pubs.acs.org/Langmuir © 2009 American Chemical Society
Electrochemical Desorption of Fibrinogen from Gold Colm T. Mallon,† Ciaran De Chaumont,‡ Niamh Moran,‡ Tia E. Keyes,*,† and Robert J. Forster*,† †
National Center for Sensor Research, School of Chemical Sciences, Dublin City University, Dublin 9, Ireland, and ‡Molecular and Cellular Therapeutics, Royal College of Surgeons in Ireland, Dublin 2, Ireland Received June 12, 2009. Revised Manuscript Received August 14, 2009
The electrochemically induced desorption of Oregon green labeled fibrinogen layers from clean gold surfaces at negative potentials has been probed using capacitance, fluorescence microscopy, and atomic force microscopy. Capacitance measurements on fibrinogen layers indicate that desorption occurs at potentials more negative than -0.8 V and that complete desorption occurs when the electrode is biased at -1.2 V. Significantly, the fluorescence intensity initially increases as the dye labeled protein is electrochemically desorbed due to a decrease in quenching by the gold surface. Following this initial increase, the protein diffuses into solution and the fluorescence intensity decreases over time. More than 90% of the dye labeled fibrinogen is desorbed and diffuses out of the confocal volume in less than 2000 s when the potential is stepped to -1.2 V. AFM before and after application of the desorbing potential confirms removal of the protein. Collection of the desorbed protein in solution reveals a surface coverage of (4.0 ( 2.3) 10-13 mol cm-2 or an area of occupation of 400 ( 140 nm2 per molecule, which indicates that the protein is not extensively spread on the bare gold surface. Significantly, SDS-PAGE analysis indicates that the adsorption-desorption cycle dramatically effects the protein structure, with the electrochemically desorbed fibrinogen showing extensive fragmentation compared to native protein.
Introduction 1,2
The factors that influence adsorption of proteins, such as the surface hydrophobicity3 or capacity for electrostatic interaction,4 have been widely documented due to the importance of protein adsorption in areas such as biofouling or biocompatibility of implants.5 However, similar factors related to desorption, which is almost always an activated process, are significantly under explored. This situation is surprising given the importance of controlling the surface coverage of proteins in applications ranging from biosensing, triggered delivery of a protein payload, to implants and ensuring the sterility of surgical instruments. Methods for protein desorption from surfaces that have been previously explored6 typically involve changing the ionic strength,4a the pH,6b or the solvent polarity of the contacting solution so as to disrupt the surface-protein electrostatic interactions. Alternatively, *To whom correspondence should be addressed. (1) (a) Brash, J.; Horbett, T. Proteins at Interfaces, Physicochemical and Biochemical Studies; American Chemical Society: Washington, DC, 1987. (b) Horbett, T.; Brash, J. Proteins at Interfaces II, Fundamentals and Applications; American Chemical Society: Washington, DC, 1995. (2) (a) Castner, D. G.; Ratner, B. D. Surf. Sci. 2002, 500, 28. (b) Nakanishi, K.; Sakiyama, T.; Imamura, K. J. Biosci. Bioeng. 