the pg/mL level, using a localized surface plasmon coupled fluorescence fiber-optic biosensor (LSPCF-FOB). We discuss a twocompartment model to achieve reaction-limited kinetics under the stagnant conditions of the reaction chamber. The association rate constant, dissociation rate constant, and the equilibrium dissociation constant, that is, ka, kd, KD, respectively, of the kinetics of binding between total prostate-specific antigen (t-PSA) and anti-t-PSA at concentrations from 0.1 pg/mL to 1 ng/mL, were measured either in PBS or in human serum. This is the first time that ka, kd, and KD have been measured at such a low concentration range in a complex sample such as human serum.

B

iomolecular binding kinetics provides information that may be qualitative, quantitative, static, or dynamic and that can be applied to the field of protein detection and separation and to function studies.14 The information is beneficial in basic research, biomedicine, biomarker discovery, and the pharmaceutical industry.1,5,6 To be clinically useful, however, the technologies for real-time monitoring biomolecular interactions must analyze low abundance proteins in serum, plasma, urine, tissues, and tumor interstitial fluid, be sufficiently specific and sensitive to support diagnostic monitoring applications.1,5 One of the most popular and efficient methods to measure the kinetics is the immunosensor, which belongs to the type of highly sensitive biosensor that uses the selective affinity between antibodies and antigens or compounds.7,8 Conventionally, binding kinetics of the interaction of biomolecules can be analyzed using a solidphase immobilized antibody and an antigen in the liquid phase or vice versa.8 Generally, different techniques can be used to monitor proteinprotein interactions in real time that allows us to visualize directly the affinity ranking, estimate comparatively the observed rate constants, and determine the equilibrium r 2011 American Chemical Society

rate constants.2,3,912 Among them, surface plasmon resonance (SPR) is the most common techniques used for this purpose.1012 However, this method is based on a single antibody assay, and the binding kinetics between the analyte and the receptor would not be analyzed if the analyte exists in a complex matrix like serum, because of interference by the highly nonspecific binding from the background. In our previous study, we derived and analyzed the kinetics of the interaction of biomolecules, based on the rapid-mixing model,13 and were able to calculate the association and dissociation rate constants from the measurement of binding kinetics by using a fluorescence-detection fiber-optic biosensor (FD-FOB).14,15 The rapid-mixing model is under reaction rate-limited conditions of interaction, and the analyte concentration is uniform around the sensor surface.13 Unfortunately, this setup cannot measure proteinprotein interactions at ultralow concentrations at the pg/mL Received: November 1, 2010 Accepted: March 21, 2011 Published: April 05, 2011 3290

dx.doi.org/10.1021/ac1028616 | Anal. Chem. 2011, 83, 3290–3296

Analytical Chemistry

ARTICLE

Figure 1. Scheme of the sandwich complex and LSPCF probe used in this study.

level. As a result, it becomes difficult in proteomics, and for disease/pharmacokinetic biomarker monitoring, to study biointeractions with extremely low affinity taken from either environmental or clinical samples.16,17 In this experiment, we proposed a method to meet the requirement of ultralow concentrations under conditions of reaction-limited kinetics. We performed an experimental demonstration on proteinprotein binding kinetics at the pg/mL level with a sandwich immunoassay, using a localized surface plasmon coupled fluorescence fiber-optic biosensor (LSPCF-FOB). In recent years, localized surface plasmons (LSPs) have been introduced into biosensing, and have proved to be one of the most efficient ways of increasing the detection limit of biosensors.1824 This is attributed to their characteristics of wavelength-selective absorption, with extremely large molar extinction coefficients, and significant enhancement of the localized electromagnetic field within 5060 nm of the surface of noble-metal nanoparticles.18,2224 In LSPCF-FOB, the fluorescence of a fluorophore-labeled detection antibody is excited with high efficiency by the LSPs near the gold nanoparticle (GNP) surface. The LSPs are excited by the evanescent wave that is generated at the stripped fiber surface and penetrates within a wavelength of the incident laser through total internal reflection (TIR). Thus, the unbound fluorophore-labeled detection antibody beyond the sensing volume of the evanescent wave will not contribute to the fluorescent signal, allowing real-time monitoring of the localized biomolecular interaction and measurement of the binding kinetics near the fiber surface. In addition, sandwich assays using two distinct antibodies that recognize different epitopes of the target antigen can largely diminish the probability of intervention from nonspecific binding from similar molecules.1,25 Therefore, the LSPCF-FOB could be an important device for performing real-time detection with high sensitivity and high specificity. It allows us not only to measure the biomolecular binding kinetics, such as the association and dissociation rate constants (ka and kd), but also to identify the specific biomolecules in biomedical applications at ultralow concentrations at the pg/mL level.

