Respiratory and photosynthetic electron transfer chains are dependent on vectorial electron transfer through a series of redox proteins. Examples include electron transfer from NapC to NapAB nitrate reductase in Paracoccus denitrificans and from CymA to Fcc3 (flavocytochrome c3) fumarate reductase in Shewanella oneidensis MR-1. In the present article, we demonstrate that graphite electrodes can serve as surfaces for the stepwise adsorption of NapC and NapAB, and the stepwise adsorption of CymA and Fcc3. Aspects of the catalytic properties of these assemblies are different from those of NapAB and Fcc3 adsorbed in isolation. We propose that this is due to the formation of NapC–NapAB and of CymA–Fcc3 complexes that are capable of supporting vectorial electron transfer.
The stepwise adsorption of oppositely charged molecules on to surfaces provides a simple means to assemble layered architectures. The resulting assemblies can support vectorial electron transfer when they contain redox proteins presenting complementary surfaces with regard to their charge, hydrophobicity and topography. When the layered architecture is assembled on an electrode, it may be exploited for electronic devices . As an example, amperometric biosensors can be developed by including an enzyme selective for redox transformation of a specific analyte. There is also the opportunity to use dynamic electrochemistry of layered architectures to provide fundamental insight into electron transfer within the protein complexes that support respiration and photosynthesis [2–4].
In the present article, we describe experiments that explored graphite electrodes as surfaces for assembling electroactive complexes from components of two microbial respiratory electron transfer chains, namely periplasmic nitrate reduction in Paracoccus denitrificans and periplasmic fumarate reduction in Shewanella oneidensis MR-1. Tetrahaem cytochromes of the NapC/NirT superfamily transfer the electrons from quinol oxidation to the terminal reductase in both of these pathways . NapC transfers electrons from ubiquinol oxidation to the NapAB nitrate reductase (Figure 1A). CymA transfers electrons from menaquinol oxidation to the Fcc3 (flavocytochrome c3) fumarate reductase. Although no structure is available for NapC or CymA at present, insight into the likely fold and arrangement of cofactors is available from the structure of another member of the NapC/NirT superfamily, namely NrfH . A periplasmic globular domain binds four haems, whereas a single α-helix anchors the proteins to the inner membrane (Figure 1A). The haems are predicted to support facile intraprotein electron transfer since the nearest neighbours are in close proximity (edge-to-edge distance <6 Å; 1 Å=0.1 nm).
Illustration of (A) the pathway of electron transfer from ubiquinol to nitrate via NapC and NapAB in P. denitrificans, and (B) the strategy to assemble an electroactive complex of NapCsol and NapAB on graphite electrodes
Nitrate reduction by stepwise adsorption of NapC and NapAB
NapAB is a heterodimer for which a molecular description is provided by a crystal structure of the homologous enzyme from Rhodobacter sphaeroides . The catalytic NapA subunit houses a molybdopterin active site and an [Fe–S] cluster that supports electron transfer from the dihaem-containing NapB. The cofactors in NapB are positioned close to the surface of the NapAB heterodimer, whereas those in NapA are deeply buried. As a consequence, NapB is proposed to dock with NapC for intersubunit electron transfer. To explore the possibility of co-adsorbing NapAB and NapC in a complex supporting vectorial electron transfer, we took advantage of the previously described route to access the tetrahaem-containing periplasmic domain of NapC, termed NapCsol . The water-soluble NapCsol can be purified more readily than full-length NapC, but should present the same surface for binding NapAB.
Figure 2(A) illustrates a typical cyclic voltammogram measured at pH 7 with a freshly polished basal plane pyrolytic graphite electrode that has been exposed to NapCsol. The data are presented after subtraction of the response obtained from an identical experiment in which the electrode had not been exposed to NapCsol. Clear peaks are resolved for oxidative (positive) and reductive (negative) processes arising from redox transformation of NapCsol between +100 and −400 mV relative to the SHE (standard hydrogen electrode). The peaks were retained after rinsing the NapCsol-coated electrode and placing it in a fresh solution of buffer electrolyte. As a consequence, the peaks can be attributed to electroactive molecules of NapCsol that are securely adsorbed on to the electrode.