2001, 91, 233. Rutala, W. A.; Weber, D. J. Clin. Infect. Dis. 2001, 32, 1348. (3) (a) Agnihotri, A.; Siedlecki, C. A. Langmuir 2004, 20, 8846. (b) Marchin, K. L.; Berrie, C. L. Langmuir 2003, 19, 9883. (4) (a) Jackson, D. R.; Omanovic, S.; Roscoe, S. G. Langmuir 2000, 16, 5449. (b) Jordan, C. E.; Corn, R. M. Anal. Chem. 1997, 69, 1449. (c) Chan, B. M. C.; Brash, J. L. J. Colloid Interface Sci. 1981, 82, 217. (d) Koutsoukos, P. G.; Norde, W.; Lyklema, J. J. Colloid Interface Sci. 1983, 95, 385. (e) Van Dulm, P.; Norde, W. J. Colloid Interface Sci. 1983, 91, 248. (5) (a) Ratner, B. D.; Bryant, S. J. Annu. Rev. Biomed. Eng. 2004, 6, 41. (b) Thevenot, P.; Hu, W.; Tang, L. Curr. Top. Med. Chem. 2008, 8, 270. (c) Rutala, W. A.; Weber, D. J. Clin. Infect. Dis. 2001, 32, 1348. (6) (a) Norde, W.; MacRitchie, F.; Nowicka, G.; Lyklema, J. J. Colloid Interface Sci. 1986, 112, 447. (b) Urano, H.; Fukuzaki, S. J. Colloid Interface Sci. 2002, 252, 284. (c) Giacomelli, C. E.; Norde, W. J. Colloid Interface Sci. 2001, 233, 234. (7) (a) Sarkar, D.; Chattoraj, D. K. J. Colloid Interface Sci. 1996, 178, 606. (b) Sheller, N. B.; Petrash, S.; Foster, M. D.; Tsukruk, V. V. Langmuir 1998, 14, 4535. (c) Rapoza, R. J.; Horbett, T. A. J. Colloid Interface Sci. 1990, 136, 480. (8) Cheng, Y. L.; Darst, S. A.; Robertson, C. R. J. Colloid Interface Sci. 1987, 118, 212. (9) (a) van der Veen, M.; Cohen Stuart, M.; Norde, W. Colloids Surf., B 2007, 54, 136. (b) Vasina, E. N.; Dejardin, P. Biomacromolecules 2003, 4, 304. (c) Ball, V.; Bentaleb, A.; Hemmerle, J.; Voegel, J. C.; Schaaf, P. Langmuir 1996, 12, 1614.
Langmuir 2010, 26(1), 293–298
competitive adsorption of surfactants,7 other proteins,8 or exchange into blank buffer9 are common methods to typically achieve partial desorption. These desorption studies have focused primarily on the kinetics of desorption6b,7a,9 and the effect of the adsorption/desorption cycle on the protein structure.6c Significantly, complete desorption of protein layers is difficult to achieve,7a,c,9a,c,10 which may be a consequence of a tightly bound inner layer and a more loosely bound overlayer. Electrochemical approaches are particularly attractive for achieving protein desorption since the driving force for desorption can be at least partially controlled through the applied potential. Moreover, depending on the protein structure, it may be possible to control the desorption mechanism, e.g., negatively charged, physisorbed proteins ought to be electrostatically desorbed at potentials negative of the potential of zero charge, whereas cysteine-rich proteins may chemisorb through goldthiol bond formation which will require potentials negative of ∼-1.0 V to be reductively desorbed. A second important issue is the impact of the adsorption-desorption cycle on the protein structure; e.g., does it induce conformational changes, denaturation, or protein scission? This issue is important since for some applications, e.g., the delivery of a protein at a defined time and location, the protein must retain its native structure while for other applications, e.g., the decontamination of surgical instruments, protein chain scission would be desirable.5c In this contribution, we report on the influence of electrode potential on an adsorbed fibrinogen layer at a gold surface. Fibrinogen is an important protein in blood11 and plays a central role in thrombosis by aggregating activated platelets. It also undergoes nonspecific adsorption at several surfaces.12,13 Here, (10) Bohnert, J. L.; Horbett, T. A. J. Colloid Interface Sci. 1986, 111, 363. (11) Weisel, J. W. Adv. Protein Chem. 2005, 70, 247. (12) (a) Desroches, M. J.; Omanovic, S. Phys. Chem. Chem. Phys. 2008, 10, 2502. (b) Sit, P. S.; Marchant, R. E. Surf. Sci. 2001, 491, 421. (c) Toscano, A.; Santore, M. M. Langmuir 2006, 22, 2588. (d) Clarke, M. L.; Wang, J.; Chen, Z. J. Phys. Chem. B 2005, 109, 22027. (e) Cosman, N. P.; Roscoe, S. G. Langmuir 2004, 20, 1711. (13) (a) Lin, Y.; Wan, L. T.; Fang, X. H. Ultramicroscopy 2005, 105, 129. (b) Yu, Y.; Jin, G. J. Colloid Interface Sci. 2005, 283, 477. (c) Hemmersam, A. G.; Foss, M.; Chevallier, J.; Besenbacher, F. Colloids Surf., B 2005, 43, 208.