’ EXPERIMENTAL SECTION Materials. All solvents and chemicals used in this study were of analytical grade. Ethyl acetate, 2-propanol, bovine serum albumin (BSA), mouse IgG, and goat antimouse IgG were purchased from

Sigma Co. (St. Louis, MO, USA). In addition, Cy5-conjugated rabbit antimouse IgG was purchased from Chemicon (Billerica, MA, USA). GNP conjugated with protein A (Au-PA, GNP diameter = 25 nm) were purchased from Aurion (Wageningen, Netherlands). Antibodies to total prostate-specific antigen (t-PSA), and human t-PSA purified protein were purchased from Meridian Life Science, Inc. (Saco, ME, USA) and Chemicon, respectively. The fluorescent labeling kit (Lightning-Link Atto633) was purchased from Innova Biosciences (Cambridge, U.K.). The plastic optical fiber was purchased from Mitsubishi Rayon Co., Ltd. (Tokyo, Japan). Sera samples were obtained from Taipei Veterans General Hospital (Taipei, Taiwan). Preparation of the Optical Fiber. A plastic (polymethyl methacrylate, PMMA) multimode optical fiber, 1 mm in diameter, was used in this experiment. The fiber was stripped off the cladding by immersing it in ethyl acetate. Subsequently, the stripped portion of fiber was washed with 2-propanol to clean its surface. Chemical adsorption was carried out using covalent binding forces to immobilize the capture antibody on the surface of the stripped part of the fiber. The chemical adsorption protocol has been described elsewhere.2224 The modified area of the optical fiber was coated with a 250 μL volume of 4 μg/mL anti-t-PSA and goat antimouse IgG at room temperature for 2 h, corresponding to the experiments involving the detection of PSA and IgG, respectively. After blocking with 10 mg/mL BSA-PBS solution for 1 h, 250 μL of serially diluted t-PSA in PBS, the serially diluted t-PSA of 10-fold diluted healthy human serum, and serially diluted mouse IgG in PBS solution were added into each reaction chamber separately. The mixtures were incubated at room temperature for 2 h to form the Æcapture antibody/target antigenæ complex on the modified surface of the optical fiber, as shown in Figure 1. For the experiment of oriented immobilization of t-PSA, the modified region of optical fiber surface was first coated with 250 μL of a 4 μg/mL PA solution at 4 C overnight before being incubated with capture antibodies (anti-t-PSA). Similarly, samples were immersed in 10 mg/mL BSA-PBS solution for 1 h to block nonspecific binding. Then, serially diluted t-PSA in PBS was added into each reaction chamber. Finally, the fibers were washed five times before testing. Preparation of the LSPCF Probes. The LSPCF probe was composed of fluorophore-labeled detection antibodies and Au-PA, as shown in Figure 1. The detection antibody was Lightning-Link Atto633-conjugated anti-t-PSA and Cy5-conjugated rabbit antimouse IgG for the experiments with t-PSA and mouse IgG 3291

dx.doi.org/10.1021/ac1028616 |Anal. Chem. 2011, 83, 3290–3296

Analytical Chemistry

ARTICLE

Figure 2. Scheme of a two-compartment model.

Figure 3. Capture antibodies were immobilized on the modified region of the stripped optical fiber (a) with PA (b) without PA.

