Electrochemical sensing of tumor suppressor protein p53–deoxyribonucleic acid complex stability at an electrified interface

Analytica Chimica Acta 828 (2014) 1–8

Contents lists available at ScienceDirect

Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca

Electrochemical sensing of tumor suppressor protein p53– deoxyribonucleic acid complex stability at an electrified interface  Emil Pale9 cek *, Hana Cernocká, Veronika Ostatná, Lucie Navrátilová, Marie Brázdová Institute of Biophysics, Academy of Sciences of the Czech Republic, v.v.i., Královopolská 135, 612 65 Brno, Czech Republic

H I G H L I G H T S

G R A P H I C A L A B S T R A C T


Electrochemical detection of DNA binding to oncoprotein p53 at picomole level.
Disintegration of surface-attached DNA–p53 complex at negative potentials.
Discrimination between sequencespecific and non-specific DNA–p53 protein binding.

A R T I C L E I N F O

A B S T R A C T

Article history: Received 21 February 2014 Received in revised form 21 March 2014 Accepted 24 March 2014 Available online 25 March 2014

Electrochemical biosensors have the unique ability to convert biological events directly into electrical signals suitable for parallel analysis. Here we utilize specific properties of constant current chronopotentiometric stripping (CPS) in the analysis of protein and DNA–protein complex nanolayers. Rapid potential changes at high negative current intensities (Istr) in CPS are utilized in the analysis of DNA–protein interactions at thiol-modified mercury electrodes. P53 core domain (p53CD) sequencespecific binding to DNA results in a striking decrease in the electrocatalytic signal of free p53. This decrease is related to changes in the accessibility of the electroactive amino acid residues in the p53CD– DNA complex. By adjusting Istr and temperature, weaker non-specific binding can be eliminated or distinguished from the sequence-specific binding. The method also reflects differences in the stabilities of different sequence-specific complexes, including those containing spacers between half-sites of the DNA consensus sequence. The high resolving power of this method is based on the disintegration of the p53CD–DNA complex by the electric field effects at a negatively charged surface and fine adjustment of the millisecond time intervals for which the complex is exposed to these effects. Picomole amounts of p53 proteins and DNA were used for the analysis at full electrode coverage but we show that even 10–20fold smaller amounts can be analyzed. Our method cannot however take advantage of very low detection limits of the protein CPS detection because low Istr intensities are deleterious to the p53CD–DNA complex stability at the electrode surface. These data highlight the utility of developing biosensors offering novel approaches for studying real-time macromolecular protein dynamics. ã 2014 Elsevier B.V. All rights reserved.

Keywords: Deoxyribonucleic acid–protein binding Tumor suppressor protein p53 Electrochemical sensing Constant current chronopotentiometry Mercury containing electrodes

1. Introduction

* Corresponding author. Tel.: +420 549 246 241; fax: +420 541 517 249. E-mail address: [email protected] (E. Pale9cek). http://dx.doi.org/10.1016/j.aca.2014.03.029 0003-2670/ ã 2014 Elsevier B.V. All rights reserved.

DNA–protein binding plays a central role in many molecular processes within organisms and cells, such as replication, transcription, DNA repair and packaging, etc. Research into the

2

E. Pale9cek et al. / Analytica Chimica Acta 828 (2014) 1–8

processes of DNA–protein binding and the nature and properties of the complexes formed between DNA and proteins is therefore of utmost importance. Various methods have been used to investigate DNA–protein interactions, including X-ray crystal analysis, NMR and fluorescence anisotropy [1–3]. Methods of electrochemical analysis have only in recent years been applied in DNA–protein interaction studies [4–6], dealing predominantly with aptamers (reviewed in Refs. [7,8]) and mostly using labeled DNA. Recently label-free methods for the detection of DNA– protein binding were introduced based on changes in the electrode charge transfer and capacity as detected by electrochemical impedance spectra [9]. Both DNA [8,10] and proteins [11,12] are electrochemically active thus making possible their label-free electrochemical determination. Such DNA determination is based on DNA oxidation or reduction signals at carbon or mercury electrodes, respectively (reviewed in Ref. [8]). Label-free determination of proteins involves either signals of irreversible oxidation at tyrosine and tryptophan residues at carbon electrodes [13,14] or reversible electrochemistry of the non-protein components of a small group of conjugated proteins (e.g., some metallo-proteins) [15]. Highly sophisticated methods for the analysis of DNA binding to [4Fe–4S] cluster-containing proteins have been developed [16,17] based on DNA charge transfer studies [16]. Recently, it has been shown that using constant current chronopotentiometric stripping (CPS) practically all proteins, including poorly soluble membrane proteins [18], produce an electrocatalytic peak H at mercury and solid amalgam electrodes [12,19–22], allowing protein determination at a much higher sensitivity (down to nanomolar and subnanomolar concentrations [12]) than the above electrochemical/voltammetric methods. At high Istr intensities the peak H is sensitive to changes in the protein structures [12], including denaturation [23,24], and aggregation [25], as well as changes resulting from mutations (single amino acid exchange) [26] or changes in the redox state [27]. The peak H was applied to detect the binding of MutS protein to DNA containing single base mismatches [5,6]. Double stranded (ds) DNA attached to magnetic beads, interacted with MutS in solution and picograms of the DNA-bound protein were detected after dissociation of the MutS from its complex with the DNA, using the CPS peak H. This technique was also applied in reverse, using protein attached to magnetic beads to detect the binding of DNA voltammetrically [28]. The tumor suppressor protein p53 is a transcription factor involved in cell defense against malignant transformation and genomic instability [29]. This protein binds DNA by several modes [30,31], and the sequence-specific binding of the p53 core domain to the dsDNA consensus site (DNACON, involved in cell cycle arrest and apoptosis) has been most intensively studied [32,33]. The consensus sequence contains two copies of the halfsite 10-base pair motif RRRC(A/T)|(T/A)GYYY separated by up to 13 bp (R represents purines while Y stands for pyrimidines in this sequence; the center of symmetry within the half site is indicated by the vertical bar). Four p53 molecules were seen to bind to the consensus sequence and the crystal structure showed that the p53 core domain could bind to one 5 bp quarter site [32]. Binding of four p53 units caused bending and twisting of the DNA [34]. Here we show that similarly to free p53 wt and mutant proteins [26], the p53–DNA sequence-specific complexes form nanolayers at thiol-modified electrodes. CPS signals of these complexes however greatly differ from those of free p53 proteins. By controlling the temperature and/or current density, it is possible to discriminate between the stabilities of different p53CD–DNA complexes on the basis of their susceptibility to the electric field effects at the electrode interface.

