Electrical control over antibody–antigen binding

Available online at www.sciencedirect.com

Sensors and Actuators B 128 (2008) 560–565

Electrical control over antibody–antigen binding Elad Brod ∗ , Shai Nimri, Boaz Turner, Uri Sivan The Russell Berrie Nanotechnology Institute, Department of Physics, Technion-Israel Institute of Technology, Haifa 32000, Israel Received 14 March 2007; received in revised form 4 July 2007; accepted 5 July 2007 Available online 19 July 2007

Abstract We show that the binding of an antibody to its antigen can be controlled electrically in a reversible manner. The antibody–antigen interaction is monitored by an electrochemical surface plasmon resonance (SPR) instrument. The antigen is immobilized on the working electrode while the antibody is injected in solution. After binding, application of a bias more negative than −0.5 V versus Ag/AgCl reference electrode causes rapid detachment of the antibody molecules from the antigens. Removal of the applied voltage restores the antigen ability to bind antibody molecules. The mechanism underlying the reported phenomenon is traced to deprotonation of positively charged amino acids, particularly lysine, by hydroxyl ions generated at the electrode/solution interface. Our finding facilitates exquisite control over one of the main interactions responsible for biomolecular recognition, namely, the attraction between positively and negatively charged residues. Potential applications to diagnostics and sensing are briefly discussed. © 2007 Elsevier B.V. All rights reserved. Keywords: Antibody–antigen interaction; Non-covalent interactions; Surface plasmon resonance; Hydrogen evolution reaction

1. Introduction Gaining an electrical control over the non-covalent interactions responsible for molecular recognition is an important milestone on the way to realize a functional interface between biology and electronics. Such control will facilitate triggering and suppression of biological pathways by electronic signals presented to the system and hence supplementing biology with the remarkable computing power and programmability of electronics. In the following we show that an antibody–antigen interaction, a paradigm for molecular recognition, can effectively be modulated through electrical control over proton concentration in the vicinity of the interacting molecules. Peptide antigens are immobilized on a gold electrode immersed in solution. Antibodies are injected into solution and bind the peptides. Application of a bias more negative than −0.5 V (relative to Ag/AgCl electrode), causes rapid detachment of the antibodies from the peptides.

∗

Corresponding author. E-mail address: [email protected] (E. Brod).

0925-4005/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.snb.2007.07.061

When the bias is removed the antibodies rebind the peptides. Specific binding of an antibody to its antigen depends on several interactions, including the electrostatic attraction between oppositely charged amino acids on the two molecules. In biologically relevant buffers, the short screening length requires an exquisite spatial matching of the charged residues on the two molecules, but when such matching occurs, the resulting attraction free energy is large. The electrostatic attraction between amino acids is sensitive to local proton concentration since the latter determines their side chain protonation state and thus, their net charge. Amino acids with acidic side chains turn negative above pH ≈ 4 and neutralize at pH ≤ 3. Amino acids with basic side chains are positively charged at natural pH and neutralize under alkaline conditions. Two antibody–antigen pairs were examined in this work. Studies of their binding (with no applied bias) show that the pairs dissociate monotonically with pH in the range 7.2–10. Above pH 10 no binding is observed. This range of pH values hints to deprotonation of lysine as the mechanism underlying dissociation. Indeed, both peptides carry this amino acid. Local pH next to a gold cathode can effectively be modulated by the hydrogen evolution reaction (HER) [1]. HER takes place

E. Brod et al. / Sensors and Actuators B 128 (2008) 560–565

on an electrode/solution junction when a sufficiently reductive potential is applied to the electrode [2]. In alkaline solutions, hydrogen evolves from water [3]: 2H2 O + 2e− → H2 + 2OH−