detection, respectively. To produce the LSPCF probes, 1 μg/mL of the detection antibody was mixed with Au-PA at a concentration of 3.76  109 particles/mL, and incubated for 10 h at 4 C in the dark. Measurement. For the experiments on PSA detection, a 632.8 nm HeNi laser was used as the excitation light source and a microscope objective (10, NA = 0.27) was used to match the numerical aperture of the fiber for better coupling efficiency. The fluorophore (Atto633), whose peak wavelength in the absorbance spectrum was 629 nm, was excited by the localized electromagnetic field close to the GNP surface during measurement. A long-pass filter was introduced into this setup to reduce the background noise. Then, the fluorescent signal was detected by using a photomultiplier tube (PMT). Finally, a lock-in amplifier was introduced to detect the fluorescence to enhance the signal-to-noise ratio (SNR). For the experiment on detection of mouse IgG, the excitation light source was a 658 nm diode laser because of the use of Cy5 as a fluorophore, and a band-pass filter with a central wavelength 680.4 nm was adopted to allow fluorescence detection via a PMT. Finally, a lock-in amplifier was introduced to detect the fluorescence. Kinetic Analysis of Biomolecule Binding. The LSPCF-FOB forms part of a sandwich assay using the fiber-immobilized

Figure 4. Linear relationships between the concentrations of initial t-PSA (a) with PA (b) without PA in PBS versus the concentrations of the Æt-PSA/LSPCF probeæ complex at the quasi-steady state.

Table 1. Kinetic Constants, KD, ka, and kd, of t-PSA Interaction with Anti-t-PSA in Different Experiment Were Estimated by the Analytic Model

PAþt-PSA in PBS t-PSA in PBS t-PSA in 10-fold

KD (M)

ka (M1 S1)

1.6  109

kd (S1)

(1.0 ( 0.3)  107

(1.6 ( 0.6)  102

9

(1.6 ( 0.4)  10

7

(1.8 ( 0.5)  102

9

(1.9 ( 0.6)  10

7

(1.9 ( 0.6)  102

1.2  10 1.0  10

diluted serum

capture antibody, antigen and LSPCF probe to create a Æcapture antibody/antigen/LSPCF probeæ complex, the fluorescence of 3292

dx.doi.org/10.1021/ac1028616 |Anal. Chem. 2011, 83, 3290–3296

Analytical Chemistry

ARTICLE

Figure 5. Linear relationship between the concentrations of initial mouse IgG versus the concentrations of the Æmouse IgG/LSPCF probeæ complex at the quasi-steady state.

which can only be excited by LSPs that are generated by the evanescent wave at the fiber surface. This allows us to measure the binding kinetics of interaction between the target antigen and the detection antibody, using an LSPCF-FOB by recording the fluorescent signal as a function of time. In addition, to calculate the kinetic association and dissociation rate constants, the following conditions are assumed to be satisfied experimentally: (1) the interaction of antigen/detection antibody has first-order kinetics; (2) there are a limited number of antigen binding sites, and there is no interaction between the binding sites;26 (3) the dissociation between the antigen and immobilized capture antibody is ignored under conditions where the surface density of the antigen is lower than that of the capture antibody;26 (4) the rebinding of antigen and detection antibody can be disregarded because the surface density of the antigen is lower than that of the detection antibody.12,27 A two-compartment model that is typically used to analyze the interaction in the SPR system, as shown in Figure 2,12,27,28 is employed in this measurement at ultralow concentration. After the injection of LSPCF probes into the reaction chamber, the LSPCF probes are transported into the inner compartment, and interact with the target antigens on the fiber surface. According to the given conditions in this experiment of a stagnant configuration of the reaction chamber, the concentration of LSPCF probes is designed to be much higher than that of the target antigen in the range from nanograms per milliliters to picograms per milliliters; this condition indicates that the mass transport flux is much faster than the reaction rate.2729 In the meantime, the concentration of Æantigen/LSPCF probeæ complex is much lower than that of the LSPCF probes in the outer compartment. Thus, the volume concentration of LSPCF probes rapidly becomes a uniform distribution in the reaction chamber. This means that the biomolecular reaction is limited by the interaction kinetics.

As a result, the rate equation for the proteinprotein binding kinetics can be written as d½AB ¼ ka ½A 0 ½B0   ðka ½A 0  þ kd Þ½AB dt

