TY - JOUR
AU - Reguera, Gemma
AB - Abstract Geobacter bacteria assemble a helical peptide of the Type IVa pilin subclass as conductive pili decorated with metal binding and reduction sites. We used recombinant techniques to synthesize thiolated pilin derivatives and self-assembled them on gold electrodes as a monolayer that concentrated the metal traps at the liquid interface. Cyclic and step potential voltammetry demonstrated the conductivity of the pilin films and their ability to bind and reductively precipitate divalent cobalt (Co2+) in a diffusion-controlled reaction characterized by fast binding kinetics, efficient charge transfer, and three-dimensional nanoparticle growth at discreet sites. Furthermore, cobalt oxidation at the pilin film was slower than on bare gold, consistent with a peptide optimized for metal immobilization. These properties make recombinant pilins attractive building blocks for the synthesis of novel biomaterials for the immobilization of toxic cationic metals that, like Co2+, are sparingly soluble and, thus, less mobile and bioavailable as reduced species. Electronic supplementary material The online version of this article (10.1007/s10295-019-02167-5) contains supplementary material, which is available to authorized users. Introduction Geobacter bacteria discharge respiratory electrons to extracellular metals, both soluble and insoluble, using conductive appendages of the Type IV pilus class [4, 21]. The conductive pili are homopolymers of a helical peptide [4] that retains structural and amino acid sequence signatures conserved in pilins of the Type IVa subclass [7, 21]. Yet, once assembled, aromatic amino acids of the Geobacter pilins cluster at distances and in geometric configurations that permit charge hopping and pilus conductivity at rates (~ 9 × 108 electrons per second) about one hundred times faster than the cellular rates of respiration of iron oxides, the natural electron acceptor for these bacteria [7, 11]. This reductive reaction solubilizes some of the ferric iron (Fe[III]) as Fe(II) and generates magnetite, a magnetic mineral of mixed Fe(III)/Fe(II) valence [21]. Each cell assembles several pili monolaterally, significantly expanding the redox active surface of the cell beyond the confines of the outer membrane [19]. In addition, cells alternate cycles of pilin polymerization and depolymerization in the cell envelope to grow and retract the pilus filaments, respectively [23]. Protrusion of the pilus nanowires allows the cells to bind and reduce the extracellular metal, whereas retraction provides a natural mechanism to shed off reduced mineral particles that would otherwise remain attached to the appendages [23]. This strategy maximizes access to the iron minerals, which are often dispersed in soils and sediments, and rapidly transform abiotically into more crystalline and less bioavailable mineral forms [20]. The pilus nanowires also bind the soluble uranyl cation and reductively precipitate it to a mononuclear mineral phase to gain energy for respiration and prevent the permeation of the toxic radionuclide inside the cell [4]. The extended X-ray absorption fine structure (EXAFS) spectrum of pili-bound uranium shows the radionuclide caged within two lateral bidentate carboxylate and one distal monodentate ligands [4] that match well the surface clusters of negatively charged residues revealed in an atomic resolution model of the pilus fiber optimized via molecular dynamics [7]. Figure 1 shows these anionic residues in the molecular structure of the PilA pilin peptide and their carboxy terminal (C-t) localization with the tyrosine (Y57) that is predicted to catalyze the last step of electron transfer to the caged metal [7]. Two of the anionic ligands (aspartic residues D53 and D54) form salt bridges between neighboring pilins and are unlikely to participate in metal binding [7]. The other anionic residues are exposed to the solvent and available to function as metal ligands. Further, two of them (glutamic E60 and the terminal serine S61 residues) reside in the pilin’s C-t random coil, a segment whose dynamics in solution could facilitate the binding of metals [7]. Fig. 1 Open in new tabDownload slide Molecular structure (a), domain architecture (b), and amino acid sequence (c) of the mature PilA pilin of G. sulfurreducens (PDB structure, 2m7g) in reference to the recombinant peptides PilA19 and PilA19-C. Shown are the pilins’ structural domains (α1-helix in gray; C-t random coil in blue), predicted metal trap (red-shaded circle); aromatic residues (F, phenylalanine, in green; Y, tyrosine, in orange); negatively charged side chains (in red) in the metal trap (E for glutamic and D for aspartic amino acids); and N-t alanine (A20) in PilA19 replaced by cysteine tag (A20C) in PilA19-C Because metal binding by the metal traps on the pilus filament is dominated by electrostatic interactions, cationic metals such as divalent cobalt, cadmium, and lead may be bound. Given the high toxicity of these metals to the cells, there is high interest in developing in vitro platforms that harness the binding and reductive properties of the Geobacter pili [19]. Recombinant pilin derivatives are promising building blocks for such platforms because they can be synthesized at high yields in heterologous hosts such as Escherichia coli [5]. The lack of cysteines in the amino acid sequence of the PilA pilin of G. sulfurreducens (Fig. 1) also makes this amino acid suitable as a functional tag for covalent binding of the peptides to materials (e.g., metals) typically used in electronic devices. In a previous study, we demonstrated the application of recombinant techniques to synthesize soluble pilin derivatives that carried truncations of several amino acids at the amino terminus (N-t) [5]. These pilin derivatives can be genetically engineered to carry a cysteine at the N-t for their covalent attachment to gold electrodes and spontaneous self-assembly as well-ordered, conductive planar films approximately 4 nm thick [5]. The pilins in the