2. Material and methods 2.1. Chemicals Human wild-type p53 core domain (amino acids 94–312, p53CD) and mutant p53CD-R273H were expressed from the vector pGEX-4TGST-p53CD [35] and pGEX-4TGST-p53CD-R273H [36], respectively as glutathione S-transferase (GST) fusion proteins. The proteins were expressed and purified according to a previously described protocol [36]. Final purification was achieved by preparative size-exclusion chromatography on a Superdex 200 HiLoad 26/60 column (Amersham Pharmacia Biotech) in 25 mM Hepes (pH 7.6), 200 mM KCl, 10% glycerol, 1 mM dithiothreitol (DTT) and 1 mM benzamidine. Most of the experiments were performed with the fusion p53–GST core domain. The results obtained with p53CD from which GST was removed did not significantly differ from those obtained with p53CD (see Fig. SI-3b for more details). The GST free form of p53CD was prepared by Thrombin cleavage and using GSTrapTM HP column (GE Healthcare) for GST capture. The final step of purification was dialysis to storage buffer (25 mM Hepes pH 7.6, 200 mM KCl, 10% glycerol, 1 mM DTT, 1 mM benzamidine). The protein concentration was determined spectrophotometrically using the molar extinction coefficient e280 = 60 280 M1 cm1 obtained from the ProtParam software on the EXPASY server. The following oligodeoxynucleotides synthesized by VBC Biotech (Vienna, Austria) were used: CON 50 -CGGCGATAAGAGACATGCCTAGACATGCCTCTTGATACGC30 (82.4  C) NON 50 - TATCGATATGGCCTCATAGCGCATCATAGCCTTGATATCG30 (79.7  C) CONGC 50 -CGGCGATAAGAGACATGCCTGCAGACATGCCTCTTGATACGC-30 (86.2  C) CONAT 50 -CGGCGATAAGAGACATGCCTATAGACATGCCTCTTGATACGC-30 (81.5  C) CONATAT 50 -CGGCGATAAGAGACATGCCTATATAGACATGCCTCTTGATACGC-30 (80.7  C) and (dA)50 (71.3  C) Tm values shown in brackets refer to DNA duplexes with their complementary strands. These values were calculated for 2 mM oligodeoxynucleotides in 50 mM NaCl according to Sigma-Aldrich OligoEvaluatorTM. The p53 consensus sequence is underlined. Bovine serum albumin and other chemicals of analytical grade were from Sigma–Aldrich. Solutions were prepared from triple distilled water. 2.2. Apparatus Electrochemical measurements were performed using an AUTOLAB Analyzer (EcoChemie, Utrecht, The Netherlands) in combination with VA-Stand 663 (Metrohm, Herisau, Switzerland); a hanging mercury drop electrode (HMDE, 0.4 mm2) or meniscus silver solid amalgam electrode (0.5 mm2) served as the working electrode in a standard cell in a three-electrode system. An Ag| AgCl|3 M KCl electrode was used as the reference electrode and a platinum wire as the auxiliary electrode. Experiments were carried out in open air, at a controlled temperature of 21  C unless stated otherwise. 2.3. DNA binding of p53 protein In general 2 mM p53CD with 0.5 mM DNA was incubated in 5 mM Tris–HCl pH 7.6, 0.5 mM EDTA, 50 mM NaCl and 10 mM DTT in a volume of 40 mL on ice for 30 min. For electrochemical measurements an aliquot of 5 mL was withdrawn from the mixture after incubation.

E. Pale9 cek et al. / Analytica Chimica Acta 828 (2014) 1–8

3

2.4. Electrochemical measurements The descriptive derived parameter used in this study is “(dE/ dt)1 vs. E”. It is related to chronopotentiometry, in which a suitable constant current density is passed through the working electrode and the ensuing variation of the interfacial potential change is measured in time. The resulting E–t curve is described by the Sands equation [37]. Derivative of this E–t curve yields the parameter dE/dt and its inverse can be plotted as function j = (dE/ dt)1 against potential E. The main reason for this approach is the fact that the stability of the surface-attached p53–DNA complexes is strongly influenced by short exposure to the electric field in the electrode double layer. Chronopotentiometry has been used in previous publications dealing with proteins (e.g. [12,20,21,26]). Here CPS analysis is used in an adsorptive transfer (ex situ) mode [12]: after the incubation (see above), 5 mL aliquot of incubation mixture with DTT is coadsorbed on the working electrode at open circuit potential for an accumulation time, tA of 60 s. The coadsorption of DTT and protein on HMDE was described [21]. The protein–DTT modified electrode is washed and transferred to the thermostated electrolytic cell with blank background electrolyte, 50 mM Na-phosphate, pH 7, and measured. When the effect of the electrode potential on the surface-attached p53–DNA complex is measured for longer time periods (e.g. 1 s or more, Figs. SI-4 and SI-5) the electrode is exposed (in the background electrolyte) to the potential EB for time tB, followed by CPS analysis. 3. Results Earlier we showed that under certain conditions free p53 wt and mutant protein core domains formed nanolayers at dithiothreitol (DTT)-modified hanging mercury drop electrode (HMDE) and/or solid amalgam electrode (SAE) [26]. Such protein layers catalyzed a hydrogen evolution reaction and produced CPS peaks, which reflected changes in mutant protein structures resulting from single amino acid exchange. This formed a novel approach to study dynamic changes in protein folding states using nanolayers at electrical interfaces. Here we used the same arrangement to learn how the sequence-specific and non-specific DNA binding will influence the CPS responses of free p53CD proteins. 2 mM p53CD with 10 mM DTT were coadsorbed at a HMDE [21] at open circuit potential for an accumulation time, tA of 60 s. The protein-modified DTT–HMDE was transferred into an electrolytic cell containing 50 mM sodium phosphate, pH 7.0, to perform CPS analysis at 21  C using stripping current, Istr of 35 mA. Under these conditions the p53CD yielded a well-developed peak H1 (peak potential, Ep 1.92 V) and a small peak H2 (Ep 1.97 V) (Fig. 1a). After incubation of p53CD with the consensus sequence-containing dsDNA 40-mer (DNACON) the peak H1 almost disappeared (Fig. 1a). This dramatic change corresponded to the p53 sequence-specific binding, as detected by gel electrophoresis (not shown). In contrast, incubation with the 40-mer dsDNA not containing consensus sequence (DNANON) resulted in only a slight decrease and small shift of both the peak H1 and H2 to more negative potentials (Fig.1a). Other DNAs without the consensus sequence displayed similar responses as the dsDNANON. Similarly, interaction of bovine serum albumin with the DNACON did not result in any significant change in the bovine serum albumin peak H (Fig. SI-1). Large differences between the peak H1 heights of the free p53CD and p53CD–DNACON were observed at different concentrations of these substances (Fig. SI-2a). The titration curves were in agreement with the binding of roughly four p53CD to one DNA molecule and the abundance of DNA did not interfere with the analysis (Fig. SI-2b). It was previously shown that the removal of zinc from the p53 core domain by treatment with EDTA resulted in the absence of the DNA sequence-specific binding [26,38]. Using EDTA we removed