Published on Web 09/11/2009
DOI: 10.1021/la902115e
293
Article
Mallon et al.
we report on the electrostatic repulsions between the negatively charged fibrinogen and a negatively charged gold electrode. In this way, new insights into the binding mechanism are obtained. The adsorption of fibrinogen on gold is characterized by AFM and capacitance measurements, and the effect of the electrode potential on interfacial protein desorption is probed using capacitance, fluorescence spectroscopy, and AFM. These investigations show that the protein desorbs as the potential is moved in a negative direction. Significantly, real time fluorescence and AFM measurements indicate that essentially all of the protein is desorbed at -1.2 V. In addition, SDS-PAGE analysis of the electrochemically released fibrinogen indicates that the protein is considerably fragmented by the adsorption-desorption cycle. This desorption procedure allows practically complete removal and significant destruction of adsorbed fibrinogen from gold.
Materials and Methods Electrochemistry was performed in a standard three-electrode cell with an Ag/AgCl reference electrode saturated with KCl and a platinum mesh as a counter electrode at 23 ( 2 °C. Gold-coated (400 nm) silicon wafers were used as working electrodes. All solutions were deoxygenated with argon, and a blanket of argon was maintained over the solutions during the experiments. Cyclic voltammetry, open-circuit potential (OCP), and impedance measurements were performed on a CH Instruments Model 600 electrochemical workstation. Impedance measurements were carried out using ac voltammetry at a frequency of 512 Hz and using an excitation signal of 5 mV. Electronic absorbance spectra were measured on a Shimadzu 3500 UV/vis/NIR spectrophotometer. Emission spectra were recorded on a Cary Eclipse fluorescence spectrophotometer. The fluorescence of the adsorbed Oregon green fibrinogen layer at the gold surface was monitored with confocal microscopy using a Horiba Jobin Yvon HR800UV spectrometer and a 10 objective lens. An argon ion laser provided an excitation source of 488 nm, which was focused onto the electrode surface. Acquisitions were 1 s in length and performed once to acquire a spectrum. A custom-built cell allowed potential control over the fibrinogencoated electrode, which was situated at the bottom of a well of electrolyte. Atomic force microscopy was carried out using a Nanoscope III (Digital Instruments) and silicon tips (Veeco Probes OTESPA for contact mode and Dataprobe DP14/GP/Ti-AuBS for tapping mode). At least five images were collected from different areas of the substrates to confirm reproducibility. Images were processed and analyzed using WSxM software.14 The integrity of the protein was assessed by SDS-PAGE gel electrophoresis as previously described.15 Briefly, SDS sample buffer (10% glycerol, 62.5 mM TrisHCl, pH 6.8, 2% SDS, 5% β-mercaptoethanol, and 0.01 mg/mL bromophenol blue) was added to native or desorbed proteins. These samples were further denatured by heating to 100 °C for 3 min before being loaded onto 10% SDS polyacrylamide gels. Samples were separated at 250 V for ∼30 min or until the dye front reached the end of the gel. Gels were fixed and then stained with 0.1% Coomassie brilliant blue R-250 in 5% glacial acetic acid/50% methanol. Fibrinogen (65%) and NaHCO3 (99.7%) were obtained from Sigma-Aldrich Chemical Co. Oregon green labeled fibrinogen (90%) was obtained from Invitrogen. Fibrinogen was prepared in PBS (5 mL) (phosphate buffered saline) solution and 0.1 M NaHCO3. PBS was composed of 0.01 M phosphate buffer, 0.0027 M KCl, and 0.137 M NaCl (pH 7.4 at 25 °C). These salts also acted as the supporting electrolyte for electrochemical measurements.
