Article pubs.acs.org/ac
Experimental Approach to Controllably Vary Protein Oxidation While Minimizing Electrode Adsorption for Boron-Doped Diamond Electrochemical Surface Mapping Applications Carlee S. McClintock†,§ and Robert L. Hettich*,†,‡ †
Graduate School of Genome Science and Technology, University of Tennessee−Oak Ridge National Laboratory, 1060 Commerce Park, Oak Ridge, Tennessee 37830, United States ‡ Chemical Sciences Division, Oak Ridge National Laboratory, P.O. Box 2008, MS 6131, Oak Ridge, Tennessee 37831, United States ABSTRACT: Oxidative protein surface mapping has become a powerful approach for measuring the solvent accessibility of folded protein structures. A variety of techniques exist for generating the key reagent (i.e., hydroxyl radicals) for these measurements; however, these approaches range significantly in their complexity and expense of operation. This research expands upon earlier work to enhance the controllability of boron-doped diamond (BDD) electrochemistry as an easily accessible tool for producing hydroxyl radicals in order to oxidize a range of intact proteins. Efforts to modulate the oxidation level while minimizing the adsorption of protein to the electrode involved the use of relatively high flow rates to reduce protein residence time inside the electrochemical flow chamber. Additionally, a different cell activation approach using variable voltage to supply a controlled current allowed us to precisely tune the extent of oxidation in a protein-dependent manner. In order to gain perspective on the level of protein adsorption onto the electrode surface, studies were conducted to monitor protein concentration during electrolysis and gauge changes in the electrode surface between cell activation events. This report demonstrates the successful use of BDD electrochemistry for greater precision in generating a target number of oxidation events upon intact proteins.
H
the analyte concentration and the exposure time,5 so efforts were undertaken in this study to investigate and minimize these effects. This research focused on optimizing an electrochemical approach for achieving controllable electrochemical oxidation at the intact protein level for ubiquitin, apomyoglobin, and lysozyme. Our goal involved extensive evaluation of the capacity for the BDD electrode to generate moderately and more controllable oxidized versions of these proteins for surface mapping applications, in contrast to the more heavily modified versions reported previously.7 By maintaining use of model proteins, we focused on charting the parameter modulations necessary to achieve the desired extent of oxidation across the same range of proteins while minimizing their adsorption onto the BDD electrode. This yielded several new experimental developments to produce incrementally progressive levels of oxidized proteins, while decreasing the likelihood of direct protein−electrode interaction, notably with the use of higher flow rates (increased from 1 to 100 μL/min) for a significantly faster exposure of protein to the oxidizing environment inside the electrochemical flow cell chamber. Additionally, a different cell activation approach (variable voltage to maintain a stable
ydroxyl radicals (•OH) can be generated by a range of techniques for probing the solvent accessibility of folded proteins, but most involve approaches (radioactive sources, specialty lasers, caustic chemicals) that may not be readily accessible to researchers. The boron-doped diamond (BDD) thin-film electrode has the capacity to generate •OH from water,1 thus promoting indirect radical-mediated oxidation events while avoiding the use of harsh oxidizing reagents. However, like all other electrodes, the BDD under certain conditions can also achieve oxidation by a direct pathway,2 especially for highly reactive sulfur-containing residues,3 which could potentially involve a physical interaction with the analyte protein that perturbs the native structure. The smooth surface of the inert BDD electrode has a low adsorption profile suitable for analyzing a variety of molecules, including proteins.4 However, proteins at the liquid/solid interfaces can adsorb onto surfaces under favorable conditions,5 including the presence of previously adsorbed molecules.6 Proteins are chemically diverse macromolecules that present heterogeneous surfaces comprised of both polar and/or charged regions as well as hydrophobic patches. Therefore, contact with surfaces may be formed on the basis of both electrostatic and hydrophobic (van der Waals) interactions, in addition to a variety of possible interfacial phenomena that impact the energetic favorability for adsorption to occur.5 If indeed it does occur, the rate of protein adsorption should occur in a manner that is dependent on both © 2012 American Chemical Society
