Protein Binding and Orientation Matter: Bias-Induced Conductance Switching in a Mutated Azurin Junction

We observe reversible, bias-induced switching of conductance via a blue copper protein azurin mutant, N42C Az, with a nearly 10-fold increase at |V| > 0.8 V than at lower bias. No such switching is found for wild-type azurin, WT Az, up to |1.2 V|, beyond which irreversible changes occur. The N42C Az mutant will, when positioned between electrodes in a solid-state Au–protein–Au junction, have an orientation opposite that of WT Az with respect to the electrodes. Current(s) via both proteins are temperature-independent, consistent with quantum mechanical tunneling as dominant transport mechanism. No noticeable difference is resolved between the two proteins in conductance and inelastic electron tunneling spectra at <|0.5 V| bias voltages. Switching behavior persists from 15 K up to room temperature. The conductance peak is consistent with the system switching in and out of resonance with the changing bias. With further input from UV photoemission measurements on Au–protein systems, these striking differences in conductance are rationalized by having the location of the Cu(II) coordination sphere in the N42C Az mutant, proximal to the (larger) substrate-electrode, to which the protein is chemically bound, while for the WT Az that coordination sphere is closest to the other Au electrode, with which only physical contact is made. Our results establish the key roles that a protein’s orientation and binding nature to the electrodes play in determining the electron transport tunnel barrier.


Introduction
This document contains additional supporting experimental protocols, data, figures, and discussions that are details of supplementary information to the main text, and support for the presented results. The information presented is: Wild type Azurin was isolated from P. aeruginosa using the published procedure of Ambler and Brown 1 , 2 except that the azurin elution buffer for the CM-cellulose column was pH of 5.0 instead of 4.6. The mutated azurin N42C Az was constructed and expressed following the published procedures 3, 4 . It was isolated in its apo-form in the presence of dithiotreitol in order to prevent binding to other thiol containing molecules. To produce the holo-N42C Az, a slight excess of Cu(II) together with a five-fold ferricyanide were added to the apo-form. That yielded the typically bluecolored, C42 disulfide linked dimers which were employed for producing the monolayers. hrs, Au nanowires of ~5-6 µm long were formed in the membrane. These metal replicas are then released into the solution by dissolving the alumina membrane in an aqueous base of KOH and suspending them in water, which served as the transport medium during the assembly process.
Protein monolayers formation on Au surface: Au microelectrodes (1 µ thick), were fabricated on top of a Si wafer by using photolithography, yielding a substrate that contains 260 devices. The Au electrode surfaces were initially cleaned by bath sonication in acetone/ethanol (3 min each) Supporting information for "Protein binding and orientation matters: Biased-induced conductance switching in a mutated azurin junction" S3 and thoroughly rinsed in Milli-Q (18 MΩ) water. The Au surface is activated using ozone (UVOcleaner Model No: 3422A-220) for 10 mins, followed by treatment with hot ethanol for 20 mins.
The activated Au substrates were then rinsed with water and immediately used for incubating with the examined protein solution. Az monolayers were prepared by immersing substrates in a 1 mg/mL solution of WT Az or N42C Az mutant for 3 hrs followed by rinsing with clean water, before drying with a mild nitrogen stream. After monolayer formation, Au nanowires were dielectrophoretically trapped to complete the circuit/junction 6 . During the dielectrophoretic trapping process Au nanowires form bridges between the two micro-electrodes. Empirically it has been found that one of the two nanowire-electrode bridge contact is always shorted (See Fig. 2A). The likely cause for this shorting is the force by which the nanowire is pulled into the gap between the electrodes, which is exerted on the first electrode with which it makes contact. This force is apparently sufficient to destroy the monolayer on that electrode. The wire's other end then performs a "soft landing", preserving the monolayer on that electrode. Such nanowires can be observed using a standard optical microscope, allowing quick marking of useful junctions, i.e., those where the nanowire is clearly bridging the electrodes (see The final architecture of all measured junctions contains only a single Au nanowire between the two gold microelectrodes. Cases of two or more nanowires bridging between the two contact pads were rather rare and could be easily detected using an optical microscope, prior to the electronic transport measurements.

Current-voltage and conductance-voltage Inelastic Electron Tunneling Spectroscopy, IETS:
The samples were loaded on an electrically floating sample stage and placed in a cryogenic probe station (Lakeshore TTPX). Current-voltage (I-V) measurements were performed to assess the transport efficiency across the junctions, using a Keithley 6430 sub-Femtoamp Source-Meter Unit, with a voltage scan rate of 20 mV/s. For all measurements, the substrate (bottom electrode of the junction) is grounded, while the AuNW electrode was biased, in a consistent manner (to ensure the bias polarity was in the same direction for all measurements). In each sets of experiments, scans were acquired that started and ended at 0 V (i.e., voltage sweep was from 0 to -1.2 V, -1.2 V to 1.2 V, 1.2 V to 0 V). This was done in order to check for a possible effect of the polarity of the initial voltage on the measured I-V characteristic. The measurements were done in 10 -4 to 10 -6 mbar vacuum, with the higher vacuum at lower temperatures.
Due to the mesoscopic nature of the junction, the variation in absolute currents was large (approximately 2 orders of magnitude). However, about a quarter of the junctions (from 30-40 successfully trapped nanowires per preparation), measured by I-V revealed high current magnitude. Such junctions that were identified as partially shorted were not analyzed in the current-voltage distribution.
Supporting information for "Protein binding and orientation matters: Biased-induced conductance switching in a mutated azurin junction" S5 For conductance and IETS measurements, the system was cooled to ~10K (p=1×10 -6 mbar). First and second derivatives of the current were recorded as the first and second harmonic signals, respectively, using lock-in techniques (two Stanford 830 lock-in amplifiers) with a DC The UV/Vis absorption spectrum of the N42C Az dimer is essentially identical to that of the wild type Az 8 , λmax 630 nm which indicates that no structural changes around the copper site have occurred. Figure S4: Overlay of the typical LMCT band of the type 1 Cu(II) site of the N42C Az mutant with that of the WT Az.