The gold substrates were electrochemically cleaned by scanning the potential in 0.5 M H2SO4 between -0.25 and 1.45 V until stable voltammograms were obtained. Fibrinogen layers were prepared on the cleaned electrodes by immersing the electrode in fibrinogen solutions (g60 μM, 5 mL) for at least 16 h. The fibrinogen-coated electrodes were rinsed copiously with Milli-Q water before use. Bulk adsorption-desorption was performed to obtain sufficient desorbed protein for SDS PAGE analysis. A 3 cm2 gold electrode in a solution of 50 μM fibrinogen was subjected to a series of potential steps from OCP to -1.2 V. The time spent at OCP was 12 min, and kinetic studies suggested that this time was sufficient to form at least 95% coverage at this fibrinogen concentration. The potential was then stepped to -1.2 V for 3 min to desorb the protein. This process was repeated, under gentle solution agitation, for 90 h to ensure all of the protein went through the adsorption-desorption process at least once.
(14) Horcas, I.; Fernandez, R.; Gomez-Rodrı´ guez, J. M.; Colchero, J. Rev. Sci. Instrum. 2007, 78, 013705. (15) Larkin, D.; Murphy, D.; Reilly, D. F.; Cahill, M.; Sattler, E.; Harriott, P.; Cahill, D. J.; Moran, N. J. Biol. Chem. 2004, 279, 27286.
Fibrinogen Layer Properties. Figure 1A shows the doublelayer capacitance, Cdl, of bare gold electrodes (gray, upper curve)
294 DOI: 10.1021/la902115e
Figure 1. (A) Capacitance-potential profiles for bare (gray, upper curve) and fibrinogen-coated (black, lower curve) gold electrodes (0.5 cm2 area) in PBS (0.01 M phosphate buffer, 0.0027 M potassium chloride and 0.1 M 0.137 M sodium chloride) obtained using ac voltammetry at a frequency of 512 Hz and using an excitation signal of 5 mV. The error bars represent the standard deviations of three independent measurements. (B) Contact mode AFM image of fibrinogen-coated gold electrode.
Results and Discussion
Langmuir 2010, 26(1), 293–298
Mallon et al.
Figure 2. CVs of fibrinogen-coated gold electrode (0.5 cm2 area) in PBS (0.01 M phosphate buffer, 0.0027 M potassium chloride and 0.1 M 0.137 M sodium chloride) scanned to increasingly negative potentials. The scan rate is 0.1 V s-1.
and gold electrodes that were immersed overnight in an aqueous buffered solution containing 100 μM fibrinogen (black, lower curve). The capacitance values for the bare electrodes are typical of those observed for a clean gold electrode16 in contact with an aqueous solution, and as expected, there is a marked potential dependence. In contrast, electrodes exposed to a 100 μM solution of fibrinogen for 16 h show significantly lower capacitance values for all potentials investigated. Moreover, the overall sensitivity to the applied potential is significantly less than that found for the bare electrode. The lower capacitance and reduced sensitivity to potential are both consistent with the formation of an organic layer on the electrode surface. Figure 1B shows an AFM image of the fibrinogen-modified electrode. This image shows a mesh like network with a root-mean-square (rms) roughness of 9.0 nm, which is in stark contrast with the AFM images obtained at bare gold electrodes (no significant features and rms roughness of 2.0 nm). Therefore, exposure of the gold electrode to the fibrinogen solution results in the spontaneous formation of a densely networked fibrinogen layer.12b,13a,17 There are two dominant electrochemical processes that are likely to be involved in protein binding to gold. First, electrostatic interactions between the negatively charged protein and electrode when Eapp is positive of the PZC. Second, binding of thiols from cysteines within the protein to the gold. Figure 1A shows that the capacitance of the fibrinogen-coated electrode is independent of the applied potential, Eapp, for values between approximately þ0.1 and -0.4 V before increasing significantly for potentials more negative than about -0.6 V. Overall, Cdl, increases by a factor of ∼3 on going from -0.4 to -1.0 V, suggesting that the surface coverage of fibrinogen may be lower at negative potentials. Figure 1A shows that for the unmodified gold electrode a minimum in Cdl occurs at ∼-0.6 V, which is consistent with the PZC observed by others in PBS buffer.18 Therefore, taking into account the shifts in PZC expected upon modification with fibrinogen, it appears that for potentials negative of ∼-0.4 V there is sufficient electrostatic (16) McNally, A.; Forster, R. J.; Keyes, T. E. Phys. Chem. Chem. Phys. 2009, 11, 848. (17) (a) Ohta, R.; Saito, N.; Ishizaki, T.; Takai, O. Surf. Sci. 2006, 600, 1674. (b) Agnihotri, A.; Siedlecki, C. A. Langmuir 2004, 20, 8846. (18) Yu, Y.; Jin, G. J. Colloid Interface Sci. 2003, 268, 288. (19) Liu, M.; Zhang, Y.; Wang, M.; Deng, C.; Xie, Q.; Yao, S. Polymer 2006, 47, 3372.