ð1Þ

where [AB] is the time-dependent concentration of the Æantigen/LSPCF probeæ complex, which is proportional to the intensity of the fluorescence signal of the measurement, [A0] is the concentration of the injected LSPCF probes, and [B0] is the surface concentration of the initial target antigen, averaged over the fiber surface. In addition, ka is the association rate constant describing the rate of molecular-complex formation between the antigen and the detection antibody of LSPCF probe, and kd is the dissociation rate constant involving the stability of the complex of the sandwich assay. The reaction under quasi-steady-state approximation,28,30 that is, d[AB]/dt =0, eq 1 gives the expression as  1 kd þ1 ½ABeq  ¼ ½B0  ð2Þ ka ½A 0  where [ABeq] is the concentration of the Æantigen/LSPCF probeæ complex, averaged over the fiber surface at the quasisteady state. [ABeq] is measured by the detected fluorescence signal and [B0] is linearly dependent on [ABeq] in eq 2. KD = kd/ka is the equilibrium dissociation constant.15 Additionally, eq 1 presents a linear relationship between d[AB]/dt and [AB] under the condition of reaction-limited kinetics satisfied during measurement in the transit state. Therefore, a combination of the transient state and quasi-steady state of the detected fluorescence signal versus time at different concentration of target antigen in the measurement, can provide (ka, kd, and KD) from the slopes of eqs 1 and 2 detection accordingly. 3293

dx.doi.org/10.1021/ac1028616 |Anal. Chem. 2011, 83, 3290–3296

Analytical Chemistry

ARTICLE

Figure 6. Linear relationship between the concentrations of initial t-PSA in diluted serum versus the concentrations of the Æt-PSA/LSPCF probeæ complex at the quasi-steady state.

’ RESULTS AND DISCUSSION Effect of Random and Oriented Immobilization of Capture Antibody on Binding Kinetics. When studying the interactions

of a soluble protein with surface-immobilized binding receptors, it would be important to consider the heterogeneity of the surface binding sites.31,32 These heterogeneities may stem from random orientations of the immobilized receptor. For example, an antibody immobilized with the antigenic sites facing the substrate will be sterically restricted from binding an antigen.32 We investigated, based on a theoretical analysis, whether the orientations of the immobilized capture antibody on optical fiber surface would affect the antigen/detection antibody binding kinetics in LSPCF-FOB system. One general strategy for solving the different orientations of the antibody is binding antibodies to the Fc receptor protein A (PA) (Figure 3). In this experiment, t-PSA was tested and PA was attached to the modified region of the optical fiber surface (surface I in Figure 3a) before the region was incubated with capture antibodies. This resulted in the oriented immobilization of capture antibodies on the surface I. Meanwhile, the control was arranged at random orientation of capture antibodies in the modified region of the optical fiber surface (surface I in Figure 3b) without PA attached. The strength of the fluorescent signal increased linearly with increasing concentrations of t-PSA in PBS solution over the range from 0.1 pg/mL to 1 ng/mL (Supporting Information, Figures S1a and S1b), and a linearity between [ABeq] and [B0] was shown (Figures 4a and 4b). These results hold true with or without the

presence of PA. In accordance with eq 1 and eq 2, the equilibrium dissociation constant KD, the association rate constant ka, and the dissociation rate constant kd were estimated (Table 1). All affinity constants corresponding to both randomly oriented and specifically oriented, immobilized capture antibodies are very close values, that is, (KD, ka, kd)without PA ≈ (KD, ka, kd)with PA. KD value is in the nanomolar range which is the same order of magnitude as the ones reported in other recent studies.33,34 In this experiment, the LSPCF probe is composed of fluorophore-labeled detection antibodies and Au-PA. PA on the GNP surface not only acts as a linker connecting to the Fc region of the fluorophore-labeled detection antibody but also plays the role of spacer that the quenching of fluorescence is effectively avoided.23 Consequently, the antigen binding domains of the antibody (Fab) orient naturally away from PA of LSPCF probe, making them available for binding to their respective target antigens. Thus, the effect of the random orientation of the antibody in the modified region of the optical fiber surface is significantly reduced. Therefore, the influence of heterogeneities from the surface binding sites in the modified region of the optical fiber surface (surface I in Figure 3) on the binding kinetics between antigen and detection antibody in LSPCF-FOB can be ignored. Mouse IgG/Antimouse IgG Binding Kinetics. To verify this proposed method, the binding kinetics of mouse IgG to an antimouse IgG was conducted. In the experiment, capture IgG antibodies were randomly immobilized on the reactive region of the optical fiber surface where there is no PA immobilized. Then the interactions between the LSPCF probe and mouse IgG at 3294

