Article pubs.acs.org/ac
Single-Molecule Studies of Intrinsically Disordered Proteins Using Solid-State Nanopores Deanpen Japrung,†,‡ Jakob Dogan,§ Kevin J. Freedman,∥ Achim Nadzeyka,⊥ Sven Bauerdick,⊥ Tim Albrecht,† Min Jun Kim,# Per Jemth,§ and Joshua B. Edel*,† †
Department of Chemistry, Imperial College London, South Kensington, SW7 2AZ, London, United Kingdom National Nanotechnology Center, National Science and Technology Development Agency, Thailand Science Park, Pathumthani, 12120 Thailand § Department of Medical Biochemistry and Microbiology, Uppsala University, Uppsala, Sweden ∥ Department of Chemical and Biological Engineering, Drexel University, Philadelphia, Pennsylvania 19104, United States ⊥ Raith GmbH, Konrad-Adenauer-Allee 8, Dortmund, 44263, Germany # Department of Mechanical Engineering and Mechanics, Drexel University, Philadelphia, Pennsylvania 19104, United States ‡
S Supporting Information *
ABSTRACT: Partially or fully disordered proteins are instrumental for signal-transduction pathways; however, many mechanistic aspects of these proteins are not well-understood. For example, the number and nature of intermediate states along the binding pathway is still a topic of intense debate. To shed light on the conformational heterogeneity of disordered protein domains and their complexes, we performed single-molecule experiments by translocating disordered proteins through a nanopore embedded within a thin dielectric membrane. This platform allows for single-molecule statistics to be generated without the need of fluorescent labels or other modification groups. These studies were performed on two different intrinsically disordered protein domains, a binding domain from activator of thyroid hormone and retinoid receptors (ACTR) and the nuclear coactivator binding domain of CREB-binding protein (NCBD), along with their bimolecular complex. Our results demonstrate that both ACTR and NCBD populate distinct conformations upon translocation through the nanopore. The folded complex of the two disordered domains, on the other hand, translocated as one conformation. Somewhat surprisingly, we found that NCBD undergoes a charge reversal under high salt concentrations. This was verified by both translocation statistics as well as by measuring the ζ-potential. Electrostatic interactions have been previously suggested to play a key role in the association of intrinsically disordered proteins, and the observed behavior adds further complexity to their binding reactions.
A
domains at the single-molecule level, we have performed nanopore translocation experiments on two IDPs that bind to each other to form a structurally well-defined complex (Figure 1a). Nanopores are single-molecule sensors for the label-free detection and structural analysis of biological polymers such as DNA, RNA, polypeptides, and DNA−protein complexes in solution.4 The key component of a nanopore sensor consists of a nanoscale channel, with typical dimensions less than 100 nm,
majority of human signaling proteins contain long disordered regions. In fact, as much as one-quarter of all mammalian proteins could be disordered.1 It is becoming clear that such intrinsically disordered proteins (IDPs) are fundamental for multicellular organisms, which rely on complex signal-transduction pathways.2 One likely reason that IDPs are so common is that one IDP can have multiple binding partners.3 In these cases the IDP may adopt a distinct conformation for each partner in the bound complex. In its free form an IDP is very flexible and may sample many more structural conformations, all having a similar free energy. To learn more about the conformational heterogeneity of IDP © 2013 American Chemical Society
Received: December 3, 2012 Accepted: January 17, 2013 Published: January 17, 2013 2449
dx.doi.org/10.1021/ac3035025 | Anal. Chem. 2013, 85, 2449−2456
Analytical Chemistry
Article
Figure 1. (a) Schematic representations of ACTR and NCBD before association and the well-folded binary complex between the two intrinsically disordered protein domains (Protein Data Bank (PDB) file 1KBH). In the surface representation, blue denotes a positive surface charge while red denotes a negative charge. (b) A schematic picture of the nanopore setup. A SiNx membrane separates the cis (−) and trans (+) compartments. Initially all protein samples were added to the cis side with only buffer in the trans reservoir.
therefore unexpected to observe current modulation due to the translocations of NCBD. We show that close to physiological salt concentrations NCBD undergoes a charge reversal (i.e., the counterions in solution accumulate near the protein surface). This was validated at the single-molecule level using nanopores and in bulk solution by measuring the ζpotential. The modulation of the electrostatic environment has been previously suggested to play a key role in the association of intrinsically disordered proteins, and the observed behavior adds further complexity to their binding reactions. This highlights the effect salt may have on the electrostatics of highly charged IDPs must be considered in their interactions with binding partners. To our knowledge this is the first time such an effect has been experimentally observed in IDPs.