Langmuir 2010, 26(1), 293–298
Article
Figure 3. Capacitance of a fibrinogen layer and fluorescence intensity of an Oregon green fibrinogen layer (obtained from Figure 4) as a function of potential. The capacitance data were obtained using ac voltammetry at a frequency of 512 Hz and using an excitation signal of 5 mV in PBS (0.01 M phosphate buffer, 0.0027 M potassium chloride and 0.1 M 0.137 M sodium chloride). The arrow indicates the potential at which dΓ/dV increases abruptly.
repulsion between the negatively charged protein19 and the electrode so as to drive desorption, causing an increased Cdl to be observed. Figure 2 illustrates the effect of scanning the potential of a fibrinogen-coated electrode from the open-circuit potential, OCP (0.2 V), to increasingly negative potentials. Consistent with the potential-dependent changes in Cdl presented in Figure 1A, the background current at þ0.2 V increases as the potential is scanned to progressively more negative potentials. Figure 3 shows that scanning the potential more negative than -0.9 V triggers a large increase in Cdl, reaching 27 μF cm-2 at -1.2 V. This figure suggests that the stability of the fibrinogen layer with respect to the applied potential changes significantly at ∼-0.9 V; i.e., dΓ/dV increases abruptly. Significantly, Figure 2 reveals an irreversible redox process with a peak potential of ∼-1.0 V. This response is consistent with reductive desorption of thiols bound to the gold surface,20 and fibrinogen, which contains 29 disulfide bonds, is relatively cysteine rich. Therefore, it appears that electrostatic desorption is the dominant mechanism for potentials between the PZC and ∼-0.9 V, while reductive thiol desorption dominates for more negative values. While our current data do not allow us to draw definitive conclusions, taking the capacitance and AFM data together, it seems likely that the protein film consists of loosely bound multilayers that desorb electrostatically and a tightly bound inner layer that desorbs reductively. Desorption Kinetics. The desorption dynamics were investigated using dye labeled protein and confocal fluorescence microscopy. Oregon green labeled fibrinogen was adsorbed onto the gold substrates, and the fluorescence signal from the modified surface was monitored as a function of the electrode potential by focusing on the electrode surface. The wavelength of excitation was 488 nm (λmax of Oregon green is 496 nm), giving emission at 526 nm. Figure 4 shows that the fluorescence intensity of the layer is approximately independent of the applied potential until -0.6 V but that the fluorescence intensity increases ∼5-fold at -1.2 V compared to the OCP. In a similar way to that discussed (20) (a) Walczak, M. M.; Popenoe, D. D.; Deinhammer, R. S.; Lamp, B. D.; Chung, C.; Porter, M. D. Langmuir 1991, 7, 2687. (b) Yang, D. F.; Wilde, C. P.; Morin, M. Langmuir 1996, 12, 6570.
DOI: 10.1021/la902115e
295
Article
Figure 4. Fluorescence spectra of an Oregon green fibrinogen layer on a gold substrate held at progressively negative potentials in aqueous solution with 0.1 M NaHCO3 as the supporting electrolyte using a 488 nm laser as the excitation source. The blank curve represents the signal in the absence of the Oregon green fibrinogen layer.