(1)

leaving behind hydroxyl ions that recombine with protons to produce water and thus increase the local pH [4,5]. The distance over which the pH deviates considerably from its native value can be approximated by the recombination length of hydroxyl ions, typically, between a few to hundreds nanometers depending on the pH and buffering capacity of the solution. This out-ofequilibrium effect is thus confined to the vicinity of the cathode. As soon as the bias is removed, the native pH profile is restored by proton and hydroxyl diffusion. Given the observed antibody–antigen dissociation at basic pH values and the coincidence between an electrical triggering of that dissociation and HER, as observed in electrochemical measurements, we attribute the reported phenomenon to electrically induced enhancement of the local pH next to the cathode. The increased pH leads to deprotonation of lysines on the peptide antigens and consequently to a loss of the peptides electrostatic binding to their corresponding antibodies. When the bias is removed the native pH profile is restored, the lysines regain their charge, and the peptides resume their ability to bind their respective antibodies. Our conjecture is supported by corroboration of electrochemical and dissociation data, studies of the effect as a function of pH and bias, and mapping the effect of buffering capacity on the observed phenomenon. Since the effect is confined to the immediate vicinity of the electrode, small electrodes should facilitate spatial control over molecular recognition and protein function, which is yet another new dimension enabled by the reported phenomenon. Several potential applications reflecting these advantages are listed towards the end of the manuscript. Previous reports of electrical effects on biomolecular interactions included modulation of the activity of an enzyme immobilized on a membrane [6], redox reactions [7–9], dissociation of molecules by reducing covalent bonds [10], and

561

electrophoretic and electrostatic control over DNA hybridization [11,12]. None of these reports demonstrated an electrical control over pH for the modulation of molecular recognition next to an electrode. Non-electrical control over antibody–antigen interaction has been exercised through an anion induced conformational change in the antigen peptide [13] and the kinetics of the interaction has been controllably accelerated by application of high hydrostatic pressures [14]. We use a combination of electrochemical setup and an SPR instrument to simultaneously modify the pH in the vicinity of the antibody–antigen pair, and monitor the resulting change in binding. Electrochemical SPR experiments are widely used to investigate ion adsorption [15], Faradaic processes [16], double layer charging [17,18], and redox proteins [19]. Additionally, Heaton et al. [12] investigated effects of voltage on DNA hybridization kinetics and Hodneland and Mrksich [10] designed a self-assembled monolayer that selectively, though irreversibly, releases covalently bound biotin upon application of a reductive potential. In contrast to the latter reference, in our case the antibody is released in response to a change in the protonation states of the antigen rather than breaking of covalent bonds. The effect reported here is, hence, reversible. 2. Experimental The experiment was carried out using a ProteOn XP (lab prototype) SPR instrument developed by Bio-Rad Haifa LTD. The measuring system (Fig. 1) included six 450 ␮m wide, 100 ␮m high channels carrying running solutions in contact with the chip top surface. The chip (Fig. 1b) comprised a glass prism whose top surface was coated with a thin adhesion layer of chromium followed by a 50 nm thick polycrystalline gold layer, patterned to yield working, reference, and counter electrodes. Silver was electroplated on the central part of the reference electrode from 0.5 M AgNO3 solution. Electrochemical chlorination of the silver with 1 M HCl solution produced an Ag/AgCl layer, several microns thick, on top of the electrode. Unless specified other-

Fig. 1. (a) Monitored area of the SPR chip and electrical circuit. The areas in contact with the six fluid channels are marked by dashed lines. Areas of interest (AOI) are denoted by squares and colors correspond to different curves in Fig. 2. (b) An overall view of the patterned chip. The area covered by panel (a) is indicated. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