film are tilted, a configuration that promotes the clustering of aromatic residues in the 2 nm closer to the underlying gold electrode and provides a path for charge hopping in this region [5]. By contrast, tight interactions between the aromatic-free helical regions in the upper 2 nm of the film favor tunneling regimes [5]. The pilin self-assembled monolayer (pSAM) is also predicted to expose to the solvent the negative-charged amino acid ligands and terminal tyrosine (Fig. 1) needed for metal binding and reduction [5]. This suggests that the pSAM surface, like the native pilus nanowires [7], is decorated with electrostatic traps that could bind cationic metals and position them close to the terminal tyrosine to facilitate their reduction. To test this, we optimized conditions for the recombinant synthesis of thiolated PilA19 pilins (PilA19-C) and their self-assembly on gold electrodes. We then used voltammetric techniques (cyclic and step potential voltammetry) to investigate their conductivity and the kinetics of binding and reduction of the divalent cobalt cation (Co2+). The electrochemical studies unmasked electronic and kinetic features in the pilin films that are consistent with the presence of specialized sites for the immobilization and reductive precipitation of Co2+ and formation of Co0 nanoparticles. The suitability of pilin-based molecular assemblies for the development of hybrid devices for sensing, remediation, and reclamation of toxic cationic metals such as Co2+ is discussed. Materials and methods Recombinant expression of thiolated pilins The thiolated pilin peptide (PilA19-C) used in this study is a cysteine-tagged version of the PilA19 peptide described earlier [5]. Briefly, the pilA 19 sequence was PCR amplified from the pilA gene of G. sulfurreducens (GSU1496) with primers (forward PilA19 5′-GGTGGTTGCTCTTCCAACGCGATTCCGCAGTTCTCGGC-3′ and reverse PilA19 5′-GGTGGTCTGCAGTCATTAACTTTCGGGCGGATAGGT-3′) designed to truncate the first 19 amino acids of the mature PilA pilin. Thus, the bolded GCG codon in the forward primer encodes the first amino acid (an alanine) of the truncated pilin PilA19 (Fig. 1). The primers also include Sap I or Pst I restriction sites (underlined) for cloning the pilA 19 sequence downstream of the intein linker and chitin-binding domain (CBD) of plasmid pTYB11 (IMPACT™-CN system; New England Biolabs). The pTYB11::pilA 19 plasmid was used as a template in a PCR reaction with the QuikChange Lightning Site-Directed Mutagenesis kit (Agilent Technologies) to replace the GCG codon in pilA 19 (alanine 1) with a cysteine using primers 5′-GGTTGTTGTACAGAACTGCATTCCGCAGTTCTCGG-3′ (cysteine codon is underlined) and 5′-CCGAGAACTGCGGAATGCAGTTCTGTACAACAACC-3′. The resulting plasmid, pTYB11::pilA 19-C, was transformed into Escherichia coli Rosetta™ 2 (DE3) pLysS cells (Novagen). This strain was used for the recombinant expression of the PilA19-C peptide as a fusion protein with an intein linker-CBD domain, which served as an affinity tag for purification of the recombinant peptide. Escherichia coli cells carrying the pTYB11::pilA 19-C plasmid were grown to an OD600 of ~ 0.5 at 37 °C in 1L of LB medium with 100 µg/ml ampicillin and 20 µg/ml chloramphenicol. Recombinant expression of the PilA19-C fusion protein was then induced by adding 50 mM isopropyl β-d-1-thiogalactopyranoside (IPTG) and incubating overnight at 16 °C. Cells were harvested, lysed by sonication in 20 mM Tris HCl buffer (100 mM NaCl, 1 mM EDTA, 1% CHAPS) and centrifuged to collect the clarified lysate with the fusion protein among other soluble proteins. The protein lysate was then loaded in a chitin column (New England Biolabs) equilibrated with 20 mM Tris buffer (100 mM NaCl, 1 mM EDTA, pH 7.4) and allowed to incubate at room temperature for 20 min to promote the binding of the protein’s CBD domain to the chitin matrix. Washing the column with a buffer containing increasing salt concentrations (20 mM Tris, 1 mM EDTA, 0.6–1 M NaCl, pH 7.4) removed non-specifically bound protein. Column incubation at room temperature for 24 h with 20 mM Tris buffer (100 mM NaCl, pH 9) containing freshly made 1,4-dithiothreitol (DTT, 50 mM) then followed to induce the intein-mediated cleavage of the recombinant peptide. Elution of the cleaved peptide from the column was with the same buffer without DTT. The column eluent was collected in 2 ml aliquots and peptide-containing fractions were identified by UV absorption at 280 nm and pooled together. The peptide solution was then incubated with freshly prepared 5 mM DTT for 1 h at 37 °C in an anaerobic chamber [COY glove bag; H2:CO2:N2 (7:10:83) atmosphere] to break disulfide linkages. Lastly, the buffer was exchanged to acetonitrile in a reverse phase C18 cartridge column (3 ml column volume, Waters) inside the anaerobic chamber prior to assembly onto electrodes, as described below. Self-assembly of thiolated pilins onto gold electrodes Gold electrodes cut to size (2–2.5-cm2 squares) from commercial wafers (LGA Thin Films, with 150 Å Cr and 1000 Å Au on silicon) with a diamond-tipped scribe were freshly cleaned with piranha solution (70% concentrated sulfuric acid and 30% hydrogen peroxide), rinsed with ddH2O and dried under a stream of N2 gas. Electrode functionalization was at room temperature inside an anaerobic enclosure (COY glove bag) for 48 h using a solution of PilA19-C in acetonitrile or 1 mM 1-undecanethiol in ethanol and resulted in the formation of a conductive pilin self-assembled monolayer (pSAM) or an insulating undecanethiol self-assembled monolayer (uSAM), respectively. When indicated, freshly formed pSAMs were rinsed thoroughly with acetonitrile and dried prior to treatment with 1 mM 1-undecanethiol in ethanol for 48 h to produce pSAM–uSAM control electrodes that tested the saturating coverage of the pSAM. Once functionalized, the electrodes were removed from the glove bag, rinsed thoroughly with acetonitrile (pSAM) or ethanol (uSAM), dried under a stream of Argon gas and placed individually in a well of a 6-well cell culture