dx.doi.org/10.1021/ac1028616 |Anal. Chem. 2011, 83, 3290–3296

Analytical Chemistry

ARTICLE

different concentrations in PBS were monitored by detecting separately the temporal fluorescent signal. When the LSPCF probes were injected into the reaction chamber, the detected fluorescent intensity increased rapidly, and then an equilibrium state was reached. During the measurement, the requirement for condition of reaction-limited kinetics was satisfied throughout the concentration range of the target antigen (1 pg/mL to 1 ng/mL). A linear relationship was obtained between the fluorescent signals and mouse IgG concentrations over the range from 1 pg/mL to 1 ng/mL in PBS (Supporting Information, Figures S2). Therefore, KD was calculated as 2.0 nM by using the slope of the linearity between [ABeq] and [B0] (Figure 5), as described in eq 2 given above. Then ka = (2.9 ( 0.6)  105 M1s1 and kd = (6.0 ( 1.2)  104 s1 are determined from the slope of the linearity in eq 1. These results are consistent with the results in recent studies on KD of mouse IgG/antimouse IgG interactions.10,14,15,35,36 Binding Kinetics of t-PSA/Anti-t-PSA Interaction in Human Serum. The extent to which drugs bind serum protein is an essential factor when determines drug pharmacokinetic and activity profiles.37 In this experiment, we measured the binding kinetics of anti-t-PSA interaction with t-PSA in human serum. Similar to the foregoing experiments, a linear relationship between the fluorescence signal and t-PSA was obtained, where t-PSA was added to the 10-fold diluted healthy human serum at different concentrations over the range 0.1 pg/mL to 1 ng/mL (Supporting Information, Figures S3). A linearity between [ABeq] and [B0] is shown (Figure 6). Then, (ka, kd, and KD) are calculated accordingly (Table 1). The affinity constants corresponding to t-PSA/anti-t-PSA interaction in human serum are in good agreement with those in PBS described previously. This result demonstrated that the binding kinetics of protein protein interactions can also be obtained in clinical samples such as human serum been applicable as well. The kinetic rate constants are determined by the previously described two-compartment model, which can be used to isolate the effect of mass transport and extract the kinetic rates.12,27,28 From previous analysis, the transport effect can be ignored if the mass transport flux is much higher than the reaction rate, that is, D/h . ka[B0],2729 indicating that the reaction is dominated by kinetics. D is the diffusion coefficient of the LSPCF probe (Au-PA conjugated fluorophore-labeled detection antibodies) in solution between two compartments in the reaction chamber, and h is the depth of the evanescent wave from the fiber surface (h ≈ 600 nm in the calculation). In this experiment, ka[B0] ≈ 5  107 (m s1) was calculated, while D is inversely proportional to the viscosity η of the solution, according to the StokesEinstein equation:11 D¼

kT 6πηa

where k is the Boltzmann constant, T is the temperature, and a is the radius of LSPCF probe. Thus, D/h ∼ 1.49  105 (m s1) . ka[B0] is satisfied. This assures that the kinetics is dominated by reaction-limited rather than diffusion-limited kinetics in this experiment, whereas the target antigen is at ultralow concentration in the range from ng/mL to pg/mL. As a result, the kinetics of biomolecule interactions following a two-compartment model in a stagnant system via LSPCF-FOB is confirmed. Nevertheless, if the radius of GNP is larger than 1 μm, the condition of D/h . ka[B0] for the criterion on reaction-limited kinetics is no longer satisfied such that the mass transport effect cannot be ignored in this experiment. Generally, the two-compartment model has been developed for

flowing analyte systems to satisfy the requirement of homogeneous distribution in two compartments.13,30 However, the ability to analyze binding kinetics of molecules in a nonflow systems is very important, particularly when the analyte is available only at a low abundance to characterize a biomolecular interaction.30

’ CONCLUSIONS Measurement of the kinetic constants of biomolecule interactions is very important in characterizing the biospecific affinity in solution. In this study, we have extended the proposed LSPCFFOB with high sensitivity and high specificity to measure the binding kinetic constants of antibody interaction with antigen in human serum at low concentration in the ng/mL to pg/mL range. The analytic model is based on a two-compartment model, with proper concentrations of the LSPCF probe and the target antigen in a stagnant system, to satisfy the condition of reactionlimited kinetics. For the first time, we were able to obtain (ka, kd, and KD) simultaneously in such a low concentration range (0.1 pg/mL  1 ng/mL), and in a complex sample such as human serum. We believe that this novel method will pave the way not only to measure binding kinetic constants of biomolecular interactions at ultralow concentration, or with extremely low affinity, from either environmental or clinical samples, but also to study drug pharmacokinetic and activity profiles, considering that drugs can bind serum proteins. ’ ASSOCIATED CONTENT

bS

Supporting Information. Additional material as described in the text. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*Tel: 886-3-2118800 # 3677. Fax: 886-3-2118507. E-mail: [email protected].