that penetrates an insulating membrane. In this work solid-state silicon nitride (SiNx) membranes are used. The membrane separates two aqueous reservoirs containing both electrolyte (typically KCl or NaCl solutions) and analyte (cis reservoir only) with the nanopore acting as a conduit between them. Upon application of an electrical potential, charged molecules are driven or translocated through the nanopore (Figure 1b).5 Importantly, in such single-molecule experiments the translocation time along with the amplitude of the electrical current measured directly depends on the charge and conformation of the molecule. It is thus possible to directly observe different conformers and charge states in an ensemble of proteins. To date, there have been a handful of studies performing protein translocations within solid-state membranes. For example, transport dynamics has been extensively studied as have protein interactions with the nanopore surface.6−8 Furthermore, studies on voltage dependence on the translocation properties have also recently been investigated.8 Until now the majority of work using solid-state nanopores relates to using relatively large ordered proteins relative to the nanopore size. This study deals with understanding the interaction between small IDP domains by interpretation of the ionic current blockade. The IDP domains used in our study include the binding domain from activator of thyroid hormone and retinoid receptors (ACTR) and the nuclear coactivator binding domain of CREB-binding protein (NCBD), are involved in activation of gene expression upon activation of nuclear receptors.9−11 The interaction of these two domains is particularly interesting since both ACTR and NCBD are disordered as shown previously.12−14 It has been demonstrated that NCBD has molten globular properties with inherent flexibility, low stability, an apparent noncooperative thermal unfolding as measured by circular dichroism, a significant secondary structure content, and a small core,12,13,15,16 whereas ACTR is fully unfolded, but still transiently forms some secondary structure in its free state.13−15 Therefore, a nanopore analysis is potentially very useful to study any conformational heterogeneity in the respective protein ensemble. For example, bound and unbound protein states would be expected to occupy different volumes within the pore resulting in a unique current modulation. In this paper we show that we can clearly discern between different conformations of both ACTR (8.1 k, theoretical pI = 4.2, as calculated by ProtParam, http://web.expasy.org/ protparam/) and NCBD (6.7 kDa, theoretical pI = 11.1) when translocating through the nanopore. Furthermore the folded ACTR/NCBD complex only exhibits a single conformational state. Interestingly, at pH 8, both ACTR and the complex have a net negative charge while NCBD has a net positive charge. Under the experimental conditions used it was
■
EXPERIMENTAL SECTION Protein Preparation. The DNA sequence of human NCBD (2058−2116)12 was amplified by polymerase chain reaction (PCR) using a human brain cDNA library as template, whereas the sequence of human ACTR (residues 1018− 1088)12 was purchased from GENEART (Germany) and PCRamplified with a Met-Gln extension at its N-terminus. Each of the two sequences was inserted into a modified pRSET vector (Invitrogen) and was preceded by a DNA sequence encoding a thrombin cleavage site (LVPRGS) and an N-terminal hexahistidine-tagged lipoyl fusion protein. BL21(DE3) plysS cells (Invitrogen) were grown in 2xTY media at 37 °C until OD600 reached 0.7−0.8 and then induced with 1 mM isopropylβ-D-thiogalactopyranoside to overexpress the fusion protein at 18 °C overnight. Cells were lysed by sonication and centrifuged at 4 °C, after which the supernatant was separated from the cell debris, passed through a 0.2 μm filter (Sarstedt), and then loaded onto a Ni−sepharose fast flow (GE Healthcare) column. The column was washed with buffer (50 mM Tris pH 8.5, 400 mM NaCl, 13 mM imidazole), after which the fusion protein was eluted with a buffer containing 250 mM imidazole and then dialyzed against 20 mM Tris (pH = 8.0). The protein solution was subjected to an anion-exchange purification step using a Source-30 Q (GE Healthcare) column and a sodium chloride gradient. Thrombin (GE Healthcare) was added to the protein solution in order to cleave off the lipoyl protein and then loaded onto a Ni−sepharose column, which separated the Histagged lipoyl protein from ACTR or NCBD. This was followed by a reversed-phase chromatography step, using C-8 (ACTR) or C-18 (NCBD) columns (Grace Davison Discovery Sciences). The identity of purified ACTR or NCBD was verified by matrix-assisted laser desorption ionization time-offlight (MALDI-TOF) mass spectrometry. The NCBD concentration was determined by measuring absorbance at 280 nm. The absorbance of ACTR, which lack Phe, Tyr, and Trp, was 2450
dx.doi.org/10.1021/ac3035025 | Anal. Chem. 2013, 85, 2449−2456
Analytical Chemistry
Article
pass cutoff filter. Excitation was at 280 nm. The concentration of NCBDY2108W was held constant at 1 μM, while varying the concentration of ACTR. The dissociation rate constants were obtained by mixing a preformed NCBDY2108W/ACTR complex solution with an excess of wild-type NCBD and monitoring the fluorescence change.