plate for short-term storage at room temperature. Cyclic and step potential voltammetry Voltammetric studies used a conventional three-electrode cell equipped with a 1-mm-diameter working electrode window. The working electrode was a gold electrode, either bare or functionalized with a pSAM, uSAM or hybrid pSAM–uSAM film. The counter electrode was a platinum wire and the third electrode, a 3 M silver/silver chloride (Ag/AgCl) reference electrode (Bioanalytical Systems, Inc.). The redox analytes were 1 mM potassium ferricyanide (K3Fe[CN]6), 10 mM ferric chloride (FeCl3), and 10 mM cobalt(II) chloride (CoCl2) prepared in 50 mM Tris–HCl buffer (100 mM NaCl, pH 7). Approximately, 1 ml of the electrolyte was dispensed in the three-electrode cell for the electrochemical measurements. Cyclic voltammetry experiments were recorded using a potentiostat (Bio-Logic USA, VSP model) connected to a lab computer equipped with the EC-Bio labs software for data acquisition. Scans were always started at a positive potential and proceeded in the negative direction before reversal. Individual voltammograms were then imported to Igor Pro 6 for analyses. Each working electrode was tested in at least two independent experiments and a minimum of five sweeps per scan rate to ensure reproducibility. Representative voltammograms were generated as the average of the last three of five sweeps for each conditions/electrode tested. The midpoint potential of the reaction with each working electrode versus the Ag/AgCl reference electrode was calculated as the mean of the potential of the anodic peak current (E pa) and of the potential of the cathodic peak current (E pc) recorded in representative voltammograms collected a scan rate of 100 mV/s. Step potential experiments used the same experimental set-up with bare gold or pSAM–gold working electrodes and 10 mM CoCl2 (in 50 mM Tris–HCl buffer with 100 mM NaCl at pH 7) as the redox analyte. The experiments started with the application of a positive voltage (+ 0.8 V) to ensure that no cobalt was deposited on the working electrode. The potential was then reversed to − 0.8 V to promote the electrodeposition of Co2+ to saturation (40 s) while measuring the current response with a potentiostat. Anodic stripping was then investigated by incrementally increasing the potential (10, 50, 100, 150 and 210 mV, as indicated), which are values above those recorded in cyclic voltammograms for the most positive dominant peak of anodic current. The current response during anodic stripping was recorded with the potentiostat to estimate the time needed to reach zero current. Results and discussion Cyclic voltammetry with iron species demonstrates the conductivity of pilin monolayers We investigated the kinetics of electron transfer at the pSAM surface in cyclic voltammograms with 1 mM potassium ferricyanide using bare gold and insulating uSAM–gold working electrodes as positive and negative controls, respectively (Fig. 2a–c). As the potential (E) swept from positive to negative at a defined scan rate, bare gold electrodes reduced the ferric (Fe[III]) species in ferricyanide (K3Fe(CN)6) at the electrode surface to the ferrous form (Fe[II]), leading to the formation of the potassium ferrocyanide salt (K4Fe(CN)6). This reductive reaction was recorded in the voltammograms as a peak of negative current (cathodic peak). Reversing the potential back to positive at the same scan rate oxidized the ferrocyanide back to ferricyanide (Fe[II] to Fe[III]) and an anodic peak was recorded. Figure 2a shows, as an example, typical ferricyanide voltammograms obtained for bare gold electrodes at a scan rate of 100 mV/s, with symmetric cathodic and anodic peaks expected for the reversible cycling of ferricyanide at the electrode surface. However, the cathodic and anodic peaks disappeared when the electrode surface was insulated with a monolayer of 1-undecanethiol (uSAM) (Fig. 2a). As with bare gold, the voltammograms recorded for electrodes functionalized with the pSAM had symmetric cathodic and anodic peaks consistent with a diffusible and reversible redox reaction (Fig. 2a). Furthermore, treating the pSAM with 1-undecanethiol (pSAM–uSAM) did not have a significant effect on the voltammograms (Fig. 2a). This in agreement with atomic force microscopy and ellipsometry studies [5], which showed that thiolated pilins self-assemble on gold electrodes as a dense, confluent monolayer that fully covers the gold electrode. Thus, any redox reaction at the pSAM and pSAM–uSAM electrodes is solely mediated by the pilins. Fig. 2 Open in new tabDownload slide Cyclic voltammetry of pSAM and controls (bare gold and uSAM) with ferricyanide or ferric chloride. a, d Voltammograms collected at 100 mV/s with 1 mM K3[Fe(CN)6] (a) or FeCl3 (d) in 50 mM Tris HCl buffer (pH 7, 100 mM NaCl) using bare gold (yellow), pSAM (solid green), pSAM treated with 1-undecanethiol (pSAM–uSAM, dashed green), or insulating uSAM controls (grey). b, e Effect of scan rate (25, 50, 75, 100, 150 and 300 mV/s, darkest to lightest color) on cyclic voltammograms collected with K3[Fe(CN)6] (b) or FeCl3 (e) for pSAM (green, main plot) or bare gold (yellow, inset; scale of axes is as in main plot). c, f Absolute value of the pSAM anodic (black) and cathodic (red) peak current as a function of the square root of scan rate. The coefficient of determination (R2) of the linear regression lines are also shown in the same color Increasing the scan rates from 25 to 300 mV/s also increased the amount of current recorded at the anodic and cathodic peaks for both the pSAM-functionalized (Fig. 2b) and bare (Fig. 2b, inset) gold electrodes. As with bare gold voltammograms, cathodic and anodic currents with the pSAM increased linearly as a function of the square root of scan rate (Fig. 2c). This linearity agrees well with the Randles–Sevcik equation for a reversible electron transfer process with freely diffusing redox species [6]. From the pSAM and pSAM–uSAM voltammograms recorded at 100 mV/s (Fig. 2a), we calculated a midpoint potential for