Fig. 1. (a) Sequence-specific binding of wild type p53 core domain (p53CD) to dsDNACON as detected by CPS at a DTT-modified hanging mercury drop electrode (DTT–HMDE). Free p53CD (black, solid line), sequence-specific p53CD–DNACON complex (red, dash-dotted line), mixture of p53CD with 40-mer dsDNA not containing the consensus sequence (blue, dotted line, p53CD + dsDNANON) and dsDNACON alone (cyan, dashed line), background electrolyte, 50 mM Na-phosphate, pH 7, (dashed line). (b, c) Interaction of mutant p53CD R273H with dsDNACON (showing no DNA binding). Free p53CD R273H (black, solid line), p53CD R273H + dsDNACON (red, dash-dotted line), p53CD R273H + dsDNANON (blue, dotted line). HMDE was immersed in a 5 mL drop of solution containing 2 mM p53CD or p53–DNA complex with 10 mM DTT to adsorb on HMDE for tA of 60 s at open circuit potential. DTT–HMDE with the surface-attached protein or protein–DNA complex was then washed and transferred into the blank background electrolyte to perform the CPS analysis at stripping current, Istr (a, b) 35 mA or (c) 55 mA at 21  C. (For

the Zn ion from p53CD and after incubation with the DNACON we observed only small changes in the peak H (Fig. SI-3a), resembling those detected after p53CD incubation with the DNANON (Fig. 1a). Similarly, the mutant protein R273H [39] (which does not bind DNA) did not show any sign of DNA sequence-specific binding (Fig. 1b) even at Istr 55 mA (Fig. 1c). The large decrease in the peak H1 of the p53CD–DNACON (as compared to the free p53CD, Fig. 1a) can be tentatively explained by the greatly decreased accessibility of catalytically active amino acid residues (Lys, Arg, His and Cys) [12,40,41] in the p53CD–DNACON complex, and by other factors related to the attachment of the complex to the electrified interface (see below). The above results were obtained at high negative Istr intensities (between 30 and 60 mA). Using lower intensities or the usual voltammetric methods (see below) discrimination between free p53CD and p53CD–DNA complexes was either difficult or impossible. These results suggested that the surfaceattached p53CD–DNA complex may disintegrate as a result of prolonged exposure to negative potentials. On the other hand shortening of this exposure to milliseconds (e.g., 0.3–4.3 V/ms at Istr 45 mA) protected the complex from disintegration. Rapid potential changes in CPS were shown to save proteins from denaturation [42]. 3.1. Exposure to negative potentials To find out in which potential region the destabilization and disruption of the p53CD–DNACON complex layer may take place, we exposed the surface-attached complex to different potentials, EB for a time, tB of 1 s, followed by CPS at Istr 30 mA. Exposure to EB between 0.1 and 0.5 V had almost no effect on the CPS response of the p53CD–DNA complex (Fig. 2a), suggesting that the surface disintegration of the complex did not occur as a result of 1 s exposure to potentials close to the potential of zero charge, i.e., under the conditions in which the HMDE was modified by the impermeable chemisorbed DTT self-assembled monolayer. 1 s exposure to EB 0.8 or 1.2 V, (when the DTT Hg S bonds were already reduced) caused some disturbance of the complex structure, but not the full disintegration of the complex (Fig. 2b, c) while longer exposure times, such as 6–10 s resulted in much greater p53CD–DNACON damage (Fig. SI-4c–e). On the other hand 1 s exposure to EB 1.4 or 1.6 V evidently resulted in

4

E. Pale9cek et al. / Analytica Chimica Acta 828 (2014) 1–8

Fig. 2. Peaks H of p53CD (black, solid line) and p53CD–DNACON complex (red, dashdotted line) after 1 s waiting at potential EB of (a) 0.5 V, (b) 0.8 V, (c) 1.2 V, (d) 1.4 V, (e) 1.6 V and (f) 1.8 V. Protein–DTT-modified electrode was washed and transferred into the blank background electrolyte, kept for tB 1 s at the indicated EB followed by chronopotentiogram recording using Istr 30 mA. Other details in Fig. 1. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

disintegration of the complex as demonstrated by almost the same responses as those of p53CD and its DNA complex (Fig. 2d, e). At EB 1.8 V partial desorption of the p53CD–DNACON and p53CD molecules from the electrode surface took place (Fig. 2f). The above changes could play little role however in the p53CD–DNACON disintegration in CPS experiments (Fig. 1), as the exposure in CPS to negative potentials was much shorter than 1 s (at Istr 35 mA about 35 ms). It thus appears probable that the inability of CPS to recognize the p53CD–DNACON complex at Istr less negative than 25 mA (see below) was due to overly long exposure of the complex to negative potentials. 3.2. Dependence on temperature Previously we demonstrated that the CPS responses of wild type and mutant R175H of superstable p53CD [1] were strongly dependent upon the temperature at which the CPS was performed [26]. Here we measured the p53CD–DNACON complex and free p53CD at temperatures between 10 and 33  C at Istr 35 mA. Between 10 and 24  C the peak H heights of the p53CD–DNACON were close to zero (Figs. 3a–e and SI-5a). Between 27 and 30  C an abrupt increase in peak H1 of the p53CD–DNACON complex took place (Figs. 3f, g and SI-5d–h) suggesting changes in the protein dynamics, followed by disruption of the surface p53CD–DNACON complex layer at temperatures above 27  C. A detailed study of the effect of temperature between 27 and 30  C (Fig. SI-5d–h) is contained in Supplementary Information. The behavior of the p53CD protein and p53CD–DNA complex at constant Istr and varying temperatures (Figs. 3 and SI-5a, d–h) as well as the effects of Istr intensity on the p53CD–DNA responses at 45  C (Fig. SI-5b, c) suggest that temperature-induced changes in the protein structure dynamics significantly influence the susceptibility of the surfaceattached p53CD layer to the electric field effects. 3.3. Dependence on Istr intensities We looked at the dependence of the p53CD–DNACON signals on the Istr as compared to the free p53CD protein at a constant

