Shallow conductance decay along the heme array of a single tetraheme protein wire

Abstract

Multiheme cytochromes (MHCs) are the building blocks of highly conductive micrometre-long supramolecular wires found in so-called electrical bacteria. Recent studies have revealed that these proteins possess a long supramolecular array of closely packed heme cofactors along the main molecular axis alternating between perpendicular and stacking configurations (TST = T-shaped, stacked, T-shaped). While TST arrays have been identified as the likely electron conduit, the mechanisms of outstanding long-range charge transport observed in these structures remain unknown. Here we study charge transport on individual small tetraheme cytochromes (STCs) containing a single TST heme array. Individual STCs are trapped in a controllable nanoscale tunnelling gap. By modulating the tunnelling gap separation, we are able to selectively probe four different electron pathways involving 1, 2, 3 and 4 heme cofactors, respectively, leading to the determination of the electron tunnelling decay constant along the TST heme motif. Conductance calculations of selected single-STC junctions are in excellent agreement with experiments and suggest a mechanism of electron tunnelling with shallow length decay constant through an individual STC. These results demonstrate that an individual TST motif supporting electron tunnelling might contribute to a tunnelling-assisted charge transport diffusion mechanism in larger TST associations.

Fig. 1. STC crystal structure (left panel) from PDB ID 1M1P indicating the position of the solvent-exposed engineered cysteine (Cys) residue at position 87. Right panel shows the TST-pairing array of the four heme cofactors in the STC structure. Each heme cofactor is axially coordinated by two histidine residues (green) and covalently attached to the peptide backbone (green ribbon) via internal (non-solvent exposed) Cys residues (yellow).

Fig. 2. (a) 400 × 400 nm² STM image of Cys87-modified STC proteins on Au surface. Each bright spot in the image corresponds to an individual STC. Characteristic triangulated atomically flat Au(111) terraces are >200 nm wide. Colour legend on the right spans from 0 to 1.2 nm. (b) Semi-log scale conductance (G) histogram accumulating 1350 selected G(z) traces (representative ones in the inset) at an applied 200 mV bias voltage. Individual Gaussian fits of the peaks (coloured solid lines) are used to extract the maxima values corresponding to the average conductance of each of the four features. The dotted line (----) is a cumulative plot including all the Gaussian peaks for visual guidance. Figure inset shows illustrative individual G(z) traces displaying plateau features used to build the 1D histogram. The X-axis scale bar at the left-bottom corner corresponds to 0.6 nm. (c) Pictorial representation of the STM break-junction process showing two possible scenarios leading to the low G (peak 1 in (b)) and high G (peak 3 in (b)) involving 4 and 2 hemes in the pathway, respectively, where STM tip asperities (represented as yellow balls) contact different points of the protein surface during different pulling cycles of the anchored (via Cys(thiol)) STC on the Au substrate.

Fig. 3. Representative G(t) blinking traces with subtracted background tunnelling current obtained at different electrode–electrode separations of the STM nanoscale gap. Insets: (a) conductance histograms accumulating blinks at the different electrode–electrode gap separations of 2.0 ± 0.3, 2.5 ± 0.3, 3.3 ± 0.3 and 4.0 ± 0.3 nm sorted by colours (see graph legend) represented in a common graph (error bars in the gap separations are calculated from the error in the experimental beta decay determination through the empty junction (see Fig. S4†)). (b) Deconvolution of the cumulative histogram of all 1686 traces without gap distance classification. The applied bias voltage was 200 mV.

Fig. 4. (a) Solvent accessible surface area (SASA) representation of the area of exposed hemes 1–4 (in red). (b) Representation of step length calculated from dynamic STM-BJ G(z) plots versus the percentage of exposed heme area calculated using SASA of the protein crystal structure (1m1b pymol). (c) Conductance decay plot representing the STC conductance values extracted from the dynamic STM-BJ histograms against the protein crystallographic distances (Fig. 1) from the bottom contact point (Cys87) to each heme 1–4 (calculated using pymol 1m1q). Inset: semi-log representation of G/G₀versus Cys87-to-heme distance, fitted to a straight line (blue) yielding β = 2.1 ± 0.4 nm⁻¹. (d) Conductance decay plot representing STC conductance extracted from the static STM-BJ histograms against the electrode–electrode junction gap separation. Inset: semi-log representation of G/G₀versus the electrode–electrode gap separation, fitting a straight line (blue) which yields a decay constant of b = 2.3 ± 0.1 nm⁻¹. The two filled (yellow) squares are computed DFT conductance values for two averaged standing and lying junction configurations (see ESI Fig. S6†). All error bars in conductance values were extracted from the FWHM of the fitted Gaussian peaks in histograms. Error bars in the electrode–electrode gap separations were extracted from the error in the experimental β determination through the empty gap (see calibration in ESI Fig. S4†).

Fig. 5. (a) Deconvoluted Gaussian functions extracted from Fig. 3b inset represented in different colours as per each identified gap separation against the computed conductance values from the different simulated structural configurations represented as red and green arrows, respectively, for standing STCs (4 hemes probed in the junction) S1(ox) & S2(ox) and lying STCs (2 hemes probed in the junction) L1(ox) & L2(ox). The S1 and L1 computed structures are included in the inset. (b) Representation of a successful I–V collection on a blink fulfilling the experimental sequence (1–4): (1) blink recognition at the selected electrode–electrode gap separation, (2) triggered voltage ramp on the blink, (3) identifying the original blink conductance level is recovered, and (4), forced pulling curve to identify protein junction breakdown. Inset: pictorial representation of open and closed junctions. (c) I–V characteristics measured at different electrode–electrode gap separations of 4.0 ± 0.3, 3.3 ± 0.3 and 2.5 ± 0.3 and 2.0 ± 0.3 nm, shown as bands representing the total current dispersion per each gap distance. Overlayed solid lines represent computed I–V curves of STC junctions using DFT, where red lines represent S1(ox) & S2(ox) and green lines represent L1(ox)& L2(ox). Error bars in the electrode–electrode gap separations were extracted from the error in the experimental determination of β through the empty gap (see calibration in ESI Fig. S4†).