Representative baseline-subtracted cyclic voltammogram (0.02 V·s−1) of a graphite electrode in 50 mM Pipes and 100 mM NaCl (pH 7.0) at 20°C after exposing the electrode to 5 μl of 33 μM NapCsol in the same buffer (grey fill)
The shoulders on the flanks of the voltammetric peaks indicate that the haem reduction potentials are spread sufficiently across the electrochemical potential domain to distinguish contributions from individual centres. Under the experimental conditions employed to study NapCsol, an isolated centre undergoing a reversible single-electron (n=1) redox transformation is predicted to give rise to peaks with a half-height width of approximately 90 mV . In the voltammetry of the tetrahaem cytochrome NapCsol, there is no obvious indication of co-operative, i.e. n=2, electron transfer activity that would produce peaks of four times the height and half the width of n=1 centres . As a consequence, and assuming all haems within a molecule of NapCsol can equilibrate with the applied potential, the peaks arising from NapCsol were fitted to the sum of contributions from four n=1 centres using processes described previously . This produced a good description of the experimental peaks (Figure 2). Averaging the peak potentials for the n=1 contributions to the oxidative and reductive peaks in order of decreasing potential gave reduction potentials for the four haems in NapCsol of −35, −90, −145 and −222 mV (all ±10 mV). The reduction potentials compare well with the values of −56, −181, −207 and −235 mV published previously for the haems in solutions of NapCsol at pH 8 and that were determined by mediated potentiometric titration monitored by electronic absorbance spectroscopy . Reduction potentials of −55, −120, −160 and −240 mV (all ±10 mV) were defined by cyclic voltammetry of the adsorbed protein at pH 8. Thus the slightly lower reduction potentials displayed at pH 8 than at pH 7 suggest stabilization of the reduced haems by protonation of NapCsol.
The areas (coulombs) of the oxidative and reductive voltammetric peaks were comparable and independent of scan rate from 0.02 to at least 100 V·s−1. These areas describe the exchange of 1.9 μmol of electrons between NapCsol and a geometric electrode area of 0.071 cm2 and which corresponds to a surface coverage of approximately 7 pmol·cm−2 NapCsol. Taking NapCsol to have dimensions comparable with those of the NrfH globular headgroup, this coverage is close to that expected on formation of a complete monolayer (12 pmol·cm−2 calculated assuming a planar electrode surface). To assess whether the NapAB nitrate reductase would co-adsorb with NapCsol, the NapCsol-coated electrode was exposed briefly to a solution of NapAB nitrate reductase (70 μM in 50 mM Pipes and 100 mM NaCl, pH 7.0) and then rinsed before returning it to the electrochemical cell. With a scan rate of 0.02 V·s−1, the cyclic voltammetry was essentially unchanged from that of NapCsol alone. However, a strong catalytic reduction wave was observed when nitrate was introduced into the experiment (Figure 3A). NapCsol is unable to reduce nitrate and so it can be concluded that NapAB co-adsorbs with NapCsol.
Representative catalytic waveshapes describing nitrate reduction by NapAB
Our experiments demonstrate catalytic nitrate reduction by NapAB adsorbed on to a NapCsol-coated electrode. The catalytic wave has a sigmoidal shape and the first derivative of catalytic current with respect to applied potential overlays well the peaked voltammetric response displayed by NapCsol alone (Figure 3B). This property is clearly different from the catalytic response from NapAB in isolation which shows a characteristic lowering of activity when the potential is lowered beyond −0.15 V  (Figure 3A). These observations suggest that NapAB co-adsorbs with NapCsol in a manner that supports vectorial electron transfer from the electrode to NapCsol and then to NapAB. When the scan rate is raised from 0.02 to 0.5 V·s−1, the magnitude of the catalytic response at −0.45 V decreased by approximately 40% for NapAB in isolation and approximately 90% for NapAB co-adsorbed with NapCsol (Supplementary Figure S1 at http://www.biochemsoctrans.org/bst/041/bst0411249add.htm). This was unexpected as the shape and area of the non-catalytic voltammetric response from NapCsol are independent of scan rates up to 2 V·s−1 (Figure 2, inset). If NapCsol and NapAB form a stable complex with interprotein electron transfer described by a single bimolecular rate constant, theory predicts that, at sufficient over-potentials, the catalytic current will be independent of scan rate. To study the electrochemical behaviour of layered protein assemblies further, we explored the homologous system of CymA–Fcc3.