ferricyanide (+ 0.311[± 0.002] V vs Ag/AgCl) similar to bare gold (+ 0.313[± 0.002] V). The mid-potential value recorded on bare gold is within the upper ranges reported for gold electrodes by cyclic voltammetry (e.g., + 0.24 V vs Ag/AgCl [1]) and reflects differences in the assay conditions (electrolyte type and concentration of redox species), which we optimized in our study for the reversible redox cycling of ferri/ferrocyanide by pSAMs. Despite having similar midpoint potentials, peak–peak separation in the pSAM voltammograms (roughly between + 0.1 and + 0.5) was larger than on bare gold (roughly between + 0.2 and + 0.4) (Fig. 2b), consistent with a slower electron transfer process at the pSAM [2, 6]. This is not unexpected because the planar assembly of pilins also concentrates the anionic ligands that mediate metal binding and optimal coordination of the metal for its reduction. Repulsion forces between the pilins’ metal traps and the negatively charged ferri/ferrocyanide species ([Fe(CN)6]3−/4−) are, therefore, likely to arise that slow down the redox cycling of the chelated iron species at the pSAM surface compared to bare gold. The role of electrostatic interactions in metal binding and reduction at the pSAM was further investigated in cyclic voltammograms with ferric chloride (FeCl3) in the analyte (Fig. 2d–f). Being a more unstable salt of iron compared to chelated iron in ferricyanide, we predicted favorable electrostatic interactions between the ferric and ferrous iron cations (Fe3+/2+) in the chloride salt and the pilins’ metal traps. As with ferricyanide, the FeCl3 voltammograms with bare and pSAM-coated gold had symmetric cathodic and anodic peaks consistent with a diffusible and reversible redox reaction, and no current was recorded with insulating uSAM controls (Fig. 2d). Furthermore, the midpoint potential vs Ag/AgCl for FeCl3 was similar for the bare gold (+ 0.606[± 0.023] V) and pSAM (+ 0.602[± 0.012] V) electrodes. However, unlike ferricyanide voltammograms, peak–peak splitting did not increase in the presence of the pSAM (roughly between + 0.4 and + 0.8 V in both; Fig. 2e). Thus, despite the presence of a 4-nm-thick peptide monolayer on the surface of the gold electrode, electron transfer at the pSAM proceeded at rates similar to those at the gold electrode surface. Furthermore, the correlation between the pSAM cathodic current and the square root of scan rate was linear (Fig. 2f), indicative of a freely diffusing redox species [6]. Hence, electrostatic interactions between the iron cation species (Fe3+/2+) and the anionic ligands of the pilin metal traps did not limit the rates of electron transfer during the reversible cycling of FeCl3. Cyclic voltammetry demonstrates the electrodeposition of divalent cobalt at discreet pilin sites Metal-reducing bacteria like Geobacter can reduce chelated forms of Co3+ to Co2+, but the toxicity of the divalent cobalt cation is often assumed to prevent its biological reduction [8]. Yet, once in solution, Co2+ forms a stable complex with water ([Co(H2O)6]2+) that, like the hydrated uranyl cation ([UO2(H2O)4]2+), can be bound in a tetrahedral coordination geometry by the carboxyl groups of two of the anionic ligands in the pilin metal traps [4]. Thus, strong interactions and optimal coordination for electron transfer reactions are predicted between the hexaaqua cobalt(II) cation and the pSAM metal traps that would lead to the reductive precipitation of Co2+ to Co0. Cyclic voltammograms recorded with bare gold consistently showed a maximum cathodic peak at about − 0.7 V (Fig. 3a). This voltage value is higher than the potential (− 1 V or below) that triggers the reductive desorption of thiol-linked peptides on gold [12] and molecular deformations in films of helical peptides [17]. Thus, we investigated the kinetics of cobalt cycling by pSAM at a maximum negative potential voltage sweep of − 0.8 V. Fig. 3 Open in new tabDownload slide Cyclic voltammetry of bare gold and pSAM with cobalt. a Cyclic voltammograms collected with a maximum negative potential of − 0.8 V at 100 mV/s on bare gold (yellow) and pSAM (green) electrodes with 10 mM CoCl2 in 50 mM Tris–HCl buffer (100 mM NaCl, pH 7). b, c Effect of scan rate (25, 50, 75, 100, 150 and 300 mV/s, darkest to lightest shade) on cyclic voltammograms collected for bare gold (b) and pSAMs (c) with anodic peaks labelled. d, e Expanded view of cathodic peaks for both gold (d) and pSAMs (e) at scan speeds 25–100 mV/s (darkest to lightest color shade of yellow and green, respectively) In contrast to the reversible redox cycling of iron species (Fig. 2), cobalt voltammograms revealed an electrodeposition process on bare gold and pSAM whereby the Co2+ species was first reduced to the sparingly soluble elemental Co0, and the reduced Co0 then acted as nucleation site for more Co2+ deposition and nanoparticle growth. This shows in cyclic voltammograms as an asymmetric current response with a much larger cathodic peak and a distinctive crossover feature (Fig. 3). Figure 3a shows as an example asymmetric voltammograms recorded at a scan rate of 100 mV/s with bare gold and pSAM electrodes. During the forward scan from positive to negative values, current production increased sharply (dominant cathodic peak) in both electrodes, marking the reduction of Co2+ to Co0 and the reductive precipitation of more Co2+ onto the Co0 atom (nucleation or crystal growth). As cobalt-on-cobalt deposition (homodeposition) proceeds, the density of metal nuclei forming on the surface increases and more current is recorded at the cathodic peak [18, 22, 24]. In fact, more current is generated during homodeposition than during the initial deposition of cobalt onto the electrode surface (heterodeposition) [22]. The formation of a cobalt layer on the electrode surface during the forward scan also influenced the current generated during the reverse scan. Homodeposition continued even after reversing the voltage but less current was generated during the reverse scan than during the forward scan. This