562

E. Brod et al. / Sensors and Actuators B 128 (2008) 560–565

wise, all potentials mentioned in this work were referenced to this electrode. The SPR signal was monitored at three areas of interest (AOI), 400 ␮m × 400 ␮m each, located on the working electrode along each channel. In each measurement, a full surface plasmon resonance spectrum was taken and the angle of reflection corresponding to the resonance was extracted and displayed in refractive index units (R.U.) as a function of time. One R.U. corresponded to 1 pg adsorbed protein per mm2 [20]. The full surface plasmon spectra indicated an approximately rigid shift of the entire curves upon protein desorption as a result of the applied bias. The curve shift in the reference channel, carrying the same buffer without antibody, was always considerably smaller compared with the antibody-carrying channel. Electrical biasing and measurements were preformed using a 2273 Princeton Applied Research Parastat. The two tested antibody–antigen pairs yielded essentially the same results. The figures below concentrate on one of them. Both antibodies were monoclonal antihistone deacetylase 1 (HDAC1), and monoclonal anti-histone deacetylase 3 (HDAC3). Unless specified otherwise, the antibody concentration in all experiments was 48 ␮g/ml. Their corresponding antigens were peptides with the following sequences, CGGGSKEEKPEAKGVKEEVKLA and CGGGSNEFYDGDHDNDKESDVEI, respectively. The antibodies and their antigens were purchased from Sigma. The linker CGGGS at the N terminus of each peptide served to immobilize it onto the working gold electrode through the cystein’s thiol group. All solutions and buffers were prepared in the lab using doubly distilled water (18 M) and analytic grade chemicals. The peptide antigen molecules were immobilized on the working electrode by incubating the latter overnight in 1.1 ␮g/ml antigen solution in phosphate-buffered saline (PBS), pH 7.2. The resulting SPR signal corresponded to 3.3 × 1013 peptide/cm2 . To confirm the deduced peptide density we have also measured it using the nitrogen line in an X-ray photoelectron spectrometer (XPS, data not shown). The XPS measurements indicated coverage of 4.3 × 1013 peptide/cm2 . After inserting the chip into the SPR instrument, blocking solution containing bovine serum albumin (BSA) was injected to protect the sample against nonspecific antibody binding to the electrode.

Fig. 2. SPR curves demonstrating the effect of voltage pulses on antibody–antigen binding under continuous flow of antibodies. At t1 antibody solution was injected to the binding channel (1) whereas the reference channel (2) was injected with supporting buffer only. Two voltage pulses, 10 s long each, were applied at t2 and t3 , respectively. Three curves are shown for three AOIs in each channel. Curve colors correspond to AOI colors in Fig. 1a. The effect of voltage was typically weaker for the AOI distant form the counter electrode due to ohmic losses along the channel. Flow rate: 25 ␮l/min. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

experiment carried out with “wrong” antibodies (not shown) showed negligible binding. The antibodies density deduced from the SPR signal varied between 1.2 and 1.6 molecules per 100 nm2 (the antibody molecular weight is 150 kDa). Comparing this density with the antigen monolayer density we find that antibodies bind approximately 3% of the bound peptides and cover 30–80% of the surface area. The last figure depends on the assumed orientation of the 8.5 nm × 6 nm × 4 nm antibodies [21]. Potential pulses of −0.9 V applied at t2 and t3 for 10 s led to large signal deflections in the binding channel, indicating an almost complete antibody dissociation from the peptides. The signal recovered gradually when the potential pulse was over, marking reassociation of antibodies with the immobilized antigens. The effect of voltage on the antibody–antigen interaction

3. Results and discussion The main result of our research, namely the effect of voltage on antibody–antigen binding, is demonstrated in Figs. 2 and 3. In the experiment of Fig. 2, binding and reference channels marked (1) and (2), respectively, were monitored simultaneously. Three binding curves are shown for each channel, corresponding to the three AOIs with matched colors in Fig. 1a. PBS pH 7.2 was initially driven through both channels. At t1 , half concentrated PBS pH 8.9 containing HDAC1 antibody was injected to the binding channel. The same supporting buffer without antibody was injected to the reference channel. Following injection, the signal in the binding channel had increased by 3000–4000 R.U., indicating antibody association and then saturation. In the reference channel, change in ion concentration led to a modest fall in signal, followed by stabilization to a slightly lower value. The same

Fig. 3. SPR signal in the binding (1) and reference (2) channels for the case where the buffer carried no antibodies. At t1 antibody-carrying solution was injected to the binding channel. Simultaneously, the same supporting buffer without antibodies was injected to the reference channel. At t2 the antibodycarrying buffer in channel (1) was replaced with the same buffer carrying no antibodies. At t3 the buffer in both channels was replaced with PBS pH 9.0, at t4 a −0.9 V pulse was applied to the two channels for 15 s that ended at t5 . (Inset) Magnified view of SPR signal evolution during the voltage pulse. For better comparison of the voltage effects on the two channels, the two plots in the inset were shifted to coincide at t4 . Flow rate: 25 ␮l/min.