Received: September 4, 2012 Accepted: December 4, 2012 Published: December 4, 2012 213
dx.doi.org/10.1021/ac302418t | Anal. Chem. 2013, 85, 213−219
Analytical Chemistry
Article
ultraviolet (UV) detector placed after the electrochemical cell to monitor absorption at 205 and 280 nm wavelengths, and Xcalibur software (Thermo Fisher Scientific, San Jose, CA) was used to simultaneously record the signal from each wavelength. The 190 nL capacity of the UV (model ULT-UZ-M10) flow cell accommodated flow rates of 10−100 μL/min. A WaveNow potentiostat (Pine Instruments, Raleigh, NC) was used for potentiometric operation of the working electrode to measure the fluctuations in potential (voltage) that occurred during the sustained application of a particular current. Other modes of operation were utilized as follows. The open circuit voltage (OCV) reading can be elicited from the electrochemical assembly once the potentiostat has been properly connected to the working, counter, and reference electrodes. The OCV reading, also known as the resting potential, measures the resting voltage between the anode and cathode in the circuit without an electrical load from the potentiostat. This value can provide feedback on the resistance (or inversely the conductivity) between the electrodes. Cyclic voltammetry (CV) is commonly used to query a particular analyte for the electrochemical potentials at which it becomes oxidized or reduced. Initially, a wide voltage range (+3.0 V to −1.0 V) was explored with the potassium sulfate solution in the electrochemical flow chamber, and a peak was observed in the rising phase at +2.1 V. In between protein oxidation experiments, the range of potentials from +1.0 V to −1.0 V was scanned at a rate of 100 mV/s, with three cycles used in each experiment. The resulting data plots current versus voltage, allowing visualization of the current level required to hit the set voltage end points. If a detectable change of the electrode surface had occurred during a protein oxidation experiment, comparison of the CV plots resulting from scans performed before (the pre-CV) and after (the post-CV) the protein experiment could provide two pieces of critical information. The first was whether the current level at the voltage end points had changed in the first cycle, and the second was whether the change was resolved by subsequent cycles. The first scenario suggests that some level of protein adsorption had taken place. The second provided clues about whether the adsorption was reversible or irreversible. Mass Spectrometry. Intact proteins were measured by mass spectrometry (MS) to evaluate the extent of oxidation resulting from electrochemical oxidation. Control experiments were conducted by passing the protein through the electrochemical cell without voltage activation. Proteins were extracted from the nonvolatile salt solution with standard bed C4 reversephase ZipTips acquired from Millipore (Billerica, MA). Electrospray-ready aliquots were prepared for direct infusion in an aqueous−organic solution consisting of ACN/H2O/FA (50/50/0.1, v/v/v). Samples were infused at 2.5 μL/min by syringe pump (Harvard Apparatus, Holliston, MA) through PEEK tubing directly into the electrospray ionization (ESI) source of a Fourier transform ion cyclotron resonance (FTICR) mass spectrometer (Varian Inc., Palo Alto, CA) equipped with a 9.4 T actively shielded superconducting magnet (Cryomagnetics Inc., Oak Ridge, TN). Time domain ion signals were acquired using 1024 K data points, averaged over 5−50 scans, and converted by fast Fourier transform to the mass-to-charge (m/z) domain using Omega software (Varian, Palo Alto, CA). Secondary Oxidation Quenchers. Overexposure of proteins to •OH after the initial oxidation may cause partially unfolded conformations to be labeled during the experiment, so moderately reactive radical scavengers such as free glutamate
current level) was employed to improve controllability of the intact oxidized protein yield. This change was made because •OH production is presumably governed by the currentdependent Nernst equation,1 and voltage readings can rise concomitant with insulation of the electrode surface for a set current. Therefore, maintaining a set voltage could result in a reduced •OH production over time. Finally, concentrations of both electrolyte (125 mM) and analyte (1−10 μM) protein were altered to better match the oxidant flux, allowing improved control over the amount of electrochemically induced oxidation. These changes facilitated rapid oxidation of proteins in a solution of physiologically relevant ionic strength with a significantly shorter residence time (∼1 s), during which adsorption could potentially occur. These findings provide the groundwork for establishing the broader utility of the BDD electrode for protein structural studies.