caused the forward and reverse scan lines in the voltammograms to cross at the crossover potential, E c (Fig. 3a), as it is typical during electrodeposition [22]. The voltammograms revealed only one crossover event for both gold and pSAMs at 100 mV/s (Fig. 3a), but the crossover potential E c became progressively more negative at faster scan rates (Fig. 3b, c) until disappearing at scan rates greater than 150 mV/s for bare gold or lower voltages (between 50 and 200 mV/s) for pSAMs (Table S1). Cyclic voltammograms recorded with CoCl2 also showed small, secondary cathodic peaks of underpotential deposition (upd peaks) prior to the emergence of the dominant cathodic peak (Fig. 3). These secondary peaks reflect the heterogenous energy landscape of the electrode surface and unmask potentials that favor stronger interactions between the Co2+ cation and the electrode surface. This energetic preference causes the metal to first deposit from solution onto an electrode at a more positive potential (underpotential) than that recorded in the cathodic peak [14]. Bare gold and pSAM voltammograms both revealed three small cathodic peaks of underdeposition (E pc1–3) during the forward scan (Fig. 3d, e and Table S1). Upd peaks are not uncommon in voltammograms recorded on bare gold electrodes because the surface roughness creates step edges that can serve as nucleation sites for the underdeposition of metals [9]. These nucleation sites provide a heterogeneous energy landscape, with each energy level showing in the voltammograms as one of three upd peaks (Fig. 3d). The most energetically favorable nucleation site becomes occupied early in the forward sweep (E pc1), followed by the less energetically favorable nucleation sites (E pc2 and E pc3) (Fig. 3d) until reaching a sufficiently low voltage (E c) that permits surface-wide deposition (Fig. 3a). The first upd peak (E pc1) also marks a heterodeposition event, where cobalt in solution is bound and reduced at the most energetically favorable nucleation sites on the gold electrode. Because cobalt heterodeposition is diffusion limited [6], the correlation between the current measured at the first upd potential (I pc1) on gold and the square root of the scan rate is linear (Fig. 4a). These localized sites of initial cobalt deposition on gold then drive cobalt-on-cobalt deposition (homodeposition) events that lead to the formation of two new upd peaks at potentials E pc2 and E pc3 (Fig. 3d). As it is typical of electron transfer reactions involving surface-adsorbed species [6], the correlation between the current at the second and third upd peaks (I pc2 and I pc3) and the square root of the scan rate is no longer linear, but polynomial (Fig. 4a). Fig. 4 Open in new tabDownload slide Current at anodic and cathodic peaks in cobalt voltammogram peaks for bare gold and pSAM. The plots show the best fit (linear or polynomial; R2 > 0.9 in all) to the current measured at the cathodic (a, b) and anodic (c, d) peaks (I pc and I pa, respectively) versus the square root of scan speed (√(mV/s)) from representative voltammograms recorded with bare (yellow) or pSAM-functionalized (green) gold electrodes with 10 mM CoCl2 Scan rates of 50 mV/s or higher favored non-specific electrodeposition on pSAM and also produced three upd cathodic peaks (E pc1–3) (Table S1). As with gold, the relationship between current at the pSAM’s upd peaks and the square root of scan speed was initially linear (I pc1), but best fitted a polynomial curve in the later peaks (I pc2 and I pc3) once homodeposition was initiated (Fig. 4b). This behavior agrees with the model of a diffusion-limited first upd seeding event that modulates electron transfer during the second and third upd processes. Despite these similarities, the first upd event in pSAMs was recorded at a more negative potential (E pc1 in Table S1), indicating a higher energetic barrier to non-specific deposition on the pilins than on bare gold at these scan rates [6]. Thus, the energy landscape is more homogenous on pSAM, consistent with the reductions in surface roughness noted for gold electrodes functionalized with the pilin films [5]. Further, the three pSAM upd peaks progressively decreased at lower scan rates, eventually disappearing at 25 mV/s (Fig. 3e and Table S1), a behavior expected for non-specific deposition events. The disappearance of the upd peaks at low scan rates coincides with the emergence of a cathodic peak (E pcS, around − 0.4 V versus Ag/AgCl) that is unique to the pSAMs (Fig. 3e and Table S1). This peak unmasks the potential (E pcS) of the reduction of Co2+ at the pilin’s metal traps. Current at this potential is also higher in pSAM (Fig. 3c) than in gold (Fig. 3b) voltammograms, consistent with a more efficient electrodeposition at the pilin’s metal traps than on bare gold at this voltage (E pcS). Scan rates also influenced the voltage that led to the emergence of the dominant anodic peak, shifting from more negative (E pa1) to less negative (E pa2) as the speed of the scan increased (Table S1). This shift was present in both bare gold (Fig. 3b) and pSAM (Fig. 3c), but it was more pronounced in the pSAM voltammograms at the low scan rates that unmasked specific oxidation reactions at the metal traps. The linear correspondence between the current at the dominant anodic peak recorded at higher scan speeds (I pa1) and the square of the scan rate shows that non-specific events are mainly diffusion limited in both electrodes (Fig. 4c, d). By contrast, the correlation is best fitted to a polynomial curve for the dominant anodic peak I pa2 of voltammograms recorded at slow scan rates (Fig. 4c, d), as expected of electron transfer reactions involving surface-adsorbed species [6]. Current for this anodic peak of homodeposition is about threefold higher on pSAM (Fig. 4d) than on bare gold (Fig. 4c) at the lowest (25–100 mV/s) scan rates. This is because the more efficient deposition of cobalt on the pSAM surface during the forward scan increased the availability of the reduced metal for oxidation