for the capacitance data, Figure 3 shows that the fluorescence intensity progressively increases as the potential is adjusted more negative than ∼-0.9 V. Both observations are consistent with fibrinogen desorption. The fluorescence intensity increases upon desorption because emission from the surface bound protein is partially quenched by the gold surface.21 A similar effect has been reported for dye labeled DNA bound to a gold surface.21a Significantly, all potential-induced changes are irreversible, which is consistent with protein desorption followed by diffusion into the bulk solution. To probe the desorption dynamics and subsequent diffusion of the protein in solution, the intensity of the fluorescence signal was monitored as a function of time after stepping the potential of the electrode to -1.2 V. Figure 5 shows the evolution of the fluorescence spectra with time, while inset A shows the intensity at λmax as a function of time. Two distinct regions can be observed in these plots. Over the first 200 s the intensity increases as the protein is desorbed and produces a high concentration in solution within the confocal volume of the microscope. However, as the confocal volume is finite, over longer time scales the protein diffuses out of this region into the bulk, which is reflected in a decrease in emission intensity at times longer than 200 s. The diffusion layer thickness, δ, can be calculated as δ = (2Dt)1/2. The diffusion coefficient of fibrinogen is 2.2 10-7 cm2 s-1,22 leading to a depletion layer thickness of ∼95 μm after 200 s. This thickness is in satisfactory agreement with the confocal height in the z direction using a 10 lens, which is ∼150 μm. Where the response is controlled by diffusion, the fluorescence intensity should depend linearly on t-1/2, where t is time. The inset B of Figure 5 shows that for times longer than 200 s dI/dt depends linearly on t-1/2. Figure 5 also shows that the intensity falls to less than 10% of the initial value after 2000 s, suggesting that almost all of the protein has been desorbed and diffused out of the confocal volume at this time. (21) (a) Arinaga, K.; Rant, U.; Tornow, M.; Fujita, S.; Abstreiter, G. Langmuir 2006, 22, 5560. (b) Rant, U.; Arinaga, K.; Fujita, S.; Yokoyama, N.; Abstreiter, G.; Tornow, M. Langmuir 2004, 20, 10086. (c) Huang, T.; Murray, R. W. Langmuir 2002, 18, 7077. (22) Marder, V. J.; Shulman, N. R.; Carroll, W. R. J. Biol. Chem. 1969, 244, 2111.
296 DOI: 10.1021/la902115e
Mallon et al.
Figure 5. Time dependence of the fluorescence intensity of an Oregon green fibrinogen layer at a gold electrode held at a potential of -1.2 V in aqueous solution with 0.1 M NaHCO3 as the supporting electrolyte. The excitation wavelength is 488 nm. Inset A shows the intensity at λmax as a function of time, and inset B shows a plot of dI/dt vs t-1/2 for the decreasing portion of the curve (i.e., after 200 s).
AFM was used to further probe the extent of desorption ex situ. Figure 6 shows a tapping mode AFM image of the gold substrate before (A) and after (B) applying a potential of -1.2 V to the substrate for ∼200 s. Figure 6A shows a fibrous network, attributable to fibrinogen adsorption, which is absent from bare slides (rms roughness of protein covered surface is 9 nm compared to 2.5 nm for a bare surface). This image is consistent with that shown in Figure 1B but also shows large agglomeration of protein at certain areas. In stark contrast, the image obtained after applying a potential of -1.2 V to the electrode for 180 s (Figure 6B) is indistinguishable from that of a bare slide (rms of 2.3 nm). This result provides further confirmation that the protein layer can be effectively desorbed in less than 200 s by applying a sufficiently negative potential. And, most significantly, unlike more common desorption approaches the protein is ∼100% desorbed. A significant goal is to quantify the structural integrity and function of the protein after desorption, since it will influence potential applications. Figure 7A shows UV-vis spectra of desorbed fibrinogen and fluorescence emission spectra of desorbed Oregon green labeled fibrinogen. Desorption of the protein was achieved by holding a 14 cm2 electrode, on which a fibrinogen layer had been formed, at a potential of -1.2 V for 3 min. The UV-vis spectrum shows a weak but detectable band at ∼280 nm, which is attributable to the aromatic amino acids in the protein. Using a mass extinction coefficient of 15.1 g-1 L cm-1 for fibrinogen22 at this wavelength, a surface coverage of 200 ng cm-2 can be calculated. By comparing the fluorescence intensity of the desorbed protein with a sample of known concentration, the Oregon green fibrinogen yields a slightly lower, but comparable, surface coverage of 70 ng cm-2. This surface concentration yields an average coverage of (4.0 ( 2.3) 10-13 mol cm-2 for the adsorbed protein. This coverage corresponds to an average footprint of ∼400 ( 140 nm2 per molecule, which is in reasonable agreement with the dimensions of fibrinogen reported in the literature (∼45 nm long and 5-8 nm in diameter23). This would suggest the protein is not significantly (23) (a) Rechendorff, K.; Hovgaard, M. B.; Foss, M.; Zhdanov, V. P.; Besenbacher, F. Langmuir 2006, 22, 10885. (b) Wertz, C. F.; Santore, M. M. Langmuir 1999, 15, 8884. (c) Snopok, B. A.; Kostyukevich, E. V. Anal. Biochem. 2006, 348, 222.