■
MATERIALS AND METHODS Materials. All reagents were obtained at the highest purity available and used as supplied: proteins, dithiothreitol (DTT), ammonium acetate, acetic acid (AA) (Sigma-Aldrich, St. Louis, MO), formic acid (FA), potassium sulfate (EMD Chemicals, Darmstadt, Germany), and HPLC-grade solvents (Honeywell Burdick & Jackson, Muskegon, MI). Electrochemistry. The boron-doped diamond (BDD) electrode is a flat disc with a 220 mm2 surface area of which roughly 30 mm2 was accessible through an oval window in a thin Teflon gasket adjacent to the electrode surface. The BDD working electrode was seated along with stainless steel counter/ auxiliary and silver chloride (Ag/AgCl) reference electrodes inside a three-electrode electrochemical flow-by cell (model 5041, ESA, Inc.), shown in Figure 1. Using a gasket with a 50
Figure 1. Boron-doped diamond (BDD) working electrode configuration inside the electrochemical flow-by cell. (A) Interior of the electrochemical flow cell, with the BDD working electrode surface showing through the gasket window. Arrows denote the entry/ electrolysis/exit scheme that the solution travels during the flow over the electrode surface when assembled. (B) Schematic of the 50 μm gasket separating the working and counter electrodes, showing relative positions of entry/exit portals denoted by ° symbols, with dimensions used for calculating the flow chamber volume.
μm thickness to separate the working and counter electrodes, the volume inside the flow chamber was roughly 1 μL. For electrochemical oxidation experiments, a protein solution was loaded into a 1 mL loop connected to a six-port injection valve (model 7125, Rheodyne, Rohnert Park, CA) placed in-line between a Switchos HPLC pump (LC Packings, a division of Dionex, Sunnyvale, CA) and the flow-by electrochemical cell. All components were connected by polyetheretherketone (PEEK) tubing (Upchurch Scientific, Oak Harbor, WA) with 0.005 in. inner diameter, and a grounded metal union was placed immediately after the electrochemical cell. Some experiments involved the use of an Ultimate (LC Packings) 214
dx.doi.org/10.1021/ac302418t | Anal. Chem. 2013, 85, 213−219
Analytical Chemistry
Article
Figure 2. Intact protein mass spectra obtained using FTICR-MS, illustrating progression of oxidation events induced by experimental conditions reported within each spectrum for proteins (A) ubiquitin, (B) apomyoglobin, and (C) lysozyme. Adjustable parameters reported above include current level to modulate the •OH concentration (in microamperes, μA) and flow rate to vary residence time (in μL/min) from approximately 1 to 4 s. Resolvable oxidation peaks occurring at higher m/z relative to unoxidized controls are typically +16 Da heavier than the native molecular mass when deconvoluted.