and solubilization during the reverse scan and, proportionally, the current recorded at the dominant anodic peak (Fig. 3c). The reverse scan also revealed in gold and pSAM voltammograms small anodic peaks (E pa3 and E pa4) at potentials more positive than the dominant anodic peaks (E pa1 or E pa2). The correlation between the current produced at these secondary anodic peaks versus the square root of scan speed was largely linear in both electrodes (Fig. 4c, d). Yet these secondary peak potentials were much more positive for pSAM-functionalized electrodes than for bare gold (Table S1), indicating that more energy is needed to oxidize and fully resolubilize cobalt atoms bound to the pSAM. This is in agreement with the prediction of strong interactions between cobalt and ligands at the pilin metal traps, which retain the metal to the surface until the potential is sufficiently high to overcome the energy barrier. The secondary anodic peaks could also reflect differences in the rates of electron transfer across the pSAM during the reverse scan. Indeed, pSAMs show rectifying behavior such that the natural direction of electron transfer across pilins in the nanowires (from the N-t to the metal traps in the C-t region) is favored [5]. Thus, more energy may be required to move electrons across the pSAM from the reduced cobalt nanoparticles to the underlying gold electrode during the anodic sweep. Step potential voltammetry unmasks strong pilin–cobalt interactions We gained additional insights into the specific mechanism of cobalt immobilization and redox cycling by pSAMs in a three-step potentiostatic experiment that measured transient current events during cobalt electrodeposition at and resolubilization from the pilin surface compared to bare gold (Fig. 5). Cobalt was electrodeposited to saturation on the surface of bare gold or pSAM electrodes by applying a potential of − 0.8 V to the working electrode for 40 s (Fig. 5a). The electrodeposition of cobalt onto the bare gold electrode control showed a simple profile that tapered to a steady current production in less than 5 s (Fig. 5a). The current then rose rapidly after the change in voltage and began to drop until the surface of the electrode was completely covered by the cobalt layer. The rapid formation of this peak is likely influenced by the surface roughness of the gold electrode, as nucleation rates increase with the number of surface defects on the electrode surface [9]. The shape of this deposition event is also characteristic of a 3-D deposition process with multiple nucleation sites controlled by diffusion of the metal species to the surface [9, 15, 16]. Cobalt electrodeposition on pSAM also led to the rapid formation of a peak of current at − 0.7 V, but current at this peak doubled compared to bare gold and plateaued at a much higher current level (Fig. 5a). The pSAM film masks step edges and defects on the gold electrode [5], effectively reducing non-specific deposition compared to bare gold. Yet the smoother pSAM surface also increased the magnitude of the current response during metal electrodeposition (Fig. 5a). This current behavior is in agreement with a higher efficiency of binding, reduction and nucleation at the pSAM surface. Indeed, cobalt deposition on pSAM behaves closely to a surface with restricted nucleation sites [9] and interactions of cobalt with discrete metal binding sites that favor charge transfer, an important component of the deposition process [13]. This is expected from the presence on the pSAM of specialized sites for metal immobilization, reduction and nucleation. Also contributing to the efficiency electrodeposition is the more favorable direction of electron transfer across the pSAM film during metal reduction, which uses the preferred biological polarity of charge transport from the N-t to the C-t (Fig. 6) [5]. Fig. 5 Open in new tabDownload slide Current response during saturating cobalt electrodeposition (a) and step potential resolubilization (b, c). The experiments were conducted with 10 mM CoCl2 in 50 mM Tris–HCl buffer (100 mM NaCl, pH 7) on bare gold (yellow) and pSAM–gold electrodes (green) and included a 40-s-long electrodeposition step at − 0.8 V to saturate the electrode surface with the cobalt mineral followed by step potential stripping. Shown are the current responses during the first 5 s of cobalt electrodeposition, when current has reached a plateau (a) and during the oxidization of the cobalt layer from bare gold (b) or pSAM (c) electrodes at positive potentials (210, 150, 110, 50, and 10 mV; lightest to darkest color). The return of current production to zero (dashed line) marks the completion of the stripping process (b, c) Fig. 6 Open in new tabDownload slide Model of pSAM and electron flow during cobalt electrodeposition. Model at left shows a pSAM on gold with the anionic residues (in red) of the metal traps on top and the aromatic residues (tyrosines, orange; phenylalanines, green) that mediate electron transfer. The cartoon at right illustrates the flow of electrons from the underlying electrode through the pilin’s hybrid electronic path (aromatic hopping and interchain tunneling) until reaching the C-t aromatic residues that reduce Co2+ to Co0 at one of the pilin traps and promote the growth of the metal nanoparticle. The efficient process for electrodeposition at the pSAM surface is further supported by the dynamics of cobalt solubilization (Fig. 5b, c). After 40 s of cobalt deposition to saturate the electrode surface, the potential was shifted to potentials (10, 50, 110, 150, and 210 mV) more positive than the most positive cobalt anodic peak (E pa2 in cyclic voltammograms; Table S1) in order to oxidize and solubilize the cobalt layer. The current response was then measured until it returned to baseline to estimate the time needed to fully solubilize all of the cobalt from the electrode surface. The solubilization of cobalt from the surface of bare gold completed in at most 15 s and produced a simple current decay profile (Fig. 5b). By