Langmuir 2010, 26(1), 293–298
Mallon et al.
Figure 6. Tapping mode AFM images of (A) fibrinogen-coated gold substrates and (B) fibrinogen-coated gold substrates that have been held at a potential of -1.2 V for 180 s.
spread on the surface, which is consistent with the hydrophilic nature of the clean gold surface. To interrogate the impact of the adsorption-desorption cycle on the protein structural integrity, SDS-PAGE was performed for the native fibrinogen and a sample collected from solution following adsorption-desorption, and the results are illustrated in Figure 7B. Significantly, attempts to concentrate the desorbed protein before running the SDSPAGE on a spin filter with a 30 kDa cutoff failed because the fragments of protein ran with the solvent front, indicating a significant change in the protein structure. This conclusion is confirmed by the SDS page studies shown in Figure 7B. While the native fibrinogen, composed of R, β, and γ polypeptides, resolves at 65, 55, and 45 kDa, respectively, only very faint traces can be seen at this level for the desorbed sample even when the gel is clearly overstained. However, there are significant traces at the sub-10 kDa level visible for the desorbed sample, which does not appear for the parent protein. Therefore, it would appear that the electrochemical desorption results in significant fragmentation of the protein. This behavior cannot be attributed solely to electrochemical reduction of the disulfide linkages as these are reduced by the Langmuir 2010, 26(1), 293–298
Article
Figure 7. (A) UV-vis spectrum of desorbed fibrinogen and fluorescence spectra of desorbed Oregon green fibrinogen in aqueous solution with 0.1 M NaHCO3 as the supporting electrolyte. The fluorescence spectrum was obtained using an excitation wavelength of 496 nm. (B) SDS-PAGE analysis of desorbed and native fibrinogen.
β-mercaptoethanol in the SDS sample buffer for both the reference fibrinogen and the desorbed sample. It is likely that the adsorption-desorption process is responsible for the protein decomposition. The observation that protein can be essentially completely desorbed from a metal surface by applying a reductive potential to give severely fragmented strands is attractive from the perspective of protein decontamination and may lead to applications such as the removal of pathogenic proteins, such as prions.
Conclusion Extremely efficient electrochemically induced desorption of fibrinogen from gold has been demonstrated at negative potentials. Both adsorption and subsequent desorption of the protein layer were confirmed from complementary techniques such as capacitance measurements, fluorescence microscopy, and atomic force microscopy. This desorption method has several advantages over other methods described in the literature.6-9 For example, there is no need to alter the composition of the supporting solution, such as changing the ionic strength or pH or adding surfactants, and unlike other reported protocols, desorption is complete within a short time scale. Unlike other methods, this procedure also appears to DOI: 10.1021/la902115e
297
Article
remove practically all of the adsorbed protein. Furthermore, the desorbed protein appears to be extensively fragmented on the basis of both SDS-PAGE and spin filtering. This observation may have interesting consequences for sterilization procedures for surgical instruments, especially as fibrinogen is a key component of blood.
298 DOI: 10.1021/la902115e
Mallon et al.
Acknowledgment. This material is based upon work supported by the Science Foundation Ireland under Grant 05/IN.1/B30. The Irish Research Council for Science, Engineering and Technology is funded by the National Development Plan for Postgraduate Studentship Funding. Sincere thanks to Dr. Barry O’Connell for supporting the AFM measurements.
Langmuir 2010, 26(1), 293–298