design). The rapid flow rate virtually eliminated the voltage oscillation attributable to the nucleation of electrolytic bubbles (water oxidation by a BDD electrode yields gases H2, O2, and O3)11 on the electrode surface by hastening their exit from the cell, which should extend the useful lifetime of the BDD electrode. Bubble formation can damage the BDD film by gradually decreasing the electrode surface area contacted by the solvent, which threatens the dwindling interface area with a rapidly increasing current density. The performance of the original BDD electrode declined over the course of one year, likely due to a combination of factors. These included damage caused by bubble accumulation at low flow rates and fouling that occurred during electrolysis of high concentrations of organic electrolytes and analytes. Observations discussed below suggest that a moderate hysteresis effect may contribute to increasing voltages with the application of a set current level over subsequent runs, whether due to adsorption of electrolyte, protein, or both. Changes in electrochemical parameters included a decrease in the applied current from 500 μA (+2.1 V with older BDD electrode) to 100−300 μA (ranging from +1.4 V to +2.1 V over time with the new BDD electrode). Along with lower protein concentrations (from 50 μM to 1−5 μM) and higher flow rates (25−100 μL/min), these currents resulted in controllable protein oxidation along a continuum from light to heavy oxidation (Figure 2). Although higher potentials (greater than or equal to +2.3 V) are reportedly required for discharging water to generate •OH in acidic solution,12 the presence of organic molecules has been shown to reduce the water discharge potential by up to 200 mV.1 Protein oxidation was accomplished in an aqueous potassium sulfate solution (pH 6.0) at half of the previously used concentration (from 250 mM to 125 mM), and a key advantage for this study was its minimal
have been used to better control the oxidant exposure.8 This measure was not taken to avoid possible complications due to the electrochemical adsorption that would interfere with protein analysis. Inclusion of catalase tetramer in reaction solutions to decompose peroxide has been shown to recoup methionine utility in oxidative studies.9 This factor was also omitted not only to minimize a potential adsorption and/or oxidant flux problem, but also because catalase has a metal prosthetic group that could be directly oxidized. Peptidyl hydroperoxides formed during oxidation of aliphatic side chains can also oxidize methionine, so methionine amide has been used to mitigate this source of uncontrolled protein oxidation.10 The background oxidation discussed below was of more concern, so the use of DTT as an antioxidant was tested in some control experiments to determine whether this phenomenon might be due to unexpected oxidant production during the digestion process. Nitrogen gas was also used in these controls to displace oxygen in the head space atop the reaction solution.
■
RESULTS AND DISCUSSION Experimental Refinement of Electrochemical Protein Oxidation Conditions. Several key modifications for the BDD electrode operation facilitated rapid electrochemical protein oxidation, favoring an indirect •OH-mediated route. This pathway is in contrast to a direct (heterogeneous) electron transfer event that involves interaction of proteins with the electrode surface, which would be of greater concern in lower flow-rate conditions. The residence time for the protein passing through the ∼1.7 μL electrochemical flow chamber was lowered from nearly 2 min at a flow rate of 1 μL/min to 1 s at 100 μL/min (note: these calculated values assume laminar flow, though some turbulence occurs with the current flow cell 215
dx.doi.org/10.1021/ac302418t | Anal. Chem. 2013, 85, 213−219
Analytical Chemistry
Article
UV light absorption.13 An interesting consideration to note about this particular electrolyte is whether the sulfate anion used in this protocol was radicalized (SO42− → e− + •OSO3−) by direct oxidation to either mediate protein oxidation or coalesce into peroxydisulfate (S2O82−), whose reactivity is similar to hydrogen peroxide.14 If so, this reactivity could confer advantages to protein studies,15,16 with the same cautions that generally apply to peroxides. The final reaction product would effectively mirror the most common oxidative mass shift for hydroxyl radical labeling (typically +16 Da), because the protein-attached potassium sulfate (KSO4, +135 Da) would only be an intermediate state that is then converted to an alcohol.17 The oxidation−reduction potential of peroxydisulfate is +2.1 V, which corroborates the oxidation peak observed when the voltage range is scanned with the potassium sulfate solution alone inside the electrochemical flow cell (data not shown). Additionally, this potential