E. Brod et al. / Sensors and Actuators B 128 (2008) 560–565

563

Fig. 4. (a) pH dependence of the antibody–antigen interaction. HDAC1 antibody in PBS at four different pH values was injected at t = 100 to a surface covered with peptide monolayer and BSA. At t = 1040 s the antibody containing solution was replaced with pure buffer. Flow rate: 25 ␮l/min (b) Percentage of detachment vs. applied bias for four different pH values (PBS).

was therefore proved to be reversible, with no apparent damage such as desorption, breaking of covalent bonds, or loss of activity caused to the peptide monolayer by the negative voltage pulse. The much smaller signal drop in the reference channel is attributed to desorption, of specifically adsorbed ions (primarily chlorine) [22–24], charge accumulation at the gold electrode [17,18], and minor desorption of the blocking proteins. By applying longer potential pulses it was found that as long as voltage was applied, even after 40 s, no reassociation occurred. This finding proved that the effect of voltage on the antibody–antigen interaction was persistent rather than transient. Additionally, voltage pulse of opposite polarity did not yield any dissociation. Fig. 3 depicts electrical dissociation of the antibodies from the antigens for the case where the buffer did not carry the former at the time when the pulse was applied. Here too, the binding and reference channels in the figure are marked (1) and (2), respectively. At t1 , PBS pH 7.2 containing HDAC1 antibody was injected to the binding channel. Again, the resulting rise in SPR signal indicated antibody association. The same supporting buffer without antibody was injected to the reference channel. At t2 the buffer in the binding channel was replaced with the same buffer lacking antibodies. At t3 the buffer in both channels was replaced with PBS pH 9.0 and at t4 a −0.9 V pulse was applied for 15 s. The pulse ended at t5 . As expected, unlike the case of Fig. 2 where antibodies continued to flow throughout the whole experiment, here the SPR signal did not recover to the original level. The magnified scale shows the signal during voltage pulse. The two plots in the inset were shifted to coincide at t4 to stress the different effect of voltage pulse on the two channels. The signal in the antibody channel combined two effects, the fast, fully reversible reaction of the buffer and gold electrode to the pulse [17,18,22–24] and the slower antibody dissociation process. Best exponential fits to the segments t2 → t3 , t3 → t4 , t4 → t5 , respectively, yielded 250 ± 70, 150 ± 20, and 4.5 ± 1.0 s dissociation times.1 Application of bias thus enhanced the dissociation rate more than 30-fold, and was at least qualitatively equivalent to a significant increase in pH.

1 The fast decrease in the first second due to ion and electrode effects needs to be subtracted from the signal before fitting.

The effect of native pH on the antibody–antigen interaction with no applied bias is plotted in Fig. 4a. Antibodies at various pH conditions were injected at t = 100 s. The resulting rise in SPR signal clearly indicates antibody association. After additional 940 s the antibody-carrying solution was substituted with pure buffer of the same pH. As a result, antibodies started to dissociate from the peptide antigens. The dissociation rate grew dramatically as the pH approached 10. No binding was observable at pH ≥ 10. Several studies [25,26] have related the pH dependence of an antibody–antigen interaction to titration curves of the amino acid residues participating in the interaction. In the present case the pH range over which the interaction diminished points to lysine residues whose bulk pKa is about 10 [27], depending on their immediate neighborhood. Indeed, both peptides carried this residue. The effect of bias on the antibody–antigen interaction at different pH values is plotted in Fig. 4b. Antibody dissociation grew monotonically with both applied bias and buffer pH. The enhanced dissociation at elevated pH is attributed more to the limited buffering capacity of PBS at these pH values than to the actual pH (see below). The threshold bias for antibody–antigen dissociation, ≈−0.5 to −0.6 V, provided a hint to the central importance of HER to the observed phenomenon. The association of the reported phenomenon with HER was further substantiated by electrochemical measurements. Fig. 5 depicts cyclic voltammogram (−0.6 V → 0.5 V → −0.6 V) measured in PBS pH 7.2 with an unpatterned SPR chip. Voltage was measured relative to a saturated calomel electrode, thus for direct comparison with Fig. 4b the bias in Fig. 5 has to be offset by the calomel potential relative to AgCl/Ag electrode, namely, +45 mV. The shoulder at −0.4 V in the negative going voltage scan is attributed to reduction of hydrated AuOH to gold and water [17]. The sharp increase in current below −0.5 V reflects the turn-on of HER. Note this turn-on coincides with the threshold for antibody–antigen dissociation found in Fig. 4b. Further evidence for the association of the modulation of antibody–antigen binding with electrical control over proton concentration is provided by comparing experiments carried out at the same pH but different buffering capacities. According to previous studies, the elimination of protons through HER in unbuffered solutions causes an increase in pH near the electrode