measured at 205 nm, and the concentration was calculated using an extinction coefficient of 31.9 mL mg−1 cm−1, obtained from amino acid analysis. Nanopore Fabrication. SiNx membranes with 70 nm thicknesses were fabricated using standard photolithography and KOH wet etching.17,18 Nanopores with diameters ranging between 16 and 20 nm were milled using a focus ion beam (FIB)/scanning electron microscope (SEM) instrument (Carl Zeiss XB 1540 Cross-Beam; ion acceleration voltage at 30 kV; milling current at 1 pA). Some of the nanopores were also milled by Raith GmbH (Konrad-Adenauer-Allee 8, Dortmund, 44263, Germany). Membranes were cleaned using Piranha solution for 20 min followed by O2 plasma for 5 min on both sides. Binding Assay. The binding reactions were performed using 1.5 μg of each ACTR and NCBD protein in 1 M KCl, 10 mM Tris·HCl pH 8.0 (TB) or 150 mM NaCl, 20 mM NaPi, pH 7.4 (CB). All reactions were incubated at room temperature or 37 °C for 30 min before running on 8% native polyacrylamide gel (Supporting Information Figure S3). Translocation Experiments. Membranes containing a 18 ± 2 nm nanopore were assembled in a Teflon cell and sealed with poly(dimethylsiloxane) (PDMS) rings. A volume of 900 μL of TB was filled in both sides of the cell. Then, 0.125 mm Ag/AgCl electrodes were connected to the cells and patch clamp amplifier (Axopatch 200B, Molecular Devices, CA, U.S.A.). Current−voltage (I−V) curves were measured between −500 and +500 mV (Supporting Information Figure S1) to calculate the pore conductance and diameter by following the equation used previously.19 Aliquots of ACTR, NCBD, and ACTR/NCBD complex were added in the cis side, and positive bias voltages were applied. Ionic current traces filtered at 10 kHz and sampled at 50 kHz. Ionic current level and translocation times were analyzed with Clampfit 10.2 (Axon Instrument) and Origin 8.5. Dynamic Light Scattering and ζ-Potential Studies. The hydrodynamic diameter and ζ-potential of the proteins were determined using a Zetasizer Nano-ZS (Malvern Instruments Ltd.). On the basis of the molecular weight of ACTR and NCBD, a suitable protein concentration in the range of 30−80 μM was used for sizing and ζ-potential readings. All proteins were diluted in KCl (50−1000 mM) and 10 mM Tris (pH 8). To measure the hydrodynamic diameter, solutions were transferred to a clear disposable 1 mL cuvette and allowed to equilibrate for 2 min. Dynamic light scattering (DLS) analyzes the velocity distribution of a moving particle by measuring fluctuations in light scattering intensity that occur as a result of Brownian motion. Using the Stokes−Einstein equation, the bulk hydrodynamic radius, or diameter, can then be calculated. The ζ-potential measurements were conducted by first transferring the solution to another 1 mL clear ζ-potential cuvette (DTS1060, Malvern). The electrophoretic mobility of the sample was measured and converted into the ζ-potential by applying the Henry equation. All measurements were repeated at least three separate times in a temperature-controlled chamber at 25 °C. Stopped-Flow Binding Measurements. Binding kinetics measurements of NCBD/ACTR were performed at 277 K using an upgraded SX-17MV stopped-flow spectrometer (Applied Photophysics, Leatherhead, U.K.), by following the fluorescence change of a tryptophan variant of NCBD (NCBDY2108W, where the residue Y2108 was replaced by a tryptophan) upon ACTR binding (ref 16) using a 320 nm long-
■
RESULTS AND DISCUSSION ACTR and NCBD Binding Under High-Salt Conditions. It is well-documented that the translocation frequency of a charged molecule through a nanopore is improved by performing experiments at high salt concentration.20,21 Furthermore the signal-to-noise ratio of the detected ionic current blockade is also improved when using a high ionic strength. Proteins are often present as ensembles of distinct conformations. The relative stabilities and thus populations of such conformations may be modulated by the ionic strength. At neutral pH, and as calculated from the primary structures, ACTR, NCBD, and the complex have net charges of −8e, +6e, and −2e, respectively. At the same pH the SiNx nanopore walls are negatively charged with a ζ-potential of approximately −21 mV.7 By adding the proteins to the cis reservoir (negative potential), the negatively charged proteins can be driven electrophoretically from the cis to trans reservoirs. Due to the negatively charged walls, electroosmotic flow is in the opposing direction and will not contribute to inducing translocation events. Therefore, both ACTR and the complex would be expected to translocate through the nanopore into the trans side (positive potential) at neutral pH. NCBD on the other hand, due to its net positive charge, would be expected to remain in the cis reservoir. Surprisingly, as will be discussed later, translocation events between the cis and trans reservoirs are observed at pH 8 indicating that NCBD is acting as a negatively charged protein. To investigate this phenomenon, ζpotential measurements were carried out at varying KCl concentrations (50−1000 mM) in a 10 mM Tris buffer at pH 8.0. At salt concentrations below 250 mM a positive ζpotential was observed, while at higher concentrations the ζpotential gradually became more negative (Figure 2).
Figure 2. Effect of KCl concentration (50−1000 mM, pH = 8.0) on ζpotential of NCBD. A ζ-potential of 0 mV is denoted by a dashed line.