Fig. 3. Dependence of peak H of free p53CD (black, solid line) and p53CD–DNACON complex (red, dash-dotted line) on the electrolyte temperature. (a) 10  C, (b) 13  C, (c) 16  C, (d) 19  C, (e) 24  C, (f) 27  C, (g) 30  C and (h) 33  C; Istr of 35 mA. Other details as in Fig. 1. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

temperature (21  C). Between Istr 35 and 60 mA (Fig. 4a, b) the peak H1 heights and Ep of the free p53CD decreased with increasing negative Istr intensities, while the peak heights of the p53CD–DNACON were close to zero and difficult to accurately measure (Fig. 4). In this Istr region very large differences between the CPS peaks of the p53CD–DNACON and free p53CD allowed detection of DNACON binding to p53CD as shown in Fig. 1. At Istr 20 mA, the peaks of the p53CD–DNACON and free p53CD were almost identical, indicating disintegration of the p53CD–DNACON layer under these conditions (Fig. 4c). In agreement with these results, cyclic voltammograms obtained at relatively low scan rates (from 0.5 to 2 V/s) displayed reduction peaks of the p53CD–DNACON and free p53CD differing little from each other (Fig. SI-7). The dependence of the peak H of p53CD–DNACON on Istr showed a much steeper decrease in this peak with increasing negative Istr intensities than that of the free p53CD (Fig. 4b). This decrease resembled curves, indicating a structural transition in the proteins [26]. In the same Istr region, Ep of the peak H1 of the complex shifted abruptly to more negative values (Fig. 4a), suggesting a more difficult reduction process. At Istr 25 mA the peak H1 of the p53CD–DNACON was absent but additional peaks appeared (Fig. 4d) at more negative potentials, probably as a result of changes in the accessibility of some amino acid residues in the p53CD. At Istr 40 mA practically no peaks were observed in the p53CD–DNACON (Fig. 4e). It can be concluded that using the described method, strong sequence-specific p53CD binding to DNA can be detected at a wide range of Istr values, when performing CPS close to room temperature. Furthermore we studied the weaker sequencespecific, as well as non-specific DNA binding of p53CD. 3.4. Non-specific binding of p53CD to DNA p53 core domain also binds to single stranded DNA [43,44] and dsDNA [43,45] which does not contain consensus sequence. This non-specific binding is, however, much weaker than the sequencespecific binding. Fig. 1a shows that at Istr 35 mA, the non-specific binding was not recognized and did not interfere with the detection of the sequence- specific binding. Considering the above results with p53CD–DNACON (Figs. 1 and 4c–e) we were interested in whether detection of the non-specific DNA binding would be possible at higher negative Istr intensities. We measured the dependence of the non-specific p53CD–DNANON layer CPS

E. Pale9 cek et al. / Analytica Chimica Acta 828 (2014) 1–8

5

Fig. 4. Comparison of p53CD sequence-specific and non-specific DNA binding. Dependence of (a) peak H1 potential, Ep and (b) peak H1 height, j of free p53CD (black, solid line), p53CD–DNACON complex (red, dashed line) and p53CD–DNANON (blue, dotted line) on stripping current (Istr). (c–e) Peak H of p53CD (black, solid line), p53CD–DNACON complex (red, dashed line) and p53CD–DNANON (blue, dotted line) at Istr of (c) 20 mA, (d) 25 mA, (e) 40 mA and (f) 50 mA. Other details as in Fig. 1. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

responses on Istr. We found that at high negative Istr intensities between 50 and 55 mA this complex became detectable by CPS (Fig. 4b). At Istr 50 mA p53CD–DNANON interaction already resulted in a striking decrease in the peak H1 (Ep 1.99 V) accompanied by a more negative peak H2 (Ep 2.06 V) (Fig. 4f). With increasing Istr intensities the peak H1 decreased more steeply than the peak H2 suggesting that the former peak is more sensitive to potential-induced changes in the conformation of the surfaceattached p53CD–DNA complex. At Istr 55 mA both peaks of the p53CD–DNANON greatly decreased but remained higher than the peak H of the sequence-specific complex p53CD–DNACON (Fig. 4b, inset). At Istr 60 mA the p53CD–DNANON yielded a single peak H, which was still slightly higher than that of the sequence-specific p53CD–DNACON complex. Fig. 4e, f shows striking changes in the peak H1 of the p53CD–DNANON at Istr between 40 and 55 mA. Using single-stranded DNACON similar results as with the p53CD– DNANON were obtained (not shown). Our results (Figs. 1 and 4) show that differences in the stabilities of sequence-specific and non-specific p53CD–DNA complexes can be recognized by CPS (Fig. 4) and that, at appropriate Istr intensities, the weaker p53CD– DNANON complex can either be destroyed at the electrode without disturbing the more stable p53CD–DNACON (Fig. 1) or (at higher negative Istr intensities) the formation of this weaker complex can be detected.