XPS results:
The presence and characteristics of the N42C Az mutant protein on the Au surface was verified and examined by XPS measurements. High resolution XPS spectra of C 1s, N 1s, Cu 2p

Polarization Modulation Infra-Red Reflection-Absorption Spectroscopy (PM-IRRAS):
In addition to XPS, PM-IRRAS was used to find evidence for protein presence on Au surface. Figure S6A and S6B shows the PM-IRRAS spectrum of N42C Az, at 2750-3050 cm -1 and 1200-1800 cm -1 wavelength ranges, respectively. The bands associated with the different functional groups of the monolayer are clearly visible in the spectrum. The bands located at 2850, 2930 and 2970 cm -1 are assigned to the symmetric (ʋ ss CH2), asymmetric (ʋ as CH2) and asymmetric (ʋ as CH3) stretching vibrations of the CH2 and CH3 groups. The bands at 1677 and 1547cm -1 are attributed to the amide I and amide II modes of the group, respectively.  were very stable till ~0.5 V (for statistics see Fig. 3 main text). Once the bias is increased above 0.75 V the stability drops quickly. Therefore, after acquiring the results till 0.5 V, the bias was increased from 0.5 V to 0.75 V, and the conductance and I-V data till 0.75 V were collected. When the junctions were still stable, we moved to 1.0 V and finally to 1.2 V. WT Az junctions show high stability at higher bias compared to those of N42C mutant. This could be due to the high current/conductance value reached via the mutant Az compared to the WT Az at a given bias.

Overlay of the IETS spectrum of WT Az and N42C Az:
Supporting information for "Protein binding and orientation matters: Biased-induced conductance switching in a mutated azurin junction" S12 Fig. S9 shows the current-voltage and dI/dV-Voltage plots of two different WT Az junctions, where no conductance switching behavior is observed at high bias. The results presented here (Figs. S8 and Fig. S9) are similar to those presented in the main text Fig. 4, i.e., conductance switching behavior is observed only for N42C Az mutant.   Supporting information for "Protein binding and orientation matters: Biased-induced conductance switching in a mutated azurin junction" S14 As shown in Fig. S12, the UPS intensity logarithmic results yield much smaller HOMO onsets than those derived from the linear intensity vs. Binding Energy UPS plots shown in Fig. 5B of the main text. Most importantly is that of the N42C Az surface is ~1 eV closer to the electrode Fermi level than that of the WT Az, one which agrees with the result of the linear UPS plot presented in the main text. Due to the expected vibrational broadening of the HOMO, there will not be one correct value, but the log plot is probably the more representative one for tail states. In earlier work based on our Azurin transistor measurements 9 , we found that the LUMO of the WT Az is slightly above the electrode Fermi level, which could fit with the end of the tail, based on the estimate derived from Fig. S12. Naturally, the energetics can be further affected by the top contact, especially for the WT Az.  Supporting information for "Protein binding and orientation matters: Biased-induced conductance switching in a mutated azurin junction" S15

I-V curve fitting analysis:
We fit the experimental I-V curves of both the WT Az and the N42C Az mutant from Fig. 3A and 3D (main ms.) to third order cubic polynomials and then extract the required parameters using a simplified one-energy level Landauer model 10 (see Fig. S14). Using this method, we extracted the electrode-molecule coupling strength values (Г), the effective energy barrier (ε0, in eV), effective conductance at zero bias (Geq= dI/dV) and a dimensionless symmetry parameter (α = 0 for a symmetrical I−V plot). All the experimental I-V curves can be fit to third-order polynomials with a high correlation (R 2 ~ 1). See Table-1    The parameter values extracted from fitting the I-V curves of both the WT and the mutant N42C Az to a third order polynomial.
The electronic coupling (Г) and the energy barrier values (ε0) that we find, give us an idea of the effective energy scales involved in the ETp process. Comparing the results for the WT Az and N42C Az mutant shows that over the given bias range, the Az mutant 's coupling to the electrodes is higher than that of the WT. The calculated effective barrier off both the WT Az and the N42C Az mutant are > 0.5 V, consistent with that, below 0.5 V ETp via both the WT Az and the N42C Az is by off-resonant tunneling transport (further supported experimentally by the temperature Supporting information for "Protein binding and orientation matters: Biased-induced conductance switching in a mutated azurin junction" S16 independence of the current, Fig. 1 main ms.). In general, though, in off-resonant tunneling regime, both the WT and the mutant behaves similarly, with a slightly higher current and the conductance values for the mutant than the WT Az, which can be traced due to better coupling to the electrodes.