Experimental translocation events were obtained using a KCl concentration of 1000 mM. At this concentration the ζpotential was measured to be −1.1 ± 0.4 mV, which supports the translocation of NCBD (the cis to trans migration). The sign reversal is attributed to the buildup of a counterion cloud surrounding NCBD which not only results in neutralizing the existing charge but through the high interaction energies between negative ions results in a large local negative charge dominating. Although a decrease in ζ-potential as a function of 2451
dx.doi.org/10.1021/ac3035025 | Anal. Chem. 2013, 85, 2449−2456
Analytical Chemistry
Article
Figure 3. Translocation traces of ACTR, NCBD, and the ACTR/NCBD complex. It should be noted that the background noise levels in all examples are identical and only appear different due to the scales at which they are plotted.
to I = 0.20 M.16 The dissociation rate constant of binding was similar at the three salt concentrations (1.8 s−1 at I = 0.023 M, 2.6 s−1 at I = 0.20 M, and 1.4 s−1 at I = 0.98 M). Hence, the association kinetics is clearly affected going from low to physiological salt conditions, but less so upon further increase of ionic strength. Thus, formation of a counterion cloud better explains the effect on konapp than classical long-range electrostatic screening. In agreement with this notion, mechanismspecific electrostatic effects were recently demonstrated for the interaction between the cell cycle regulator p27 upon binding to cyclin A.27 ACTR and NCBD Translocation Studies. Translocation studies were performed in order to (1) further understand the translocation behavior of NCBD and (2) attempt to understand the folding and binding pathways of ACTR, NCBD, and the complex at the single-molecule level. Examples of current time traces for each sample at an applied bias of +300 mV are shown in Figure 3. Prior to going through a detailed statistical analysis of the observed signal, it is important to make several qualitative observations. First, the magnitude of the most likely translocation times ranges from several tenths of milliseconds for NCBD and the complex, to up to several milliseconds for ACTR. This is indeed very slow compared to other much larger biopolymers, such as λ-DNA, which typically translocates in a few milliseconds under similar conditionsdespite the fact that it is orders of magnitude longer and the charge density is roughly comparable.28 On the other hand the time scales for NCBD and the complex are not significantly different to those reported by other groups for similarly sized proteins.29,30 For example, Stefureac et al. performed translocation and folding studies on HPr, which is a 9.1 kDa protein with a net charge of −2e.30 Translocation times were also on the order of a tenth of a millisecond. Interestingly, they came to the conclusion that only proteins smaller than the dimension of the pore can translocate but do so in a folded conformation. This is analogous to the regime we are working in. The translocation
salt concentration is known, very few studies exist where a charge reversal has been observed.22 For example, Firnkes et al. also observed strong ζ-potential dependence on the KCl concentration for avidin (pI = 9.4, net charge at pH 8.0 = +4e); however, in their case a charge reversal was not observed.7 Our results are perhaps complementary to the studies previously performed that experimentally and theoretically observe either a decrease in the net charge or a charge reversal as a function of increasing salt concentration.23−26 For example, Salis et al. observe a charge reversal for bovine serum albumin (BSA) when the salt concentration is increased from 10 to 500 mM.23 The negative charge found below the pI is a logical consequence of charge reversal especially at higher salt concentrations. This is due to the electrostatic interactions from the positively charged NCBD pulling the counterions in as the salt concentration is increased until the point of neutralization. After this point nonelectrostatic interactions such as ionic dispersion take over increasing the local counterion concentration which gets physisorbed to the protein resulting in a net negative charge. In light of the sign reversal of the ζ-potential at high salt concentrations for NCBD, we complemented our previous binding measurements using the stopped-flow technique16 by performing experiments at a low salt concentration. Thus, binding experiments were done at three different ionic strength conditions at pH = 7.4 in sodium phosphate/NaCl buffers to investigate the role of salt in binding of ACTR: I = 0.023, 0.20, and 0.98 M, respectively. We were not able to obtain a precise value of the apparent association rate constant (konapp) at the lowest salt concentration due to elevated observed rate constants (kobs), but judging from kobs at low ACTR concentrations, it seems that the association rate constant konapp is about an order of magnitude higher compared to the value obtained at the more physiological ionic strength (∼30 × 107 vs 2.8 × 107 M−1 s−1). At the highest ionic strength, konapp changed only marginally (1.6 × 107 M−1 s−1) when compared 2452
dx.doi.org/10.1021/ac3035025 | Anal. Chem. 2013, 85, 2449−2456
Analytical Chemistry
Article
Figure 4. (a) Scatter plot of current change and time is shown for ACTR. The experiments were performed at an applied voltage of +300 mV (black) and +500 mV (red). Note that two clusters (cluster 1 and cluster 2) of translocation events were observed from both voltages. Histograms of the ionic current distribution (ΔIb) and translocation time (tD) of ACTR are also shown for +300 mV. A similar analysis is performed for (b) NCBD and (c) the ACTR/NCBD complex. In all cases Gaussian fits are shown as a red line.