Fumarate reduction by stepwise adsorption of CymA and Fcc3
Previously, we have shown that CymA adsorbs as a stable electroactive layer on graphite as well as modified gold electrodes [12,13]. On graphite electrodes, cyclic voltammetry of CymA reveals peaks with an area that corresponds to adsorption of approximately 65% of a close-packed monolayer assuming a planar electrode surface. When the CymA-coated electrode is placed in a solution of 0.07 μM Fcc3 and 2 mM fumarate, a catalytic reduction wave appears with a magnitude that increases over seven cycles between 0.2 and −0.6 V at 0.02 V·s−1. Adsorption of Fcc3 on to the CymA-coated electrode was confirmed when the catalytic response was maintained after the electrode had been rinsed and the electrode placed in a solution lacking Fcc3, but containing 2 mM fumarate (Figure 4A).
Catalytic voltammetry of Fcc3 co-adsorbed with CymA
The catalytic waveshape arising from Fcc3 adsorbed in isolation on the electrode [14,15] is very similar to that displayed by Fcc3 co-adsorbed with CymA. As a consequence, this feature of the catalytic response provides no immediate evidence for vectorial electron transfer from CymA to Fcc3. Nevertheless, the scan-rate-dependence of the catalytic current magnitude arising from these systems differed in a manner that may support the formation of a complex in which CymA can pass electrons to Fcc3. The magnitude of the catalytic response from Fcc3 in isolation was independent of scan rate to at least 1 V·s−1 (Figure 4B). That from Fcc3 co-adsorbed with CymA was reduced by over 90% at 0.2 V·s−1 (Figure 4B). This is despite isolated CymA giving rise to peaks with areas and shapes that are independent of scan rate from 0.01 to 0.5 V·s−1.
Discussion and conclusions
Catalytic substrate reduction is detected from NapAB co-adsorbed with NapCsol and from Fcc3 co-adsorbed with CymA. The different catalytic behaviour displayed by the co-adsorbed enzymes when compared with the enzymes in isolation is proposed to arise from the formation of complexes between the co-adsorbed proteins that support vectorial electron transfer and that may mimic the complexes that underpin respiratory electron transfer in P. denitrificans and S. oneidensis MR-1.
The scan-rate-dependence of the catalytic responses displayed by these complexes was unexpected. One possible explanation is that a redox-driven conformational change in the electrode supported assemblies limits the rate of electron transfer from NapCsol to NapAB and from CymA to Fcc3. Whether the same behaviour is displayed by these complexes during respiratory electron transfer remains to be established. It is possible that neither NapCsol nor CymA adsorb on the electrode to present exactly the same properties as the proteins displayed on the surface of the bacterial inner membrane. We have recently explored the possibility that Fcc3 may form an electrocatalytically active complex with CymA that is contained in an electrode-supported bilayer . Our initial experiments demonstrate that this allows CymA and Fcc3 to form complexes capable of supporting vectorial electron transfer and it will be of interest to establish whether these assemblies display evidence for redox-dependent conformational change.
Bioenergetics in Mitochondria, Bacteria and Chloroplasts: Third Joint German/UK Bioenergetics Conference, a Biochemical Society Focused Meeting held at Schloss Rauischholzhausen, Ebsdorfergrund, Germany, 10–13 April 2013. Organized and Edited by Fraser MacMillan (University of East Anglia, Norwich, U.K.) and Thomas Meier (Max Planck Institute of Biophysics, Frankfurt am Main, Germany).
This research was supported by the Biotechnology and Biological Sciences Research Council [grant numbers BB/G009228, BB/E0219991 and BB/D5230191], the University of East Anglia and the U.S. Department of Energy, Office of Biological and Environmental Research (BER) through the Subsurface Biogeochemical Research (SBR) Program. The paper represents a contribution from the Pacific Northwest National Laboratory (PNNL) SBR SFA. PNNL is operated for the Department of Energy by Battelle.