most often results in moderate protein oxidation when flow rate was adjusted appropriately. Mass Spectrometry Analysis of Oxidized Intact Protein. All three model proteins were reproducibly oxidized at variable levels depending on the controlled flow rate (25− 100 μL/min) or electrode current (100−300 μA) employed, while occasionally producing some evidence of interaction with the electrode surface. After passage through the activated electrochemical cell, proteins were measured by ESI-FTICRMS for the oxidation level as a function of voltage-dependent •OH flux and exposure time. As shown in Figure 2, intact proteins consistently showed an increase in overall number of oxidation events when either flow rate or protein concentration was decreased, or when current magnitude was increased. Ideally, proteins would be oxidized on the microsecond timescale or less, but it is impractical to reduce the protein residence time inside this electrochemical flow chamber to that level. The current design of this electrochemical flow cell forces the protein solution to undergo a 90° change in direction while entering and exiting the flow chamber, producing fluidic turbulence and backpressure that caused leakage within the cell assembly at flow rates higher than 100 μL/min. In order to accommodate higher flow rates, a linear flow cell design that facilitates laminar flow would be necessary for achieving a significantly faster exposure time, possibly on the order of milliseconds. Since the timescale reported here was not reduced to an optimal level for surface-mapping applications, it was appropriate to omit the usage of scavengers to prevent magnifying any hysteresis effect. Regardless, sulfur-containing residues are electroactive, so they may be targeted by both indirect (mediated by radicals or peroxides) and direct (electron transfer) oxidation mechanisms. Experimentation with small molecules suited for the electrochemical cell that prohibit overoxidation by radical-mediated or peptidyl hydroperoxide sources would be advantageous to further development of this technique. Ubiquitin revealed a bias toward the mono-oxidized form in a variety of experimental conditions, which remained the base peak upon overoxidation (Figure 2A). Methionine is the solitary sulfur-containing residue present in ubiquitin, and its N-terminal location minimizes the likelihood of structural distortion upon oxidation. However, even a sample that appears primarily mono-oxidized may contain an ensemble of overoxidized proteins, as the most relatively abundant monooxidized form may be masking all other species present. An indication of this may be whether the native form is present
alongside the mono-oxidized version, with its absence suggesting overoxidation has occurred. Apomyoglobin showed a unimodal distribution in the moderately oxidized sample (300 μA at 50 μL/min) centered around +16 Da, with flanking peaks representing the native and dioxidized forms (Figure 2B). Similar to ubiquitin, this signature seems to correspond with mono-oxidation of the two methionines present in myoglobin. With heavier oxidation, additional distributions appear starting at greater masses than the dioxidized species, suggesting the presence of partially unfolded conformers that expose more reactive sites. Lysozyme has ten sulfur-containing residues, including eight cysteines which can each be triply oxidized (+48 Da), so this correlative phenomenon is not readily observable due to the large number of mass shifts possible (28 events of +16 Da each) on these residues alone (Figure 2C). It should be noted that the oxidation signature in the larger proteins suffered interference by satellite peaks that were also observed in the unoxidized stock solution. Investigation of Possible Protein−Electrode Interactions. Higher flow rates effectively increased the rate of protein mass transport through the electrochemical flow cell to reduce the timeframe for adsorption opportunities. Inherent physical factors summarized in Table 1, however, may influence the Table 1. Physicochemical Features of Model Proteins Subjected to Electrochemical Oxidation protein
molecular weight
isoelectric point (pI)
hydropathicity (GRAVY)
Cys
Met
ubiquitin apomyoglobin lysozyme
8565 16951 14306
6.6 7.4 9.3
−0.49 −0.40 −0.47
0 0 8
1 2 2