polarization36 or (partial) separation of the protein from its solvating ion cloud upon entry into the pore, may result in current enhancement but are expected to be too short-lived or too small in intensity to account for the observations made here. Only processes that affect the steady-state transport of ions through the pore result in a large enough effect that is sufficiently long-lived. In accordance with this interpretation, we now propose a model that could explain both the long duration of the ACTR translocation events as well as the sign of the current modulation. Namely, we hypothesize that it is a combination of adsorption to the inner pore surface, which slows down the translocation process, and modulation of the inner surface charge density that leads to the observed effects. Note that the experimentally determined conductance of the pore corresponds to a cylindrical nanopore with a diameter of 16 nm, neglecting surface contributions and access resistance for simplicity. If the nanopore is noncylindrical, e.g., hourglassshaped, then the narrowest part is most likely smaller and the above value represents on upper limit. Accordingly, the surface contribution Gsurf to the total pore conductance G rather increases compared to the bulk contribution Gbulk. Equation 1 expresses the pore conductance G for a negatively charged cylindrical pore of high aspect ratio with a length (L) much greater than the diameter (d) immersed in a KCl containing electrolyte solution:37
times for ACTR are much longer than expected. Interactions between the translocating molecule and the inner pore surface, or even electroosmotic effects, have been used to rationalize long translocation times or even utilized to characterize the analyte.7,31,32 For example, Niedzwiecki et al. characterize the translocation and binding of negatively charged BSA to the walls of a SiNx nanopore.31 Furthermore, Rosso et al. have more recently published an article highlighting the adsorption of proteins to silicon-based surfaces.33 This is not entirely unreasonable to expect for IDPs as the lack of conformational homogeneity can potentially result in significantly more interaction between the nanopore wall and the protein than a perfectly ordered or folded protein (i.e., a much greater proportion of the surface area of an IDP is exposed to the nanopore walls). A further observation can be made in terms of the translocation process. Although all samples translocate from the cis to trans reservoirs, both NCBD and the complex exhibit current enhancement, while ACTR exhibits a current decrease. This in its own right warrants further discussion. Ionic Current Blockade or Enhancement? Somewhat surprising is that the sign of the current modulation is different for NCBD, the complex, and ACTR even though the ζpotentials are very close to 0 in the TB buffer (ACTR, −5.20 ± 0.384 mV; NCBD, −1.13 ± 0.416 mV; complex, −0.955 ± 0.356 mV). In a simple volume-exclusion model, current blockade is expected due to the exclusion of electrolyte solution by an (uncharged) analyte. On the other hand, current enhancement may be rationalized by a high charge of the translocating species,34,35 which temporarily increases the ion flux and hence the current through the pore upon translocation.21 However, the latter clearly does not apply to NCBD or the complex. Transient effects, such as concentration
G = G bulk + Gsurf =
π d2 d (μK + μCl )nKCle + πσμK 4 L L (1)
where μK and μCl are the electrophoretic mobilities of K+ and Cl−, nKCl is the number density of K+ or Cl− ions, σ is the 2453
dx.doi.org/10.1021/ac3035025 | Anal. Chem. 2013, 85, 2449−2456
Analytical Chemistry
Article
surface charge density, and e the elementary charge. Normalizing to Gbulk yields Grel =
4σμK G =1+ G bulk d(μK + μCl )nKCle
distinctive conformational states. A further possibility is that the orientation of the protein entering the pore could potentially produce a different modulation in the current. However, this is unlikely as the pore dimensions are much larger than the protein size or excluded volume. We can only speculate about the species giving rise to event c. It may be an intermediate structure due to the two current levels being identical to that of events a and b. Another possibility is that NCBD interconverts in the pore, from the event a species to that of event b. A third possibility is that one of the events results from voltagedependent denaturation of NCBD, which is consistent with the low stability of NCBD. Unfortunately, due to the relatively rare occurrence of these events, a complete statistical analysis based on our data is not possible and is outside the scope of this manuscript. Four structures of NCBD/target complexes have thus far been determined, showing that NCBD adopts two different conformations depending on the bound ligand.12,43,44,45 Further, NMR studies