’ ACKNOWLEDGMENT L.-C.S. and Y.-F.C. contributed equally to this research. This research was supported partially by the National Science Council of Taiwan through research grants NSC 98-2221-E-182-063-MY3 and NSC 98-2221-E-182-064-MY3. This research was also supported partially by Chang Gung University through research grant UERPD280201. ’ REFERENCES (1) Stoevesandt, O.; Taussig, M. J. Proteomics 2007, 7, 2738–2750. (2) Liu, Y.; Yu, X.; Zhao, R.; Shangguan, D.-H.; Bo, Z.; Liu, G. Biosens. Bioelectron. 2003, 18, 1419–1427. (3) Liu, Y.; Yu, X.; Zhao, R.; Shangguan, D.-H.; Bo, Z.; Liu, G. Biosens. Bioelectron. 2003, 19, 9–19. (4) Taussig, M. J.; Stoevesandt, O.; Borrebaeck, C. A. K.; Bradbury, A. R.; Cahill, D.; Cambillau, C.; de Daruvar, A.; D€ubel, S.; Eichler, J.; Frank, R.; Gibson, T. J.; Gloriam, D.; Gold, L.; Herberg, F. W.; Hermjakob, H.; Hoheisel, J. D.; Joos, T. O.; Kallioniemi, O.; Koegl, M.; Konthur, Z.; Korn, B.; Kremmer, E.; Krobitsch, S.; Landegren, U.; van der Maarel, S.; McCafferty, J.; Muyldermans, S.; Nygren, P.- Å.; Palcy, S.; Pl€uckthun, A.; Polic, B.; Przybylski, M.; Saviranta, P.; Sawyer, A.; Sherman, D. J.; Skerra, A.; Templin, M.; Ueffing, M.; Uhlen, M. Nat. Methods 2007, 4, 13–17. 3295

