PNNL

Abstract

Electron transfer in proteins, which are considered electronic insulators1,2, occurs through tunneling or hopping a few nanometers via inorganic cofactors1,2. However, the common soil microorganism Geobacter sulfurreducens transfers electrons over hundreds of micrometers, to insoluble electron acceptors3 or syntrophic partner species4, using protein filaments called pili, in order to survive in environments that lack membrane-permeable electron acceptors such as oxygen5. The conductivity of wild-type pili exhibits temperature dependence5,6 similar to that of disordered metallic polymers7. However, metallic conductivity is considered impossible in proteins due to lack of periodicity in protein structure, thermal fluctuations, and low conductivity values1,2. Truly metallic conductivity7 has never been observed in proteins. Here we report the hallmarks of classic metallic conductivity in pili. Using the 4-electrode method on individual wild-type pili, intrinsic conductivity measurements exhibited metallic behavior at physiological temperatures. Surprisingly, tryptophan-substituted pili showed room-temperature conductivities > 500 S/cm that increased exponentially upon cooling to 2 K, which is consistent with the theory developed for quasi-one-dimensional metallic polymers8,9. Using terahertz time-domain spectroscopy, we measured frequency-dependent optical conductivity of pili that was also consistent with metal physics. The electron decay length in pili was 1000 times lower than other proteins2 or DNA10, comparable to that of synthetic metallic nanostructures11. X-ray diffraction studies revealed that improved metallic nature of pili was associated with increased 3.6 Å periodicity, consistent with higher p-stacking between aromatic residues. This results in elongation of the effective conjugation length and longer mean free path7. We conclude that the electronic structure of proteins can be systematically tuned from insulators to semiconductors and metals, using rational design of amino acid sequence. This may allow development of next-generation semiconducting technology using synthetic biology.