contrast, cobalt oxidation from the pSAM took 50–65 s (from highest to lowest potentials, respectively) and involved a more complex process, as indicated by the peaks in current that appeared at different times during the oxidation reaction (Fig. 5c). This complex current profile results from the unique binding kinetics of cobalt at the pSAM, whose metal traps promote the immobilization of the metal and thus retain cobalt at the surface more strongly. Rectification across the pSAM may have also contributed to the slower solubilization process but this was likely restricted to the lowest potentials (< 100 mV). Indeed, the polarity of electron transfer across the pSAM during the oxidation reaction is from the cobalt deposits at the metal traps (C-t) to the N-t cysteine linker and the underlying electrode (Fig. 6). But rectification is only measurable at the low (100 mV) potentials that operate biologically [5]. Conclusions The electrochemical studies on pSAM electrodes highlight material properties of planar assemblies of conductive recombinant pilins that make them excellent in vitro tools for metal immobilization. Figure 6 shows a model of the pSAM that illustrates the high concentration of anionic metal traps at the liquid interface. Some of the negatively-charged side chains are located in the flexible C-t segment of the pilin, whose exposure to the solvent facilitates the immobilization of soluble cationic metals. The exposure and coordination of these anionic side chains favors electrostatic interactions with the hexaaqua cobalt(II) cation and position the Co2+ atom close to the terminal tyrosine (Y57) for its reductive precipitation to Co0. Electron transfer across the pSAM is favored in this direction thanks to the presence of aromatic dopants (hopping regime) close to the electrode. The charge hopping regime then transitions to interchain tunneling in the top region of the pSAM to connect the metal traps to the electrode [5]. This allows for the heterodeposition of more Co2+ onto the Co0 atom and the growth of the Co0 nanoparticle in the metal trap (Fig. 6). Voltammetry methods proved to be useful tools to describe these catalytic steps at the pilus metal traps during the electrodeposition of divalent cobalt. Similar approaches could be used to investigate the kinetics of binding, reduction and nucleation of other cationic metals. Indeed, the in vitro pSAM platform provides a rapid mean to screen the spectrum of metals that the nanowires can bind and reduce, and to identify electrochemical features of the binding and reduction kinetics. These findings can inform in vivo studies of specific biological reactions that Geobacter pili could mediate. For example, Geobacter bacteria can reduce chelated forms of Co3+ to Co2+ (aqueous) [3], but the reduction of Co2+ to Co0 [midpoint potential (E m) versus standard hydrogen electrode (SHE), − 0.28 V] is only thermodynamically possible if involving inner membrane carriers such as menaquinones (E m= − 0.26 V vs SHE [10]) [20]. The reduction of Co2+ in the cell envelope would lead to the accumulation of Co0 nanoparticles in the periplasm and the disruption of the cell envelope’s essential functions, eventually killing the cell. Yet, the pSAM studies suggest that the pili’s metal traps can bind and reductively precipitate Co2+, a strategy that could allow cells to mineralize the toxic cation extracellularly to prevent its permeation and non-specific reduction in the periplasmic space, as demonstrated for the uranyl cation [4]. Planar pilin assemblies could also provide the foundation of novel platforms for environmental sensing, the remediation of toxic cationic metals, and the reclamation of precious and rare metal cations. The recombinant production of pilin building blocks alleviates sustainable manufacturing concerns that currently limit the development of peptide and non-peptide building blocks for similar applications [19]. As shown herein, recombinant pilins can also be functionalized via genetic engineering to modulate their chemistry and interactions with support materials, a critical consideration for the manufacturing of peptide arrays with multiplex functionality. Genetic engineering can also be applied to modulate the conductivity of the peptides and their interactions with metals to tune the specificity of metal binding in order to functionalize electrodes with the sensitivity and selectivity required for environmental sensing applications [19]. Pilin-based biomaterials could also be engineered with the specificity needed to reclaim toxic and rare metals in the often complex environmental mixes and to synthesize metallic nanoparticles with defined properties and functionalities. The electrostatic nature of pilin–metal interactions suggests that many cationic metals could be immobilized using pilin-based materials, opening opportunities to harness their catalytic properties in the biosensing, bioremediation and bioreclamation of a wide range of metals. Acknowledgements This work was supported by Grant EAR1629439 from the National Science Foundation. G.R. also acknowledges support from the USDA National Institute of Food and Agriculture (Hatch project 1011745). Compliance with ethical standards Conflict of interest GR. is co-founder and Chief Scientific Officer of BioElectrica Inc., a startup that uses microorganisms to advance waste to energy technologies. References 1. Alleman KS , Weber K, Creager SE Electrochemical rectification at a monolayer-modified electrode J Phys Chem 1996 100 17050 17058 10.1021/jp962193b Google Scholar Crossref Search ADS WorldCat 2. Brownson DAC , Banks CE Brownson DAC, Banks CE Interpreting electrochemistry The handbook of graphene electrochemistry 2014 London Springer 23 77 Google Scholar Crossref Search ADS Google Preview WorldCat COPAC 3. Caccavo F Jr, Lonergan DJ, Lovley DR, Davis M, Stolz JF, McInerney MJ Geobacter sulfurreducens sp. nov., a hydrogen- and acetate-oxidizing dissimilatory metal-reducing microorganism Appl Environ Microbiol 1994 60 3752 3759 201883 Google