half-sites would be manifested in the CPS profiles of the p53CD– DNA complexes. We used three sequences corresponding to DNACON containing AT (CON-AT), GC (CON-GC) and ATAT (CON-ATAT) spacers (see Section 2) and measured the resulting p53CD–DNA complexes by CPS at different Istr. Using higher negative Istr intensity than that used in most of our experiments (Figs. 1–3) with p53CD–DNACON complexes (without any spacer in DNACON), the three spacer-containing p53CD–DNA complexes produced peak H greatly differing both from those of p53CD–DNACON and from free p53CD (Fig. 5a). These results suggested that at faster potential changes the spacer-containing complexes were less stable than p53CD–DNACON, but not completely disintegrated at the electrode surface. Under optimum Istr conditions for the detection of p53CD binding to DNACON (where the p53CD–DNA complex yielded almost no peak H), p53CD–DNA complexes with spacer-containing DNACON sequences yielded peak H significantly higher than the peak H of p53CD–DNACON (Fig. 5b). Under these conditions peak H of p53CD–DNACON–ATAT complex differed only little from the peak of free p53CD and peak H of p53CD–DNACON–GC was only slightly lower and less negative suggesting that this complex was not very

3.5. The effect of spacers interspersed between the half-sites It has been shown that most of the DNA sequences involved in cell cycle arrest bind p53 with a high affinity and do not have spacers interspersed between the two half sites [46,47]. In contrast the DNA sequences involved in apoptosis show large variations in p53 binding affinity and contain interspersed spacers between the two half-sites [46,47]. These results suggest that the DNA sequence and structure might modulate the affinity of p53 for its binding sites [48,49]. Recent high resolution crystal structures of sequencespecific p53CD–DNA complexes using DNA consisting of contiguous half-sites, showed non-canonical Hoogsteen base pairing and displayed enhanced protein–DNA and protein–protein interactions as compared to non-contiguous half-sites [49]. We were interested in whether the presence or absence of spacers between the two

Fig. 5. Peak H of p53CD (black) and p53CD complexes without (DNACON, red) and with spacer-containing dsDNAs: DNACON–GC (cyan), DNACON–AT (magenta) and DNACON–ATAT (blue) at Istr of (a) 38 mA, (b) 35 mA (c) 33 mA at 21  C. Other details as in Fig. 1. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

6

E. Pale9cek et al. / Analytica Chimica Acta 828 (2014) 1–8

far from complete disintegration at the electrode surface. On the other hand p53CD–DNACON–AT displayed much smaller and more negative peaks, suggesting better stability of this complex as compared to the other spacer-containing p53CD–DNA complexes. At Istr 33 mA all spacer containing p53CD–DNA complexes yielded peaks almost identical with that of free p53CD (Fig. 5c) suggesting their complete disintegration at electrode surface. Under these conditions the p53CD–DNACON complex produced two peaks more negative and smaller than peak H of free p53CD, suggesting that at slower potential changes this complex was disturbed but not disintegrated at the electrode surface. 4. Discussion Emerging concepts in the protein science field include development of novel physical interfaces that reveal nanoparticle assembly and disassembly. The combination of mercury electrodes with CPS forms a novel solid phase that includes the added advantage of integrating an electrical current that impacts on nanolayers or macromolecular dynamics. This paper shows that the surface layer of the sequence-specific p53CD–DNA complex is disintegrated at a negatively charged DTT–HMDE electrode (Fig. 1). This disintegration is potential- and time-dependent (Scheme 1). The electrocatalytic signals (peaks H) appear at highly negative potentials but in CPS we are able to control the time for which the complex is exposed to such potentials. By applying high negative current intensities to a relatively small working electrode, this exposure can be reduced to milliseconds (Scheme 1 and Fig. SI-6). Under such conditions the p53CD–DNACON complex yields only a very small, or no protein signal, contrasting with the large signal of free p53CD (Fig. 1). The decrease in the peak H of the complex can be explained by changes in the accessibility of electrocatalytically active amino acid residues (particularly Arg, Lys, His and Cys) [12,40,41]. These changes result in more difficult electrocatalysis,

manifested by shifting of the peak H to more negative potentials and eventually to the disappearance of this peak. Our preliminary results suggest that some other DNA–protein complexes also behave in a similar way. Moreover, we show that p53CD–DNA complexes can be discriminated according to their stabilities at the electrode surface (Fig. 4). What are the forces, which cause the disintegration of the complex at a negatively charged surface? In the vicinity of the electrode surface a non-homogenous electric field exists (up to 109 V m–1) [23,50]. The width of the inner double layer corresponds to a few atomic diameters. The strength of the electric field in the outer layer decreases in dependence on the layer depth, which is dependent upon ionic strength. Considering the size of the p53CD–DNACON complex [29], it is possible to imagine that the adsorbed complex will be affected by the electric field at the bare HMDE to different extents depending on various factors, such as the length of exposure to this field, orientation of the complex at the electrode surface and the ionic conditions. Thus longer exposure of the p53CD–DNACON complex layer to highly negative potentials (Fig. SI-4), when the DTT layer is already reduced, may result in partial or full denaturation of the protein at the electrode surface [20,23]. Moreover, it has been shown that dsDNA unwinds at a bare mercury electrode charged to potentials close to 1.2 V [8]. Unwinding of dsDNA at other negatively charged surfaces has also been subsequently described. Thus unwinding of dsDNA (starting from the ends of the dsDNA molecules) could contribute to the p53CD–DNA complex destabilization. The main force involved in the disintegration of the complex may likely be a combination of the destabilization of the surface-attached protein and the strong repulsion of the polyanionic DNA from the highly negatively charged mercury surface [8]. The electric field could also have other effects, which may contribute to the destabilization of the surface-attached biomacromolecules, including alteration of the charge distribution in the

Scheme 1. Effect of Istr intensities (current densities) on CPS response of p53CD–DNA complex (red, b and b) and free p53CD (black, c and c0 ). In CPS the rate of potential changes increases non-linearly with current density (i.e., with Istr at constant electrode size) reaching very high scan rates, such as 6000 V/s at Istr 65 mA (see SI and Fig. SI-7 for details). P53CD–DNA does not show any sign of disintegration (a) when adsorbed at DTT–HMDE close to the potential of zero charge (p.z.c.). This complex is however disintegrated (b) when exposed to negative potentials for 1 s (Fig. 2) as indicated by almost the same peak H as that of free p53CD (c). Using high negative Istr intensities, the length of exposure of the surface-attached p53CD–DNA complex can be reduced to milliseconds, saving the complex from disintegration as indicated by very little or no CPS signal (b) at potentials of the peak H of free p53CD greatly differing from peak H of free p53CD under the same condition (c0 ). The susceptibility of the p53CD–DNA complex to electric field effects at negative potentials decreases with decreasing temperature. By other words, the method depicted in this scheme is based on (i) the ability of the electric field to disintegrate at negative potentials the surface attached DNA–protein complex, (ii) controlling the time for which this complex is exposed to negative potentials by applying appropriate polarizing current intensities in CPS, (iii) the ability of the CPS peak H to reflect accessibility of aa residues in surface-attached free p53 protein, as compared to the p53–DNA complex (after their short exposure to negative potentials). (For interpretation of the references to color in this legend, the reader is referred to the web version of this article.)