propensity of a protein to interact with a proximate electrode surface. The protein-specific isoelectric point (pI) carries implications that a protein will become charged in proportion to the magnitude of difference between the intrinsic pI value relative to the solution pH, although electrolyte ions can provide electrostatic shielding with counterions negating the charged functional groups. It is worth noting that the oxidation process can alter the pI by causing losses of both acidic (decarboxylation, D/E is −30 Da) and basic (deguanidation, R is −43 Da) functional groups.18 Furthermore, surface electron density, and thus polarity, may be increased through the addition of electronegative oxygen atoms via hydroxyl or carbonyl groups.18 The grand average of hydropathicity (GRAVY) index19 reflects the overall relative hydrophobicity or hydrophilicity of the macromolecule with more positive or negative values, respectively. Enhancement of hydrophilicity by oxygen incorporation would cause the GRAVY value to become increasingly negative. Before such oxidative alterations, the intrinsic values suggest that ubiquitin would be the least likely of the three model proteins tested to experience any interaction with the BDD electrode surface. Taken together, lysozyme should be the most electrostatically attracted, while apomyoglobin has a higher propensity for making hydrophobic contacts. Lysozyme also contains cysteines, and previous studies have documented the adsorption of free cysteine to various electrode surfaces.20 However, BDD electrodes were shown to perform much better than glassy carbon electrodes because they did not suffer the same magnitude of cysteine adsorption.21 The large bovine serum albumin (BSA) protein 216
dx.doi.org/10.1021/ac302418t | Anal. Chem. 2013, 85, 213−219
Analytical Chemistry
Article
minutes. Regardless, the UV traces in Figure 3 indicate that evidence of adsorption appears relatively minor, as virtually all of the protein continues passing through the activated electrochemical cell. The extended time for protein to flush out of the cell suggests that some protein lingers during the exponential decay of the electrode voltage; however, the control experiments (data not shown) revealed a similar phenomenon possibly attributable to the aforementioned turbulence, resulting in incremental dilution over time rather than a distinctly exiting protein bolus. Interrogation of the BDD electrode surface integrity was gauged by cyclic voltammetry (CV) scans conducted between protein oxidation experiments. Changes were often noted in current magnitudes and/or the overall shape of the CV plots after a protein experiment relative to the one acquired just prior to it (Figure 4), which suggested that the electrode surface had been altered in some manner. In Figure 4A, an oxidation experiment with apomyoglobin resulted in more gradual current changes instead of the sharper inflections seen in the earlier plot, especially in the rising phase from cathodic (negative) potentials. A dramatic reduction in current magnitude was observed for the +1.0 V end point in the first cycle, followed by generally lower current magnitudes corresponding to the voltage range end points. Figure 4B also shows that the current level was sharply reduced at the first cycle end point after lysozyme oxidation, although the subsequent cycles showed different characteristics of significantly greater current magnitudes at cathodic potentials. These plots suggested that a fraction of each protein was adsorbed onto an electrode surface after the electrochemical oxidation experiment. Because these proteins are both positively charged at pH 6.0, it stands to reason that they interact with the negatively charged cathode instead of the BDD anode. However, the pI of apomyoglobin is much closer to the solution pH, so its postulated interaction may be more hydrophobic than electrostatic, which would corroborate its GRAVY value relative to the other proteins tested. Protein adsorption did not hinder recovery for further analysis even at the low micromolar concentrations tested. However, any change in protein concentration would impact oxidation kinetics due to an altered match with the flux of hydroxyl radicals over the experimental time frame23 and interfere with the structural interpretation of circular dichroism (CD) results.24 Two of these model proteins were already shown to be structurally resilient to oxidative damage by CD measurements and intact protein mass spectrometry,25 so the latter metric alone was used for gauging structural integrity.
that contains more than thirty cysteines showed minimal, though not nonzero, adsorption onto a BDD electrode.22 In an effort to capture evidence of adsorption occurring during electrolysis, Figure 3 illustrates the concentrations of
Figure 3. Protein concentration (5 μM) was monitored with a UV detector (205 nm) to identify changes during electrode activation for (A) ubiquitin, (B) apomyoglobin, and (C) lysozyme. Baseline absorption of potassium sulfate solution was obtained prior to injecting protein solution at a flow rate of 50 μL/min into the electrochemical cell flow chamber over a 20 min duration. Five minutes after protein introduction, the BDD electrode was activated with a 200 μA current, and oxidized protein samples were collected once a steady state was reached, as evidenced by a voltage plateau around +2.05 V.