dx.doi.org/10.1021/ac1028616 |Anal. Chem. 2011, 83, 3290–3296

Analytical Chemistry

ARTICLE

(5) Haab, B. B.; Paulovich, A. G.; Anderson, N. L.; Clark, A. M.; Downing, G. J.; Hermjakob, H.; LaBaer, J.; Uhlen, M. Mol. Cell. Proteomics 2006, 5, 1996–2007. (6) Santamaria, E.; Munoz, J.; Fernandez-Irigoyen, J.; Prieto, J.; Corrales, F. J. Liver Int. 2007, 27, 163–173. (7) Vo-Dinh, T.; Sepaniak, M. J.; Griffin, G. D.; Alarie, J. P. ImmunoMethods 1993, 3, 85–92. (8) Klenin, K. V.; Kusnezow, W.; Langowski, J. J. Chem. Phys. 2005, 122, 214715. (9) Aebersold, R.; Mann, M. Nature 2003, 422, 198–207. (10) Quinn, J. G.; O’Kennedy, R. Anal. Biochem. 2001, 290, 36–46. (11) Yu, F.; Persson, B.; Lofas, S.; Knoll, W. Anal. Chem. 2004, 76, 6765–6770. (12) Shank-Retzlaff, M. L.; Sligar, S. G. Anal. Chem. 2000, 72, 4212–4220. (13) Sapsford, K. E.; Liron, Z.; Shubin, Y. S.; Ligler, F. S. Anal. Chem. 2001, 73, 5518–5524. (14) Chou, C.; Hsu, H.-Y.; Wu, H.-T.; Tseng, K.-Y.; Chiou, A.; Yu, C.-J.; Lee, Z.-Y.; Chan, T.-S. J. Biomed. Opt. 2007, 12, 024025. (15) Lin, C.-H.; Chen, H.-Y.; Yu, C.-J.; Lu, P.-L.; Hsieh, C.-H.; Hsieh, B.-Y.; Chang, Y.-F.; Chou, C. Anal. Biochem. 2009, 385, 224–228. (16) Wang, Y. C.; Han, J. Lab. Chip. 2008, 8, 392–394. (17) Xiang, M.; Xu, X.; Liu, F.; Li, N.; Li, Ka. J. Phys. Chem. B 2009, 113, 2734–2738. (18) Ng, M.-Y.; Liu, W.-C. Opt. Express 2009, 17, 5867–5878. (19) Cui, R. J.; Huang, H., P.; Yin, Z. Z.; Gao, D.; Zhu, J.-J. Biosens. Bioelectron. 2008, 23, 1666–1673. (20) Endo, T.; Yamamura, S.; Nagatani, N.; Morita, Y.; Takamura, Y.; Tamiya, E. Sci. Technol. Adv. Mater. 2005, 6, 491–500. (21) Thaxton, C. S.; Mirkin, C. Nat. Biotechnol. 2005, 23, 681–682. (22) Hsieh, B.-Y.; Chang, Y.-F.; Ng, M.-Y.; Liu, W.-C.; Lin, C.-H.; Wu, H.-T.; Chou, C. Anal. Chem. 2007, 79, 3487–3493. (23) Chang, Y.-F.; Chen, R.-C.; Lee, Y.-J.; Chao, S.-C.; Su, L.-C.; Li, Y.-C.; Chou, C. Biosens. Bioelectron. 2009, 24, 1610–1614. (24) Huang, J. C.; Chang, Y.-F.; Chen, K.-H.; Su, L.-C.; Lee, C.-W.; Chen, C.-C.; Chen, A. Y.-M.; Chou, C. Biosens. Bioelectron. 2009, 25, 320–325. (25) Kingsmore, S. F. Nat. Rev. Drug Discovery 2006, 5, 310–320. (26) Smith, R. H.; Lemon, W. J.; Erb, J. L.; Erb-Downward, J. R.; Downward, J. G.; Ulrich, O. E.; Wittliff, J. L. Clin. Chem. 1999, 45, 1683–1685. (27) Glaser, R. W. Anal. Biochem. 1993, 213, 152–161. (28) Myszka, D. G.; He, X.; Dembo, M.; Morton, T. A.; Goldstein, B. Biophys. J. 1998, 75, 583–594. (29) Karlsson, R.; Roos, H.; Fagerstam, L.; Persson, B. Methods: Companion Methods Enzymol. 1994, 6, 99–110. (30) Vijayendran, R. A.; Ligler, F. S.; Leckband, D. E. Anal. Chem. 1999, 71, 5405–5412. (31) Svitel, J.; Boukari, H.; Ryk, D. V.; Willson, R. C.; Schuck, P. Biophys. J. 2007, 92, 1742–1758. (32) Vijayendran, R. A.; Leckband, D. E. Anal. Chem. 2001, 73, 471–480. (33) Saerens, D.; Kinne, J.; Bosmans, E.; Wernery, U.; Muyldermans, S. J. Biol. Chem. 2004, 279, 51965–51972. (34) Katsamba, P. S.; Navratilova, I.; Calderon-Cacia, M.; Fan, L.; Thornton, K.; Zhu, M.; Bos, T. V.; Forte, C.; Friend, D.; Laird-Offringa, I.; Tavares, G.; Whatley, J.; Shi, E.; Widom, A.; Lindquist, K. C.; Klakamp, S.; Drake, A.; Bohmann, D.; Roell, M.; Rose, L.; Dorocke, J.; Roth, B.; Luginb€uhl, B.; Myszka, D. G. Anal. Biochem. 2006, 352, 208–221. (35) Paetz, A.; Sack, M.; Thepen, T.; Tur, M. K.; Bruell, D.; Finnern, R.; Fischer, R.; Barth, S. Biochem. Biophys. Res. Commun. 2005, 338, 1811–1817. (36) Ivnitski, D.; Wolf, T.; Solomon, B.; Fleminger, G.; Rishpon, J. Bioelectrochem. Bioenerg. 1998, 45, 27–32. (37) Cooper, M. A. Nat. Rev. Drug Discovery 2002, 1, 515–528.

3296

dx.doi.org/10.1021/ac1028616 |Anal. Chem. 2011, 83, 3290–3296