564

E. Brod et al. / Sensors and Actuators B 128 (2008) 560–565

Fig. 5. Cyclic voltammogram measured with unpatterened SPR chip (−0.6 V → 0.5 V → −0.6 V) in PBS solution, pH 7.2. Voltages were measured relative to a KCl saturated calomel reference electrode. Area of electrode: 0.69 cm2 , scan rate: 50 mV/s.

while in buffered solutions this effect is small [1]. Such comparison between PBS and pyrophosphate buffer, both at pH 9, is depicted in Fig. 6. PBS with its three pKa values: 2.15, 7.2, and 12.33 has weak buffering capacity at pH 9. Pyrophosphate, on the other hand, having four pKa values: 0.91, 2.10, 6.70, and 9.32, buffers well at that pH. The two binding channels were injected at t1 with antibody solution in PBS pH 7.2. At t2 antibody injection was terminated and a slow decline had set in when pure PBS buffer of the same pH continued to flow. At t3 PBS and pyrophosphate buffers, both at pH 9, were injected to their respective channels accelerating antibody dissociation as expected (t3 in Fig. 3). At t4 a potential pulse of −0.9 V was applied for 15 s causing antibody dissociation. While the antibody dissociation rates for t3 < t ≤ t4 were similar for the two buffers, the voltage pulse at t4 released about three times more antibodies in PBS compared with the pyrophosphate channel. In fact, antibody–antigen dissociation in the first case was almost complete. Given the similar ionic

Fig. 6. Antibody dissociation in PBS and pyrophosphate buffers, both at pH 9. Antibody in PBS pH 7.2 supporting buffer was injected to two binding channels from t1 to t2 . At t2 the buffer was replaced with pure PBS buffer of the same pH. At t3 one channel was injected with PBS buffer pH 9 (pink plot) and the other with pyrophosphate buffer of the same pH (blue plot). At t4 voltage pulse of 15 s was applied to both channels. Note the large dissociation jump at t4 in PBS compared with the case of pyrophosphate buffer. Flow rate: 25 ␮l/min. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

strength of the two buffers we can safely attribute the difference to buffering capacity. Similar experiments at pH 9 with solutions having high buffering capacity (borate and Tris) as well as low buffering capacity (acetic acid) supported the same conclusion. The same conclusion also emerged from a series of experiments carried out at different buffer concentrations and the same ionic strength provided by sodium chloride. As predicted, antibody dissociation was maximal for pure NaCl and diminished monotonically with buffer concentration. The effect in the former case was similar in magnitude to that in PBS at pH 9.5. The compilation of all data univocally points to hydrogen evolution and the resulting increase in local pH as the mechanism controlling antibody–antigen desorption in our experiments. The pH profile is richer than expected since the same negative bias applied to the gold in order to generate current, also generates an electrostatic field that attracts protons from the solution, and reduces the pH in the immediate vicinity of the cathode. At low bias, the latter effect is amplified by chlorine ion chemisorption to the gold. Fortunately, electrostatics and current manifest themselves on two different scales. While the former decays on scales comparable to the screening length (λ ≈ 1 nm for 0.1 M NaCl), the effect of current is long range. Full calculation should take into account chlorine desorption from the cathode by the applied bias, HER, and ion recombination. We conclude by listing a few potential applications. The simplest would be reversible and tunable attachment and detachment of biomolecules to specific sites in protein arrays. In that scheme the antibody is fused to a molecule of interest and the couple is localized or released from a given electrode at a flip of a switch, freeing that electrode for attachment of other moieties. Another straightforward application would be mapping of the full pH dependence of an antibody–antigen interaction either by scanning the bias applied to one electrode or by application of different bias to individual electrodes interacting with the same sample. Applications to biological systems are also possible despite the strong buffering capacity characterizing them. One solution would be 100-nm-scale compartments that limit the access of further buffer molecules. A particularly interesting possibility would be application of bias across compartments or polymers that pass protons but block hydroxyl ions. The modulation of pH in that case should be very efficient due to the suppressed recombination. The phenomenon reported here extends beyond antibody–antigen interaction. The electrostatic interaction between positively and negatively charged amino acids is central to protein folding and their function. The applications of programmable local control over protein function with nanometer scale resolution are obviously very broad. Acknowledgements We are in debt to Prof. Israel Rubinstein for indicating to us the importance of HER next to gold electrodes, Tal Rosenzweig and Ariel G. Notcovich for helping us with the setup and helpful discussions, Prof. Yoram Reiter and Arbel Artzy for numerous discussions, and Prof. Steve Lipson for critical