Scholar Crossref Search ADS PubMed WorldCat 4. Cologgi DL , Lampa-Pastirk S, Speers AM, Kelly SD, Reguera G Extracellular reduction of uranium via Geobacter conductive pili as a protective cellular mechanism Proc Natl Acad Sci USA 2011 108 15248 15252 10.1073/pnas.1108616108 Google Scholar Crossref Search ADS PubMed WorldCat 5. Cosert KM , Steidl RJ, Castro-Forero A, Worden RM, Reguera G Electronic characterization of Geobacter sulfurreducens pilins in self-assembled monolayers unmasks tunnelling and hopping conduction pathways Phys Chem Chem Phys 2017 19 11163 11172 10.1039/c7cp00885f Google Scholar Crossref Search ADS PubMed WorldCat 6. Elgrishi N , Rountree KJ, McCarthy BD, Rountree ES, Eisenhart TT, Dempsey JL A practical beginner’s guide to cyclic voltammetry J Chem Educ 2018 95 197 206 10.1021/acs.jchemed.7b00361 Google Scholar Crossref Search ADS WorldCat 7. Feliciano GT , Steidl RJ, Reguera G Structural and functional insights into the conductive pili of Geobacter sulfurreducens revealed in molecular dynamics simulations Phys Chem Chem Phys 2015 17 22217 22226 10.1039/C5CP03432A Google Scholar Crossref Search ADS PubMed WorldCat 8. Hau HH , Gilbert A, Coursolle D, Gralnick JA Mechanism and consequences of anaerobic respiration of cobalt by Shewanella oneidensis strain MR-1 Appl Environ Microbiol 2008 74 6880 6886 10.1128/AEM.00840-08 2583509 Google Scholar Crossref Search ADS PubMed WorldCat 9. Holzle MH , Zwing V, Kolb DM The influence of steps on the deposition of Cu onto Au(111) Electrochim Acta 1995 40 1237 1247 10.1016/0013-4686(95)00055-J Google Scholar Crossref Search ADS WorldCat 10. Kishi S , Saito K, Kato Y, Ishikita H Redox potentials of ubiquinone, menaquinone, phylloquinone, and plastoquinone in aqueous solution Photosynth Res 2017 134 193 200 10.1007/s11120-017-0433-4 5645442 Google Scholar Crossref Search ADS PubMed WorldCat 11. Lampa-Pastirk S , Veazey JP, Walsh KA, Steidl RJ, Tessmer SH, Reguera G Thermally activated charge transport in microbial protein nanowires Sci Rep-Uk 2016 6 23517 10.1038/srep23517 Google Scholar Crossref Search ADS WorldCat 12. Love JC , Estroff LA, Kriebel JK, Nuzzo RG, Whitesides GM Self-assembled monolayers of thiolates on metals as a form of nanotechnology Chem Rev 2005 105 1103 1169 10.1021/cr0300789 Google Scholar Crossref Search ADS PubMed WorldCat 13. Mendoza-Huizar LH , Palomar-Pardave M, Robles J Quantum chemical study of the electrochemical reduction of the [Co(H2O)6]2+ and [Co(NH3)5(H2O)]2+ ions Electrochim Acta 2001 46 2749 2755 10.1016/S0013-4686(01)00504-7 Google Scholar Crossref Search ADS WorldCat 14. Mendoza-Huizar LH , Robles J, Palomar-Pardave M Nucleation and growth of cobalt onto different substrates Part I. Underpotential deposition onto a gold electrode J Electroanal Chem 2002 521 95 106 10.1016/S0022-0728(02)00659-9 Google Scholar Crossref Search ADS WorldCat 15. Mendoza-Huizar LH , Robles J, Palomar-Pardave M Nucleation and growth of cobalt onto different substrates Part II. The upd-opd transition onto a gold electrode J Electroanal Chem 2003 545 39 45 10.1016/S0022-0728(03)00087-1 Google Scholar Crossref Search ADS WorldCat 16. Palomar-Pardave M , Gonzalez I, Soto AB, Arce EM Influence of the coordination sphere on the mechanism of cobalt nucleation onto glassy carbon J Electroanal Chem 1998 443 125 136 10.1016/S0022-0728(97)00496-8 Google Scholar Crossref Search ADS WorldCat 17. Pawlowski J , Juhaniewicz J, Tymecka D, Sek S Electron transfer across α-helical peptide monolayers: importance of interchain coupling Langmuir 2012 28 17287 17294 10.1021/la302716n Google Scholar Crossref Search ADS PubMed WorldCat 18. Pletcher D , Greff R, Peat R, Peter LM, Robinson J Pletcher D, Greff R, Peat R, Peter LM, Robinson J Electrocrystallisation Intrumental methods in electrochemistry 2010 Cambridge Woodhead Publishing 283 316 Google Scholar Crossref Search ADS Google Preview WorldCat COPAC 19. Reguera G Harnessing the power of microbial nanowires Microb Biotechnol 2018 11 979 994 10.1111/1751-7915.13280 6201914 Google Scholar Crossref Search ADS PubMed WorldCat 20. Reguera G Microbial nanowires and electroactive biofilms FEMS Microbiol Ecol 2018 94 fiy086 10.1093/femsec/fiy086 Google Scholar Crossref Search ADS WorldCat 21. Reguera G , McCarthy KD, Mehta T, Nicoll JS, Tuominen MT, Lovley DR Extracellular electron transfer via microbial nanowires Nature 2005 435 1098 1101 10.1038/nature03661 Google Scholar Crossref Search ADS PubMed WorldCat 22. Soto AB , Arce EM, Palomar-Pardave M, Gonzalez I Electrochemical nucleation of cobalt onto glassy carbon electrode from ammonium chloride solutions Electrochim Acta 1996 41 2647 2655 10.1016/0013-4686(96)00088-6 Google Scholar Crossref Search ADS WorldCat 23. Speers AM , Schindler BD, Hwang J, Genc A, Reguera G Genetic identification of a PilT motor in Geobacter sulfurreducens reveals a role for pilus retraction in extracellular electron transfer Front Microbiol 2016 7 1578 10.3389/fmicb.2016.01578 5065972 Google Scholar Crossref Search ADS PubMed WorldCat 24. Zafferoni C , Cioncoloni G, Foresti ML, Dei L, Carretti E, Vizza F, Lavacchi A, Innocenti M Synergy of cobalt and silver microparticles electrodeposited on glassy carbon for the electrocatalysis of the oxygen reduction reaction: an electrochemical investigation Molecules 2015 20 14386 14401 10.3390/molecules200814386 6331807 Google Scholar Crossref Search ADS PubMed WorldCat Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. © Society for Industrial Microbiology 2019 This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © Society for Industrial Microbiology 2019
TI - Voltammetric study of conductive planar assemblies of Geobacter nanowire pilins unmasks their ability to bind and mineralize divalent cobalt
JO - Journal of Industrial Microbiology and Biotechnology
DO - 10.1007/s10295-019-02167-5
DA - 2019-10-01
UR - https://www.deepdyve.com/lp/oxford-university-press/voltammetric-study-of-conductive-planar-assemblies-of-geobacter-wRF59wur8s
SP - 1239
EP - 1249
VL - 46
IS - 9-10
DP - DeepDyve
ER -