each protein exiting the electrochemical cell, as measured by ultraviolet (UV) detection of peptide bonds at a 205 nm absorbance. Electrochemical cell activation induced a 10−30% increase relative to the baseline protein UV absorbance that persisted throughout the duration of electrolysis. Conversely, a transient drop below the protein baseline was observed upon either electrode activation or deactivation in all three cases. Both ubiquitin and apomyoglobin showed an initial absorbance drop of 2−3% relative to the total protein signal just as the BDD was activated. After the initial dip, ubiquitin absorbance ramped up to a plateau, whereas apomyoglobin absorbance increased briefly and then sharply declined to roughly 5% below the baseline for a 2 min duration before ramping back up to a plateau for the last half of the electrolysis duration. Lysozyme did not suffer an initial drop, but the absorbance dipped by 5% upon BDD deactivation before returning to the baseline. These results suggested that while ubiquitin appeared to undergo an early electrolytic interaction, both apomyoglobin and lysozyme suffered more adsorption by a mixture of phenomena due to the differing UV absorbance signatures. We speculate that a high cysteine content contributed to the later evidence of slight lysozyme release from the BDD electrode surface, whereas the delayed apomyoglobin dip may have required early nucleation to induce a hydrophobic buildup that saturated after a few
■
CONCLUSIONS This research has demonstrated that proteins can be rapidly and controllably oxidized under variable electrochemical conditions. Several attributes of the BDD electrode are particularly attractive for this work, including a documented ability to produce reactive oxygen species directly from water within its wide working potential window. The additional contribution to the oxidant flux from the sulfate radical may offer an interesting advantage for surface mapping when better understood and properly harnessed. Other studies have demonstrated the minimal adsorption of the BDD electrode due to its smooth and relatively inert surface, although our studies suggest a nonzero value, as some evidence of adsorption was detected in this investigation. The technology holds potential for mapping the dynamically exposed surfaces of 217
dx.doi.org/10.1021/ac302418t | Anal. Chem. 2013, 85, 213−219
Analytical Chemistry
Article
Figure 4. Monitoring electrode changes with cyclic voltammograms (CVs) performed before (gray) and after (black) oxidation of (A) 5 μM apomyoglobin and (B) 5 μM lysozyme. The BDD working electrode potential range was scanned from −1.0 V and +1.0 V, at a rate of 100 mV/s over six scan segments (2.5 cycles). During these CV experiments, only potassium sulfate solution was pumped through the electrochemical flow cell. Left panels depict typical CVs as potential versus current, while right panels provide chronological detail showing changes in the measured current occurring over successive cycles.
solution-phase proteins, as the flow-by design is highly compatible with increasingly rapid exposures of analyte protein to the oxidizing environment inside the electrochemical chamber. However, modification of the cell design would be necessary to achieve laminar flow, which would facilitate a faster exposure time frame. Overall, this methodology for performing and detecting electrochemical protein oxidation has generally expanded knowledge of how the BDD electrode may become valuable to future protein research.