E. Brod et al. / Sensors and Actuators B 128 (2008) 560–565

reading of the manuscript. The research was supported by the Israeli Science Foundation (ISF) and the Deutsch-Israelische Projektkooperation (DIP). References [1] T.N. Andersen, B.S. Dandapani, J.M. Berry, Hydrogen evolution studies in neutral media, J. Electroanal. Chem. 357 (1–2) (1993) 77–89. [2] G.J. Burg, M. Sluyters-Rahbach, J.H. Sluyters, The kinetics of the reduction of protons at polycrystalline and monocrystalline gold electrodes, J. Electroanal. Chem. 181 (1984) 245–266. [3] K.J. Vetter, Electrochemical Kinetics, Academic Press, New York, 1967, pp. 516–614. [4] C.R. Catherine, B. Finlayson, Y. Paul, Formation of natural pH gradients in a microfluidic device under flow conditions: model and experimental validation, Anal. Chem. 73 (3) (2001) 658–666. [5] M. Bier, O.A. Palusinski, R.A. Mosher, D.A. Saville, Electrophoresis: mathematical modeling and computer simulation, Science 219 (4590) (1983) 1281–1287. [6] D. Vallin, C. Tran Minh, Electric field effects on immobilized urease activity, Biochim. Biophys. Acta Enzymol. 571 (2) (1979) 321–332. [7] N.J. Tao, Probing potential-tuned resonant tunneling through redox molecules with scanning tunneling microscopy, Phys. Rev. Lett. 76 (21) (1996) 4066–4069. [8] L. Andolfi, D. Bruce, S. Cannistraro, G.W. Canters, J.J. Davis, H.A.O. Hill, J. Crozier, M.Ph. Verbeet, C.L. Wrathmell, Y. Astier, The electrochemical characteristics of blue copper protein monolayers on gold, J. Electroanal. Chem. 565 (1) (2004) 21–28. [9] J.J. Davis, D. Bruce, G.W. Canters, J. Crozier, H. Hill, O. Allen, Genetic modulation of metalloprotein electron transfer at bare gold, Chem. Commun. (Camb.) 5 (2003) 576–577. [10] C.D. Hodneland, M. Mrksich, Biomolecular surfaces that release ligands under electrochemical control, J. Am. Chem. Soc. 122 (2000) 4235–4236. [11] C.F. Edman, D.E. Raymond, D.J. Wu, E. Tu, R.G. Sosnowski, W.F. Butler, M. Nerenberg, M.J. Heller, Electric field directed nucleic acid hybridization on microchips, Nucleic Acids Res. 25 (24) (1997) 4907–4914. [12] R.J. Heaton, A.W. Peterson, R.M. Georgiadis, Electrostatic surface plasmon resonance: direct electric field-induced hybridization and denaturation in monolayer nucleic acid films and label free discrimination of base mismatches, PNAS 98 (7) (2001) 3701–3704. [13] Y. Katakura, T. Miyazaki, H. Wada, T. Omasa, M. Kishimoto, Y. Goto, K. Suga, Control of antibody–antigen interaction using anion-induced conformational change in antigen peptide, Protein Eng. 13 (2000) 719–724. [14] D.J. Green, G.J. Litt, J.A. Laugharn, Use of high pressure to accelerate antibody:antigen binding kinetics demonstrated in an HIV-1 p24:Anti-HIV1 p24 assay, Clin. Chem. 44 (1998) 341–342. [15] Y. Iwasaki, T. Horiuchi, M. Morita, N. Osamu, Time differential surface plasmon applied for electrochemical analysis, Electroanalysis 9 (16) (1997) 1239–1241. [16] Y. Iwasaki, T. Horiuchi, M. Morita, N. Osamu, Electrochemical reaction of Fe(CN)6 3−/4− on gold electrodes analyzed by surface plasmon resonance, Surf. Sci. 427–428 (1999) 195–198. [17] J.E. Garland, K.A. Assiongbon, C.M. Pettit, D. Roy, Surface plasmon resonance transients at an electrochemical interface: time resolved measurements using a bicell photodiode, Anal. Chim. Acta 475 (2003) 47– 58.