E. Pale9 cek et al. / Analytica Chimica Acta 828 (2014) 1–8

DNA and protein, brought about by shifts in the acid–base equilibrium toward the ionized forms, alignment of the molecular dipoles, displacement of the charged residues, and polarization of the hydrogen bonds [51]. Our results show that the formation of the p53CD–DNA surface layer may occur under the most convenient conditions [including temperature (Figs. 3 and SI-5) and Istr intensities (Fig. 4)] in a complete disappearance of the peak H1 displayed by free p53 protein (Fig. 1a). Decrease of this peak can be explained due to a decreased accessibility of electroactive groups as mentioned above. The crystallographic structure of the p53CD–DNA complex illustrated the involvement of a number of Arg residues (such as Arg175, Arg249, Arg282, Arg248, Arg273) as well as some Lys and Cys residues in the p53 sequence-specific binding to DNA [1]. But are these interactions sufficient for complete abolition of the p53CD-induced catalytic hydrogen evolution reaction? This appears somewhat improbable as we found similar CPS behavior in different p53CD–DNA complexes, such as those shown in Figs. 1, 5 and SI-3b as well as in DNA complexes with p63 and p73 and in the superstable p53CD–DNA complex (unpublished [35]). It is hard to imagine that the DNA complexes of all these proteins can be oriented at the surface in the same way with all of their electroactive amino acid residues inaccessible and catalytically silent. On the other hand the different adsorption and changed electrical properties of the complex as compared to free p53CD might contribute to the inhibition of the catalytic hydrogen evolution reaction. We studied the properties of the layers of the p53CD protein and its DNA sequence-specific complex at potentials close to zero charge on bare and DTT–HMDEs and we did not observe any dramatic differences between the charge transfer resistance and permeability of the free p53CD vs. the p53CD–DNA complex layers (Fig. SI-8). Unfortunately it is not possible to study such properties at negative potentials, where a peak H1 is produced, because of too fast disintegration of the p53CD–DNA complex at negative potentials (Fig. SI-4). Some conclusions can be drawn however, on the basis of our CPS measurements. We demonstrated that the sequence-specific p53CD–DNACON complex produced almost no peak H (Fig. 1a), while under the same conditions the relatively unstable nonsequence-specific p53–DNANON layer yielded almost the same CPS response as that of free p53CD (Figs. 1a and 4e). In contrast, the sequence-specific complexes of p53CD with the spacer-containing DNACON produced appreciable peaks H, which were (at sufficiently negative Istr) more negative than the peak H of the free p53CD and differed from CPS response of p53CD–DNACON complex, showing almost no peak H (Fig. 5a, b). At Istr 33 mA peaks of all complexes of p53CD with the spacer-containing DNACON indicated disintegration of these complexes (Fig. 5c). Our results obtained at lower current intensities (Figs. 3c and 4d) suggested that the p53CD– DNACON complex was under the given conditions disturbed at the electrode due to prolonged exposition to negative potentials. It appears that this disturbance induced changes in the environment of some amino acids, making them accessible for the electrocatalysis at more negative potentials than in the free p53CD. Our data thus suggest that even complete disappearance of the peak H in very stable sequence-specific p53–DNA complexes need not be interpreted as a complete loss of the catalytic hydrogen evolution reaction. We cannot exclude that in some cases, a highly negative CPS peak reflecting a difficult electrode process, can be too close to the background and merge with it. Considering that decreasing of the peak height and shifting Ep to more negative potentials represent signs of more difficult p53CD electrocatalysis, then among the complexes of p53CD with spacer-containing DNACONs (Fig. 5), the p53CD–DNACON–AT complex, displaying at Istr 35 mA peak H greatly differing from that that of free p53CD, might represent the most stable spacer-containing complex (Fig. 5b).

7

Such conclusion may however be limited to the given conditions at the electrode surface. For more definite conclusions about stabilities of these complexes systematic studies of a number of DNACON samples with different spacers and CONs, measured under different conditions will be necessary. 5. Conclusion Investigation of DNA–protein interactions is one of the most important fields in the current molecular biology, which affects various areas of biomedicine. When inspecting the recent edition of DNA–Protein Interactions Principles and Protocols [52] you can notice over 35 methods. No one of them is however electrochemical in spite of a wide application of electrochemistry of DNA [8,53,54] and protein analysis [15,55]. Here we proposed an electrochemical method, which for the first time utilizes principles of constant current chronopotentiometry for the analysis of DNA– protein interactions. In this method testing of the susceptibility of the surface-attached DNA–protein complex to the effect of the electric field for very short time intervals is performed by investigating changes in accessibility of aa residues responsible for the electro-catalytic process. In this way not only the DNA protein binding can be detected, but also stability of the DNA– protein complex can be tested. Compared to the frequently used gel shift method, CPS in combination with solid amalgam electrodes [19,22] is easily amenable for parallel analysis, yields instant results and does not require radioactive or fluorescence labeling of the DNA. Using 2 mM p53CD, picomole amounts of the p53CD–DNA complex can be determined (Fig. SI-9). Our preliminary results suggest that these amounts can be decreased in at least two ways: (a) by decreasing the electrode size and (b) by increasing the yield of electrons in the electrocatalysis. The specific dependence of the CPS responses of the p53CD–DNA complex on Istr intensities and temperature (Figs. 4 and SI-5b, c) opens the door for other applications of CPS in bioanalysis. In addition to stabilities of different protein–DNA and other protein complexes at electrically charged surfaces, such applications can include drug discovery/chemical biology screens, chromatin assembly, and chaperone dynamics. More details about the analytical applications of this method can be found in Supplementary Information. Acknowledgments šek for The authors are indebted to J. Janata, T. Hupp and B. Vojte critical reading of the manuscript. This work was supported by Czech Science Foundation P301/11/2055 project to EP, P301/13/ 00956S to VO and 13-36108S to MB. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.aca.2014.03.029. References [1] A.C. Joerger, A.R. Fersht, Structure–function-rescue: the diverse nature of common p53 cancer mutants, Oncogene 26 (2007) 2226–2242. [2] J. van Dieck, T. Brandt, D.P. Teufel, D.B. Veprintsev, A.C. Joerger, A.R. Fersht, Molecular basis of S100 proteins interacting with the p53 homologs p63 and p73, Oncogene 29 (2010) 2024–2035. [3] G.D. Stormo, Introduction to Protein–DNA Interactions. Structure, Thermodynamics, and Bioinformatics, Cold Spring Harbor Laboratory Press, New York, 2013. [4] P. Horakova, H. Macickova-Cahova, H. Pivonkova, J. Spacek, L. Havran, M. Hocek, M. Fojta, Tail-labelling of DNA probes using modified deoxynucleotide triphosphates and terminal deoxynucleotidyl transferase. Application in electrochemical DNA hybridization and protein–DNA binding assays, Organic and Biomolecular Chemistry 9 (2011) 1366–1371.