have shown that NCBD displays dynamics on a wide range of time scales;13 for instance, there is significant line broadening in the 1H−15N heteronuclear single-quantum coherence spectrum of NCBD, indicative of exchange between different states on the microsecond to millisecond time scale.13,15 It is clear that the conformation of the NCBD polypeptide affects its effective charge. We speculate that one of the translocation events results from a denaturation and the other two from differences in charge related to the ion cloud, which in turn depend on small differences in structure. It is known that highly charged proteins may be surrounded by ions such that the overall charge is neutral or even of opposite sign.23−26 IDPs have a higher number of charged groups than globular proteins, and electrostatics are believed to govern their binding reactions. Finally, translocation studies of the ACTR/NCBD complex were also performed. Much like NCBD, current enhancement was observed; however, in this case only one event type was observed consisting of a blockade duration of 0.2 ± 0.1 ms and a current amplitude of 0.7 ± 0.2 nA at an applied potential of +300 mV. Importantly, the blockade duration decreased linearly as a function of applied potential (Supporting Information Figure S2) which is a good indirect indication of translocation and not “bumping”. It is also worth noting that the width of the blockade distribution is relatively compact when compared to that of the NCBD. The calculated volume and hydrodynamic diameter of the complex are 12 nm3 and 3.6 nm, respectively, which is much smaller than the total pore volume; this ensures any surface interaction taking place would be expected to be uniform for all translocation events. Furthermore, only a single event type was observed which supports the fact that only a single conformation is being translocated through the nanopore and is in support of previously published NMR studies, which show that ACTR and NCBD obtain well-defined structures following their coupled folding and binding.12
(2)
Taking the pore diameter as d = 16 nm, μK = 7.616 × 10−8 m2/ (V·s), μCl = 7.909 × 10−8 m2/(V·s), nKCl = 6.022 × 1026 ions/ m3, and σ = 0.06 C/m2,38 Grel is increased by roughly 7.6%, relative to an uncharged pore, and further to 8.9% for σ = 0.07 C/m2. Accordingly, a decrease in σ causes a decrease in Grel (and thus the pore current). Hence, if adsorption of a biomolecule leads to an increase in the local surface charge density, the pore current is enhanced and the duration of the current modulation is related to the adsorption/desorption kinetics, rather than the transport process itself. This could imply that the shape of the current modulation could in principle be directly correlated to adsorption and desorption rather than the conformation of the protein itself. Assuming adsorption leads to an increase in local surface charge density, the conductance is enhanced by 1.3% during this process, which is similar to the conductance modulation observed for NCBD and the complex. Larger effects are predicted for smaller pore dimensions, and it should be noted that the diameter used above represents an upper limit, especially if the nanopore channel is noncylindrical. Hence, given the approximate nature of the model, its predictions are in reasonable agreement with experimental data and validate the possibility of obtaining either current blockade or enhancement based on the level of interaction between the nanopore, the protein, and local ionic environment. Further studies are required to quantify this effect. Folding and Binding of ACTR and NCBD. Scatter plots and histograms of event duration and change in current for ACTR, NCBD, and the complex are shown in Figure 4. For ACTR (Figure 4a), two clear current blockade populations were observed (cluster 1 and cluster 2). Cluster 1 had a ΔIb of 0.230 ± 0.004 nA and short blockade durations below 0.05 ms. This is likely due to the protein touching or entering the nanopore but not physically translocating.39,40 Cluster 2 on the other hand had a significantly higher ΔIb = 0.42 ± 0.08 nA and an average translocation time of 5.8 ± 1.3 ms. The excluded atomic volume of ACTR was calculated to be 28.7 nm3 using Λ(t) ≈ (ΔIbVp)/I0, where ΔIb is the amplitude of current blockade, I0 is the open pore current (17.8 nA), and Vp is the pore volume (1256 nm3). The excluded volume is in good agreement with the theoretical volume, 10 nm3, based on standard residue volumes in solution,41 indicating ACTR translocation in the unfolded form.42 A similar analysis was performed with NCBD, and in this case three different blockade events were observed (Figures 3b and 4b). First, a small fraction of “bumping” events (2%) were observed and excluded from further analysis. These events are likely due to proteins approaching the nanopore without translocating. Second, three distinct translocation types (a, b, and c) were observed. Event a had a mean amplitude of ΔIb = 1.0 ± 0.1 nA and consisted of 13.4% of all events, while event b had a mean amplitude of ΔIb = 0.6 ± 0.1 nA and contributed to 81.2% of all translocations. Finally, a third population was observed, event c, which consisted of 5.4% of all events. Assuming the surface interactions between these three event types and the wall of the nanopore are similar, these three modulations can be attributed to NCBD populating three
