"There has long been this breadcrumb trail of evidence that proteins behave unusually electronically," explains Lindsay, director of the Biodesign Center for Single Molecule Biophysics at Arizona State University. "All the experiments you can shoot down because you don’t know the state of the protein or how many proteins you have there – here, for the first time, we trap a single protein in a well defined gap and in a condition in which the protein is native."

Lindsay worked alongside researchers at Arizona State University in the US and Eötvös Loránd University in Hungary to characterize the proteins both using a scanning tunnelling microscope (STM) similar to other groups, as well as with a "fixed-gap device" junction developed in work on DNA sequencing. Characterizing proteins by STM raises several issues because the precise chemistry and geometry of the STM tip are not known, and the native environment of these proteins differs greatly from a vacuum, where the physics is well established. However, Lindsay and his colleagues found that their less error-prone fixed-gap device also gave conductances several orders of magnitude greater than expected, and that they fluctuated.

"We were shocked," says Lindsay. "We shelfed these experiments while looking for what must be the mistakes in them, but we couldn’t find any mistakes." Reports of similar behaviour in other proteins persuaded Lindsay that the results may have genuine significance. In particular, a report by Gábor Vattay at Eötvös Loránd University – also an author on this current work – identified quantum critical point behaviour in several functional proteins, prompting Lindsay on to further analysis that revealed the same quantum critical point statistics in their data.

What possible evolutionary pressure might have advocated proteins that exist at this quantum critical point remains a mystery. However, Lindsay does note that the large conductances were only observed above a threshold voltage of around 50–100 mV – a range of values that coincide with the cell membrane potential arising from ion differences inside and outside cells.

The fixed junction device

Lindsay is also a world expert in DNA sequencing, for which he and his group devised recognition tunnelling technology, currently being licensed. Here, tunnelling electrodes are functionalized with specific chemical linkage molecules that recognize and trap the target molecule. "We realized that the recognition tunnelling approach is physical and so is applicable to any molecule including proteins," says Lindsay.

The researchers’ fixed junction device comprises a 10 nm thick Pd electrode on a SiN substrate, followed by 4 to 5 nm of Al2O3 deposited by atomic layer deposition, and then a second 10 nm-thick Pd electrode. An ion-beam-milled gap in the sandwiched electrode structure provides the junction.

The protein

The researchers investigated the αVβ3 integrin protein – "a passive blob" in electrochemical terms since it has no redox centres. Ionization from redox interactions could lead to additional conductance paths that would complicate the experiment.

Functionalizing the electrode with an "RGD" peptide, which is specific to the αVβ3 protein, led to a distinct two-level current signal and conductances several orders of magnitude higher than expected. Functionalizing with a less selective peptide, or measuring an off-target protein, led to far fewer signals, indicating that the protein was native. The signals were also observed in human serum – a mucky substance with a lot of background noise, suggesting the approach may be useful for single molecule detection.

Future work

The αVβ3 protein has only one binding site for the RGD peptide, and the results suggest the one strong contact that this provides is crucial for detecting the two-level electronic signals. The weak contact on the other side may play a role in the fluctuations. However, the number of previous STM experiments yielding supporting results, and the strong fit of these robust fixed-junction device measurements with the modified Poisson distribution typical of a quantum critical point, remains convincing.

Lindsay and his collaborators are now in the process of testing other proteins for the same behaviour. Time will tell whether this quantum critical point behaviour is a property of all functional proteins, and what the evolutionary significance of this might be.

Full details are reported in Nano Futures.