In Situ Scanning Tunneling Microscopy and Microcantilever Investigations of yeast cytochrome c on gold

Proteins immobilized at surfaces are a research field of fundamental interest with important applications in bio- and nanotechnology, such as biosensors, biomimetics materials, and supramolecular structures. Knowledge of structure and functionality of protein layers is thus essential. In this report a comprehensive approach is used to assess such properties of Saccharomyces Cerevisiae yeast cytochrome c (YCC) on gold surfaces, both macroscopically and at the single-molecule level.

The gold-sulfur interaction was used to detect YCC with microcantilevers. YCC contains a nearsurface cysteine (Cys102) usable for immobilization to gold. Gas phase adsorption of 1-hexanethiol on gold-coated microcantilevers was first investigated to optimize surface cleaning for applications in liquid.

Cleaning by Aqua Regia (AR) and Oxygen Plasma (OP) was tested. AR provided a faster and larger surface stress signal than OP. 0.7 N/m of compressive stress was observed, while OP gave 0.25 N/m compressive stress. The signal for AR cleaned microcantilevers deteriorated, however, with successive cleanings, while the signal from OP cleaned cantilevers was stable.

OP cleaning was hence used for detection of YCC by gold-coated microcantilevers in buffer solution. 0.3 N/m of tensile stress evolved upon introduction of YCC into the microcantilever cell. The signal increased to 0.6 N/m when YCC physisorbed on the silicon nitride reference cantilever was flushed off with buffer.

The rate of surface stress change increased with increasing concentration of YCC in the range 5-100 µM. A range of other sulfur-containing molecules were investigated to elucidate further adsorption properties of both the gold-coated working cantilever and the reference cantilever. Cysteine gave tensile stress (0.2-0.3 N/m) when introduced to OP cleaned microcantilevers.

The stress signal could be returned to zero when the microcantilever cell was flushed, indicating physisorbed cysteine. Compressive stress (0.2-0.4 N/m) evolved, however, when cysteine was adsorbed on microcantilevers cleaned in Argon plasma (AP) and a fresh gold layer (FG) was deposited on the working cantilever before measurement.

The cysteine layer was stable towards flushing with buffer, indicating chemisorbed cysteine. Adsorption of 1-octadecanethiol onto AP cleaned microcantilevers with FG from a dilute solution in an ethanol/water mixture provided a tensile stress of –0.16 N/m possibly due to interaction with the reference cantilever or the mixed solvent.

Adsorption of 6-mercaptohexanol (MCH) onto AP cleaned microcanilevers with FG caused a fast concentration-independent tensile surface stress signal of –0.7 N/m followed by a slow compressive signal. The first signal resembled electrostatic binding to the reference cantilever. The second signal presumably adsorption to the gold layer.

YCC was imaged on Au(111) by in situ STM with molecular resolution, showing globular structures with average area 13-22 nm2. This is slightly larger than expected from the bulk size of YCC, presumably due to tip convolution. Coverages were up to 18 % or 8 × 1011 molecules⋅cm-2. Prolonged adsorption increased the molecular density to 18 × 1011 molecules⋅cm-2, but STM contrast was attenuated.

The YCC molecular layer was stable in the potential range –0.4 → -0.1 V vs. SCE. The contrast of YCC did not vary with potential in this range. The layers were, however, unstable towards potential scanning while imaging, with significant desorption within minutes. Capacitance (CAP) measurements, Linear Sweep (LSV), and Differential Pulse Voltammetry (DPV) were conducted for YCC at Au(111).

The capacitance of the YCC coated electrode was 9 µFcm-2, confirming that the electrode surface was covered with material (YCC) of low dielectric constant. A reductive desorption signal at app. -0.76 V indicated that YCC is bound to gold via sulfur. The charge transferred was 10 µCcm-2. The charge of a full monolayer desorbing via one gold-sulfur bond corresponds to 1.5 µCcm-2.

The excess charge might result from capacitive currents or binding of YCC via other sulfur residues than Cys102. YCC contains five sulfur atoms. Besides a weak DPV signal at 0.01 V from native YCC, DPV and CAP showed a peak at -0.45 V corresponding to a non-native conformation of YCC where the heme axial ligation to Met80 is lost.

X-ray Photoelectron Spectroscopy (XPS) showed peaks at 162.2 and 163.4 eV for both YCC and Horse heart cytochrome c (HHCC), which does not contain cysteine at position 102. Binding energies of 162.2 and 163.3 eV are fingerprints of gold-thiolate bonds. The HHCC spectrum contained a minor peak at 164.8 eV and the spectrum of YCC minor peaks at 161.2 and 164.6 eV.

Hence the sulfur peak is broader for YCC than for HHCC. Gold-thiolate peaks for HHCC could be explained by unfolding and exposure of internal sulfur residues induced by the UHV environment. YCC is structurally less stable than HHCC. The broader sulfur peak for YCC could therefore be explained by a more unfolded structure than HHCC.

The Au photoelectron signal was lower for YCC, indicating a higher coverage of YCC than HHCC. Broad characterization of YCC layers on gold with regards to adsorption kinetics, functionality, structure, and binding mode has thus been performed. YCC was found to be bonded to gold via sulfur and to exist partially in a non-native but functionable state.