■
CONCLUSIONS Our translocation experiments suggest that the disordered NCBD populates distinct conformations in the free state, which was previously shown by NMR studies to occur in the bound state with different ligands, and very recently also in the free state.15,46 Moreover, the ζ-potential of NCBD in its free form goes from positive to negative at high salt concentrations around 200 mM KCl emphasizing the importance of salt conditions in modulating the electrostatic environment around 2454
dx.doi.org/10.1021/ac3035025 | Anal. Chem. 2013, 85, 2449−2456
Analytical Chemistry
Article
(9) Chen, H.; Lin, R. J.; Schiltz, R. L.; Chakravarti, D.; Nash, A.; Nagy, L.; Privalsky, M. L.; Nakatani, Y.; Evans, R. M. Cell 1997, 90 (3), 569−580. (10) Leo, C.; Chen, J. D. Gene 2000, 245 (1), 1−11. (11) Goodman, R. H.; Smolik, S. Genes Dev. 2000, 14 (13), 1553− 1577. (12) Demarest, S. J.; Martinez-Yamout, M.; Chung, J.; Chen, H.; Xu, W.; Dyson, H. J.; Evans, R. M.; Wright, P. E. Nature 2002, 415 (6871), 549−553. (13) Ebert, M. O.; Bae, S. H.; Dyson, H. J.; Wright, P. E. Biochemistry 2008, 47 (5), 1299−1308. (14) Keppel, T. R.; Howard, B. A.; Weis, D. D. Biochemistry 2011, 50 (40), 8722−8732. (15) Kjaergaard, M.; Teilum, K.; Poulsen, F. M. Proc. Natl. Acad. Sci. U.S.A. 2010, 107 (28), 12535−12540. (16) Dogan, J.; Schmidt, T.; Mu, X.; Engström, A.; Jemth, P. J. Biol. Chem. 2012, 287 (41), 34316−34324. (17) Chansin, G. A.; Mulero, R.; Hong, J.; Kim, M. J.; DeMello, A. J.; Edel, J. B. Nano Lett. 2007, 7 (9), 2901−2906. (18) Kim, Y. R.; Min, J.; Lee, I. H.; Kim, S.; Kim, A. G.; Kim, K.; Namkoong, K.; Ko, C. Biosens. Bioelectron. 2007, 22 (12), 2926−2931. (19) Ayub, M.; Ivanov, A.; Hong, J.; Kuhn, P.; Instuli, E.; Edel, J. B.; Albrecht, T. J. Phys.: Condens. Matter 2010, 22 (45), 454128. (20) Chou, T. J. Chem. Phys. 2009, 131 (3), 034703. (21) Smeets, R. M.; Keyser, U. F.; Krapf, D.; Wu, M. Y.; Dekker, N. H.; Dekker, C. Nano Lett. 2006, 6 (1), 89−95. (22) He, Y.; Gillespie, D.; Boda, D.; Vlassiouk, I.; Eisenberg, R. S.; Siwy, Z. S. J. Am. Chem. Soc. 2009, 131 (14), 5194−5202. (23) Salis, A.; Boström, M.; Medda, L.; Cugia, F.; Barse, B.; Parsons, D. F.; Ninham, B. W.; Monduzzi, M. Langmuir 2011, 27 (18), 11597− 11604. (24) Boström, M.; Parsons, D. F.; Salis, A.; Ninham, B. W.; Monduzzi, M. Langmuir 2011, 27 (15), 9504−9511. (25) Lee, K. K.; Fitch, C. A.; Lecomte, J. T. J.; Garcia-Moreno, B. Biochemistry 2002, 41 (17), 5656−5667. (26) Parsons, D. F.; Boström, M.; Maceina, T. J.; Salis, A.; Ninham, B. W. Langmuir 2010, 26 (5), 3323−3328. (27) Ganguly, D.; Otieno, S.; Waddell, B.; Iconaru, L.; Kriwacki, R. W.; Chen, J. H. J. Mol. Biol. 2012, 422 (5), 674−684. (28) Smeets, R. M.; Kowalczyk, S. W.; Hall, A. R.; Dekker, N. H.; Dekker, C. Nano Lett. 2009, 9 (9), 3089−3096. (29) Cressiot, B.; Oukhaled, A.; Patriarche, G.; Pastoriza-Gallego, M.; Betton, J. M.; Auvray, L.; Muthukumar, M.; Bacri, L.; Pelta, J. ACS Nano 2012, 6 (7), 6236−6243. (30) Stefureac, R. I.; Trivedi, D.; Marziali, A.; Lee, J. S. J. Phys.: Condens. Matter 2010, 22 (45), 454133. (31) Niedzwiecki, D. J.; Grazul, J.; Movileanu, L. J. Am. Chem. Soc. 2010, 132 (31), 10816−10822. (32) Wanunu, M.; Sutin, J.; McNally, B.; Chow, A.; Meller, A. Biophys. J. 2008, 95 (10), 4716−4725. (33) Rosso, M.; Nguyen, A. T.; de Jong, E.; Baggerman, J.; Paulusse, J. M. J.; Giesbers, M.; Fokkink, R. G.; Norde, W.; Schroen, K.; van Rijn, C. J. M.; Zuilhof, H. ACS Appl. Mater. Interfaces 2011, 3 (3), 697−704. (34) Chang, H.; Kosari, F.; Andreadakis, G.; Alam, M. A.; Vasmatzis, G.; Bashir, R. Nano Lett. 2004, 4 (8), 1551−1556. (35) Fan, R.; Karnik, R.; Yue, M.; Li, D.; Majumdar, A.; Yang, P. Nano Lett. 2005, 5 (9), 1633−1637. (36) Das, S.; Dubsky, P.; van den Berg, A.; Eijkel, J. C. Phys. Rev. Lett. 2012, 108 (13), 138101. (37) Ayub, M.; Ivanov, A.; Instuli, E.; Cecchini, M.; Chansin, G.; McGilvery, C.; Hong, J. G.; Baldwin, G.; McComb, D.; Edel, J. B.; Albrecht, T. Electrochim. Acta 2010, 55 (27), 8237−8243. (38) Keyser, U. F.; Krapf, D.; Koeleman, B. N.; Smeets, R. M.; Dekker, N. H.; Dekker, C. Nano Lett. 2005, 5 (11), 2253−2256. (39) Kasianowicz, J. J.; Brandin, E.; Branton, D.; Deamer, D. W. Proc. Natl. Acad. Sci. U.S.A. 1996, 93 (24), 13770−13773. (40) Fologea, D.; Ledden, B.; McNabb, D. S.; Li, J. Appl. Phys. Lett. 2007, 91 (5), 539011−539013.