8

E. Pale9cek et al. / Analytica Chimica Acta 828 (2014) 1–8

[5] M. Masarik, K. Cahova, R. Kizek, E. Palecek, M. Fojta, Label-free voltammetric detection of single-nucleotide mismatches recognized by the protein MutS, Analytical and Bioanalytical Chemistry 388 (2007) 259–270. [6] E. Palecek, M. Masarik, R. Kizek, D. Kuhlmeier, J. Hassmann, J. Schulein, Sensitive electrochemical determination of unlabeled MutS protein and detection of point mutations in DNA, Analytical Chemistry 76 (2004) 5930–5936. [7] E.J. Cho, J.W. Lee, A.D. Ellington, Applications of aptamers as sensors, Annual Reviews in Analytical Chemistry 2 (2009) 241–264. [8] E. Palecek, M. Bartosik, Electrochemistry of nucleic acids, Chemical Reviews 112 (2012) 3427–3481. [9] C. Tersch, F. Lisdat, Label-free detection of protein–DNA interactions using electrochemical impedance spectroscopy, Electrochimica Acta 56 (2011) 7673–7679. [10] E. Palecek, Oscillographic polarography of highly polymerized deoxyribonucleic acid, Nature 188 (1960) 656–657. [11] E. Palecek, M. Bartosik, V. Ostatna, M. Trefulka, Electrocatalysis in proteins, nucleic acids and carbohydrates, Chemical Record 12 (2012) 27–45. [12] E. Palecek, V. Ostatna, Electroactivity of nonconjugated proteins and peptides. Towards electroanalysis of all proteins, Electroanalysis 19 (2007) 2383–2403. [13] T.A. Enache, A.M. Oliveira-Brett, Peptide methionine sulfoxide reductase A (MsrA): direct electrochemical oxidation on carbon electrodes, Bioelectrochemistry 89 (2013) 11–18. [14] V. Ostatna, H. Cernocka, K. Kurzatkowska, E. Palecek, Native and denatured forms of proteins can be discriminated at edge plane carbon electrodes, Analytica Chimica Acta 735 (2012) 31–36. [15] O. Hammerich, J. Ulstrup, Bioinorganic Electrochemistry, Springer, Dordrecht, Netherlands, 2008. [16] P.A. Sontz, N.B. Muren, J.K. Barton, DNA charge transport for sensing and signaling, Accounts of Chemical Research 45 (2012) 1792–1800. [17] A.A. Gorodetsky, A.K. Boal, J.K. Barton, Direct electrochemistry of endonuclease III in the presence and absence of DNA, Journal of the American Chemical Society 128 (2006) 12082–12083. [18] J. Vacek, M. Zatloukalova, M. Havlikova, J. Ulrichova, M. Kubala, Changes in the intrinsic electrocatalytic nature of Na+/K+ ATPase reflect structural changes on ATP-binding: electrochemical label-free approach, Electrochemistry Communications 27 (2013) 104–107. [19] P. Juskova, V. Ostatna, E. Palecek, F. Foret, Fabrication and characterization of solid amalgam electrodes array for bioanalytical applications, Analytical Chemistry 82 (2010) 2690–2695. [20] V. Ostatna, H. Cernocka, E. Palecek, Protein structure-sensitive electrocatalysis at DTT-modified electrodes, Journal of the American Chemical Society 132 (2010) 9408–9413. [21] V. Ostatna, H. Cernocka, E. Palecek, Simple protein structure-sensitive chronopotentiometric analysis with dithiothreitol-modified Hg electrodes, Bioelectrochemistry 87 (2012) 84–88. [22] B. Yosypchuk, J. Barek, Analytical applications of solid and paste amalgam electrodes, Critical Reviews in Analytical Chemistry 39 (2009) 189–203. [23] E. Palecek, V. Ostatna, Ionic strength-dependent structural transition of proteins at electrode surfaces, Chemical Communications 13 (2009) 1685– 1687. [24] E. Palecek, V. Ostatna, Potential-dependent surface denaturation of BSA in acid media, Analyst 134 (2009) 2076–2080. [25] C.D. Borsarelli, L.J. Falomir-Lockhart, V. Ostatna, J.A. Fauerbach, H.H. Hsiao, H. Urlaub, E. Palecek, E.A. Jares-Erijman, T.M. Jovin, Biophysical properties and cellular toxicity of covalent crosslinked oligomers of alpha-synuclein formed by photoinduced side-chain tyrosyl radicals, Free Radical Biology & Medicine 53 (2012) 1004–1015. [26] E. Palecek, V. Ostatna, H. Cernocka, A.C. Joerger, A.R. Fersht, Electrocatalytic monitoring of metal binding and mutation-induced conformational changes in p53 at picomole level, Journal of the American Chemical Society 133 (2011) 7190–7196. [27] V. Dorcak, E. Palecek, Electrochemical determination of thioredoxin redox states, Analytical Chemistry 81 (2009) 1543–1548. [28] K. Nemcova, L. Havran, P. Sebest, M. Brazdova, H. Pivonkova, M. Fojta, A labelfree electrochemical test for DNA-binding activities of tumor suppressor protein p53 using immunoprecipitation at magnetic beads, Analytica Chimica Acta 668 (2010) 166–170. [29] B. Vogelstein, D. Lane, A.J. Levine, Surfing the p53 network, Nature 408 (2000) 307–310. [30] E. Kim, W. Deppert, The versatile interactions of p53 with DNA: when flexibility serves specificity, Cell Death & Differentiation 13 (2006) 885–889. [31] B. Ma, Y. Pan, J. Zheng, A.J. Levine, R. Nussinov, Sequence analysis of p53 response-elements suggests multiple binding modes of the p53 tetramer to DNA targets, Nucleic Acids Research 35 (2007) 2986–3001.