■
(3) Terashima, C.; Rao, T. N.; Sarada, B. V.; Kubota, Y.; Fujishima, A. Anal. Chem. 2003, 75, 1564−1572. (4) Fujishima, A.; Einaga, Y.; Rao, T. N.; Tryk, D. A., Diamond Electrochemistry: Elsevier B.V.: Amsterdam, 2005. (5) Haynes, C. A.; Norde, W. Colloids Surf., B 1994, 2, 517−566. (6) Tie, Y.; Calonder, C.; Van Tassel, P. R. J. Colloid Interface Sci. 2003, 268, 1−11. (7) McClintock, C.; Kertesz, V.; Hettich, R. L. Anal. Chem. 2008, 80, 3304−3317. (8) Hambly, D. M.; Gross, M. L. J. Am. Soc. Mass Spectrom. 2005, 16, 2057−2063. (9) Sharp, J. S.; Tomer, K. B. Biophys. J. 2007, 92, 1682−1692. (10) Saladino, J.; Liu, M.; Live, D.; Sharp, J. S. J. Am. Soc. Mass Spectrom. 2009, 20, 1123−1126. (11) Michaud, P. A.; Panizza, M.; Ouattara, L.; Diaco, T.; Foti, G.; Comninellis, C. J. Appl. Electrochem. 2003, 33, 151−154. (12) Marselli, B.; Garcia-Gomez, J.; Michaud, P. A.; Rodrigo, M. A.; Comninellis, C. J. Electrochem. Soc. 2003, 150, D79−D83. (13) Aitken, A.; Learmonth, M. The Protein Protocols Handbook; Walker, J. M., Ed.; Humana Press Inc.: New York, 1996; pp 3−6. (14) Snook, M. E.; Hamilton, G. A. J. Am. Chem. Soc. 1974, 96, 860− 869. (15) Bridgewater, J. D.; Vachet, R. W. Anal. Biochem. 2005, 341, 122−130. (16) Gau, B. C.; Chen, H.; Zhang, Y.; Gross, M. L. Anal. Chem. 2010, 82, 7821−7827. (17) Sethna, S. M. Chem. Rev. 1951, 49, 91−101. (18) Xu, G.; Chance, M. R. Chem. Rev. 2007, 107, 3514−3543. (19) Kyte, J.; Doolittle, R. F. J. Mol. Biol. 1982, 157, 105−132. (20) (a) Hager, G.; Brolo, A. G. J. Electroanal. Chem. 2003, 550, 291−301. (b) Liu, Z.; Wu, G. Spectrochim. Acta 2006, 64, 251−254. (21) Spãtaru, N.; Sarada, B. V.; Popa, E.; Tryk, D. A.; Fujishima, A. Anal. Chem. 2001, 73, 514−519. (22) Shin, D.; Tryk, D. A.; Fujishima, A.; Merkoçi, A.; Wang, J. Electroanalysis 2004, 17, 305−311. (23) Tong, X.; Wren, J. C.; Konermann, L. Anal. Chem. 2007, 79, 6376−6382. (24) (a) Kelly, S. M.; Jess, T. J.; Price, N. C. Biochim. Biophys. Acta 2005, 1751, 119−139. (b) Miles, A. J.; Whitmore, L.; Wallace, B. A. Protein Sci. 2005, 14, 368−374.
AUTHOR INFORMATION
Corresponding Author
*Email:
[email protected]. Present Address §
University of Tennessee, Department of Biochemistry/ Cellular & Molecular Biology, Walters Life Sciences Building, 1414 Cumberland Avenue, Knoxville, TN 37996.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors extend thanks to Dr. Vilmos Kertesz for providing electrochemical expertise. C.M. acknowledges financial support from the UTK-ORNL Graduate School of Genome Science and Technology. This research was supported by the National Institutes of Health General Medicine Section under Grant R01-GM070754. Oak Ridge National Laboratory is managed by University of Tennessee−Battelle LLC for the Department of Energy.
■
REFERENCES
(1) Kapalka, A.; Foti, G.; Comninellis, C. Electrochim. Acta 2009, 54, 2018−2023. (2) Zhi, J. F.; Wang, H. B.; Nakashima, T.; Rao, T. N.; Fujishima, A. J. Phys. Chem. B 2003, 107, 13389−13395. 218
dx.doi.org/10.1021/ac302418t | Anal. Chem. 2013, 85, 213−219
Analytical Chemistry
Article
(25) (a) Venkatesh, S.; Tomer, K. B.; Sharp, J. S. Rapid Commun. Mass Spectrom. 2007, 21, 3927−3936. (b) Downard, K. M.; Maleknia, S. D.; Akashi, S. Rapid Commun. Mass Spectrom. 2012, 26, 226−230.
219
dx.doi.org/10.1021/ac302418t | Anal. Chem. 2013, 85, 213−219