the protein. This is also reflected in the kinetics of the interaction, which we measured using the stopped-flow technique, as a function of salt concentration. Thus, the physiological salt concentration allows highly uniformly charged IDPs, such as NCBD, to interact with their targets very specifically but with reduced affinity compared to lower salt conditions. This effect could thus optimize the affinity of signaling complexes, in particular since the charge reversal occurs just above physiological ionic strength. Any local intracellular change in ionic strength or pH could therefore have a significant effect on the binding properties of NCBD. At the same time IDPs may still retain the disordered nature since they are enriched in polar and charged residues and depleted of bulky hydrophobic residues, making it difficult to form a stable core, which in turn provides a way to have multiple binding partners. In general, the number of charged residues of IDPs are higher than those of globular proteins. Our results highlight the importance of including the modulation of electrostatics of uniformly highly charged IDPs by salt, when discussing or simulating interactions involving such IDPs. Furthermore, we show that the nanopore studies performed are useful in validating that charge reversal takes place at high salt concentrations, and at the same time this method is also useful in extracting conformational populations.
■
ASSOCIATED CONTENT
S Supporting Information *
Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported in part by a European Research Council Starting Grant (J.B.E.), the Swedish Research Centre (P.J.), The Lars Hierta Memorial Foundation (J.D.), Biomedical Research Council (T.A., J.B.E., D.J.), and a HFSP Research Grant (J.B.E., P.J., M.J.K.). We are grateful to Drew Parsons and Mathias Boström for sharing their expertise on charge reversal of proteins, and to Åke Engströ m for performing mass spectrometry analyses.
■
REFERENCES
(1) Oldfield, C. J.; Cheng, Y.; Cortese, M. S.; Brown, C. J.; Uversky, V. N.; Dunker, A. K. Biochemistry 2005, 44 (6), 1989−2000. (2) Uversky, V. N.; Dunker, A. K. Anal. Chem. 2012, 84 (5), 2096− 2104. (3) Oldfield, C. J.; Meng, J.; Yang, J. Y.; Yang, M. Q.; Uversky, V. N.; Dunker, A. K. BMC Genomics 2008, 9 (Suppl.1), S1. (4) Bahrami, A.; Dogan, F.; Japrung, D.; Albrecht, T. Biochem. Soc. Trans. 2012, 40 (4), 624−628. (5) Miles, B. N.; Ivanov, A. P.; Wilson, K.; Dogan, F.; Japrung, D.; Edel, J. B. Chem. Soc. Rev. 2013, 42 (1), 15−28. (6) Talaga, D. S.; Li, J. J. Am. Chem. Soc. 2009, 131 (26), 9287−9297. (7) Firnkes, M.; Pedone, D.; Knezevic, J.; Doblinger, M.; Rant, U. Nano Lett. 2010, 10 (6), 2162−2167. (8) Freedman, K. J.; Jürgens, M.; Prabhu, A.; Ahn, C. W.; Jemth, P.; Edel, J. B.; Kim, M. J. Anal. Chem. 2011, 83 (13), 5137−5144. 2455
dx.doi.org/10.1021/ac3035025 | Anal. Chem. 2013, 85, 2449−2456
Analytical Chemistry
Article
(41) Harpaz, Y.; Gerstein, M.; Chothia, C. Structure 1994, 2 (7), 641−649. (42) Counterman, A. E.; Clemmer, D. E. J. Am. Chem. Soc. 1999, 121 (16), 4031−4039. (43) Qin, B. Y.; Liu, C.; Srinath, H.; Lam, S. S.; Correia, J. J.; Derynck, R.; Lin, K. Structure 2005, 13 (9), 1269−1277. (44) Lee, C. W.; Martinez-Yamout, M. A.; Dyson, H. J.; Wright, P. E. Biochemistry 2010, 49 (46), 9964−9971. (45) Waters, L.; Yue, B.; Veverka, V.; Renshaw, P.; Bramham, J.; Matsuda, S.; Frenkiel, T.; Kelly, G.; Muskett, F.; Carr, M.; Heery, D. M. J. Biol. Chem. 2006, 281 (21), 14787−14795. (46) Kjaergaard, M.; Andersen, L.; Dalby Nielsen, L.; Teilum, K. Biochemistry 2013, in press, DOI: 10.1021/bi4001062.
2456
dx.doi.org/10.1021/ac3035025 | Anal. Chem. 2013, 85, 2449−2456