[32] N.P. Pavletich, K.A. Chambers, C.O. Pabo, The DNA-binding domain of p53 contains the four conserved regions and the major mutation hot spots, Genes and Development 7 (1993) 2556–2564. [33] I. Beno, K. Rosenthal, M. Levitine, L. Shaulov, T.E. Haran, Sequence-dependent cooperative binding of p53 to DNA targets and its relationship to the structural properties of the DNA targets, Nucleic Acids Research 39 (2011) 1919–1932. [34] A.K. Nagaich, V.B. Zhurkin, S.R. Durell, R.L. Jernigan, E. Appella, R.E. Harrington, p53-induced DNA bending and twisting: p53 tetramer binds on the outer side of a DNA loop and increases DNA twisting, Proceedings of the National Academy of Sciences of the United States of America 96 (1999) 1875–1880. [35] C. Klein, G. Georges, K.P. Kunkele, R. Huber, R.A. Engh, S. Hansen, High thermostability and lack of cooperative DNA binding distinguish the p63 core domain from the homologous tumor suppressor p53, Journal of Biological Chemistry 276 (2001) 37390–37401. [36] J. Dadova, P. Orsag, R. Pohl, M. Brazdova, M. Fojta, M. Hocek, Vinylsulfonamide and acrylamide modification of DNA for crosslinking with proteins, Angewandte Chemie 52 (2013) 10515–10518. [37] A.J. Bard, L.R. Faulkner, Electrochemical Methods: Fundamentals and Applications, Wiley, 2000. [38] M. Lokshin, Y.C. Li, C. Gaiddon, C. Prives, p53 and p73 display common and distinct requirements for sequence specific binding to DNA, Nucleic Acids Research 35 (2007) 340–352. [39] A.N. Bullock, J. Henckel, A.R. Fersht, Quantitative analysis of residual folding and DNA binding in mutant p53 core domain: definition of mutant states for rescue in cancer therapy, Oncogene 19 (2000) 1245–1256. [40] V. Dorcak, V. Ostatna, E. Palecek, Electrochemical reduction and oxidation signals of angiotensin peptides. Role of individual amino acid residues, Electrochemistry Communications 31 (2013) 80–83. [41] V. Vargova, M. Zivanovic, V. Dorcak, E. Palecek, V. Ostatna, Catalysis of hydrogen evolution by polylysine, polyarginine and polyhistidine at mercury electrodes, Electroanalysis 25 (2013) 2130–2135.  [42] H. Cernocká, V. Ostatná, E. Pale9cek, Enzymatic activity and catalytic hydrogen evolution in reduced and oxidized urease at mercury surfaces, Analytica Chimica Acta 789 (2013) 41–46. [43] E. Palecek, M. Brazdova, V. Brazda, J. Palecek, S. Billova, V. Subramaniam, T.M. Jovin, Binding of p53 and its core domain to supercoiled DNA, European Journal of Biochemistry 268 (2001) 573–581. [44] G. Bakalkin, G. Selivanova, T. Yakovleva, E. Kiseleva, E. Kashuba, K.P. Magnusson, L. Szekely, G. Klein, L. Terenius, K.G. Wiman, P53 binds singlestranded-DNA ends through the C-terminal domain and internal DNA segments via the middle domain, Nucleic Acids Research 23 (1995) 362–369. [45] D.B. Veprintsev, S.M.V. Freund, A. Andreeva, S.E. Rutledge, H. Tidow, J.M.P. Canadillas, C.M. Blair, A.R. Fersht, Core domain interactions in full-length p53 in solution, Proceedings of the National Academy of Sciences of the United States of America 103 (2006) 2115–2119. [46] H. Qian, T. Wang, L. Naumovski, C.D. Lopez, R.K. Brachmann, Groups of p53 target genes involved in specific p53 downstream effects cluster into different classes of DNA binding sites, Oncogene 21 (2002) 7901–7911. [47] R.L. Weinberg, D.B. Veprintsev, M. Bycroft, A.R. Fersht, Comparative binding of p53 to its promoter and DNA recognition elements, Journal of Molecular Biology 348 (2005) 589–596. [48] E. Palecek, V. Brazda, E. Jagelska, P. Pecinka, L. Karlovska, M. Brazdova, Enhancement of p53 sequence-specific binding by DNA supercoiling, Oncogene 23 (2004) 2119–2127. [49] M. Kitayner, H. Rozenberg, R. Rohs, O. Suad, D. Rabinovich, B. Honig, Z. Shakked, Diversity in DNA recognition by p53 revealed by crystal structures with Hoogsteen base pairs, Nature Structural and Molecular Biology 17 (2010) U423–U458. [50] E. Neumann, Chemical electric-field effects in biological macromolecules, Progress in Biophysics and Molecular Biology 47 (1986) 197–231. [51] H. Wackerbarth, P. Hildebrandt, Redox and conformational equilibria and dynamics of cytochrome c at high electric fields, ChemPhysChem 4 (2003) 714–724. [52] T. Moss, B. Leblanc, DNA–Protein Interactions: Principles and Protocols, Humana Press, New York, 2009. [53] E. Palecek, Polarographic techniques in nucleic acid research, in: J.N. Davidson, W.E. Cohn (Eds.), Progress in Nucleic Acid Research & Molecular Biology, Academic Press, New York, 1969, pp. 31–73. [54] E. Palecek, Premelting changes in DNA conformation, Progress in Nucleic Acid Research and Molecular Biology 18 (1976) 151–213. [55] E. Palecek, F. Scheller, J. Wang, Electrochemistry of Nucleic Acids and Proteins. Towards Electrochemical Sensors for Genomics and Proteomics, 1st ed., Elsevier, Amsterdam, 2005.