565

[18] C.M. Pettit, K.A. Assiongbon, J.E. Garland, D. Roy, Time resolved detection of electrochemical effects by surface plasmon resonance measurements: a simple technique using a large area single cell photodiode, Sens. Actuators B 96 (2003) 105–113. [19] D.D. Schlereth, Characterization of protein monolayers by surface plasmon resonance combined with cyclic voltammetry ‘in situ’, J. Electroanal. Chem. 464 (1999) 198–207. [20] E. Stenberg, Quantitative determination of surface concentration of protein with surface plasmon resonance using radiolabeled proteins, J. Colloid Interface Sci. 143 (2) (1991) 513–526. [21] M. Marquart, J. Deisenhofer, R. Huber, W. Palm, Crystallographic refinement and atomic models of the intact immunoglobulin molecule Kol and ˚ and 1.9 A ˚ resolution, J. Mol. Biol. its antigen-binding fragment at 3.0 A 141 (4) (1980) 369–391. [22] F. Silva, A. Martins, Surface structural effects on specific adsorption of oxoanions on gold single crystal electrodes, J. Electroanal. Chem. 467 (1999) 335–341. [23] J.P. Bellier, Structural effect on adsorption at solid electrodes halide adsorption on gold surfaces with “rails”, J. Electroanal. Chem. Interfacial Electrochem. 140 (2) (1982) 391–401. [24] J.O. Bockris, W.K. Paik, M.A. Genshaw, Adsorption of anions at the solid-solution interface ellipsometric study, J. Phys. Chem. 74 (24) (1970) 4266–4275. [25] C.J. Gibas, S. Subramaniam, A.J. McCammon, B.C. Braden, R.J. Poljak, pH dependence of antibody/lysozyme complexation, Biochemistry 36 (1997) 15599–15614. [26] A. Dejaegere, L. Choulier, V. Lafont, E. De Genst, D. Altschuh, Variations in antigen–antibody association kinetics as a function of ph and salt concentration: a QSAR and molecular modeling study, Biochemistry 44 (2005) 14409–14418. [27] L. Strayer, Biochemistry, 3rd ed., W.H. Freeman and Company, New York, 1988, p. 21.

Biographies Elad Brod, MSc student in the department of physics at the Technion-Israel Institute of Technology. He obtained his first degree in biochemistry from this institute in 2003. His current fields of interest are DNA – polycation interactions and computer simulations of electrically induced pH gradients near a metal/solution interface. Shai Nimri, PhD in chemistry (2000) from the Technion (Haifa, Israel). Presently he is head of the chemistry R&D team in Bio-Rad Haifa, in charge of the development of SPR sensor-chips of the ProteOn XPR36 system. His main fields of interest are biosensors and bioassays, applied surface chemistry. Boaz Turner, PhD in chemistry (2000) from the Technion (Haifa, Israel). Presently he is senior chemist in the chemistry R&D team in Bio-Rad Haifa, leading the development of a high performance new SPR sensor-chip for the ProteOn XPR36 system. His main fields of interest are metal evaporation processes and applied surface chemistry. Prof. Uri Sivan holds the Bertoldo Badler Chair in Physics and heads the Russell Berrie Nanotechnology Institute at Technion – Israel Institute of Technology. His current scientific interests focus on the interface between molecular biology and nanoelectronics as well as the interaction between biomolecules and electronically relevant surfaces.