The Archaellum of Methanospirillum hungatei Is Electrically Conductive

We chose the methanogen Methanospirillum hungatei for the initial search for an electrically conductive archaellum (e-archaellum) because M. hungatei is capable of reducing extracellular electron acceptors (14), archaellum expression is readily induced in M. hungatei (15), and a cryo-EM (3.4-Å) structure of the archaellum is available (13).

Initial screening of the relative conductivity of diverse bacterial pili is typically determined with conductive atomic force microscopy in which samples are deposited on a conductive surface and a conductive tip serves as a translatable top electrode (16–19). Therefore, 100 µl of a culture of M. hungatei grown in low-phosphate medium to induce archaellum expression (15) was drop-cast onto highly oriented pyrolytic graphite (HOPG), washed, dried, and then equilibrated at 40% relative humidity for conductivity measurements. This process was designed to mimic physiologically relevant conditions by avoiding chemical alteration of the archaellum structure and determining conductivity of hydrated archaella.

Cells with a polar archaellum with the expected height of 10 nm (13) were readily detected with topographic imaging in contact mode (Fig. 1a, b, and d). Conductive imaging demonstrated that the archaellum was electrically conductive (Fig. 1c to e; see also Fig. S1 and S2 in the supplemental material). Point-mode current-voltage (I-V) spectroscopy revealed a linear-like response with currents that were higher than at the same voltage with G. sulfurreducens e-pili prepared in the same manner (Fig. 1e). The pili of G. sulfurreducens strain Aro-5, which produces pili specifically designed for low conductivity (20, 21), exhibited very low currents at the same voltages (Fig. 1e). Conductance estimated from the linear portion of the I-V curves yielded conductance estimates of 16.9 ± 3.9 nS (mean ± standard deviation; n = 9; three independent points on three separate archaella; 8,000 points of measurement taken for each experimental I-V curve comprised of quadruplicate 0.6-V-bias sweeps) for the archaella, 4.5 ± 0.3 nS for the wild-type G. sulfurreducens pili, and only 0.004 ± 0.002 nS for the Aro-5 pili. The estimated conductance of the wild-type G. sulfurreducens pili was similar to the values found in previous studies that employed a comparable measurement technique (16).

These results demonstrated that the M. hungatei archaellum is conductive and suggest that a search for electrically conductive protein filaments in other Archaea as well as the Eukarya is warranted. It has been proposed that electrically conductive filaments of anaerobic methane-oxidizing archaea may be conduits for extracellular transfer to electron-accepting partners (22). Other possible benefits of archaellum conductivity might include facilitating attachment by dissipating charge barriers between cells and surfaces or electrical signaling between cells. Expression of synthetic, poorly conductive pili has played an important role in elucidating the function of e-pili in Geobacter species (1). Similar functional studies of M. hungatei will require the development of genetic tools for this microbe.

The cryo-EM structure of the M. hungatei e-archaellum (Fig. 2a), previously reported by Poweleit et al. (13), provides a much needed first opportunity to directly evaluate possible routes for long-range electron transport along a biologically produced protein filament. Aromatic rings of phenylalanine, tyrosine, and tryptophan are grouped into three well-separated regions: an outer sleeve (Fig. 2b and Fig. S5a), a middle sleeve (Fig. 2b and Fig. S5b), and a core (Fig. 2c). It was previously noted that the N-terminal phenylalanine residues in the archaellin subunits (Phe1) interact to “create a spokes effect via a π-stacking sandwich” that plays a key role in stabilizing the structure (13). Additional analysis of the distribution of aromatic amino acids (Fig. 2b and c and Fig. S5) further revealed that the aromatic rings of Phe1 and Phe13 in the core of the structure are packed almost as close as is physically possible (distances between ring centers of Phe1 and Phe13 of 4.5 and 5.1 Å), with angled T-shaped geometric orientations, which previous studies have suggested may enable π-π interactions (23). Furthermore, recent experimental evidence has indicated that, even in the absence of π-π stacking, phenylalanines within the hydrophobic core of an amino acid α-helical structure can facilitate long-range electron transport (10, 24). Therefore, our working hypothesis is that the Phe1-Phe13 core is at least one of the features contributing to the e-archaellum conductivity. Other aromatic amino acids of note include Phe20 (Fig. 2b), which is positioned close to the Phe1-Phe13 core, as well as outer and middle sleeves of aromatics that are well separated from each other and from the Phe1,13,20 core (Fig. 2b). Unlike the core, the outer and middle aromatic sleeves lack any closely spaced continuous chain of aromatics extending the length of the filament (Fig. S5).

Analogous to recent studies of G. sulfurreducens e-pili (20, 21, 25, 26), genetic manipulations to alter the positions of aromatic amino acids or other amino acids that may promote conductivity within the M. hungatei archaellum could lead to a better understanding of the structural features contributing to conductivity. The added benefit of such studies with the M. hungatei e-archaellum is that it will be possible to directly examine structural modifications to electron conductance pathways with cryo-EM. In the absence of genetic tools for M. hungatei, it will be necessary to heterologously express the gene for the M. hungatei archaellin in a genetically tractable archaeal host, similar to the expression of heterologous e-pili in G. sulfurreducens (3) or to identify a similar e-archaellum in a genetically tractable archaeon.

Microbially produced protein nanowires show substantial promise as a sustainable “green” electronic material with possibilities for functionalization and biocompatibility not available with other nanowire materials (1, 5–7). e-Archaella offer a unique opportunity to directly examine how synthetic designs to tune conductivity and/or add functionality influences protein nanowire structure, enabling a less empirical approach to the design of protein nanowire electronics.

M. hungatei was grown as previously described (1) in low-phosphate medium to induce archaellum expression. An aliquot (100 μl) of the culture was drop-cast onto highly oriented pyrolytic graphite (HOPG). Cells were allowed to attach to the HOPG for 10 min, and then the liquid was removed with a pipette tip. The surface was washed twice with 100 μl of deionized water, the surface was blotted dry at the edge with a Kimwipe, and the sample was placed in a desiccator overnight. All samples were equilibrated with atmospheric humidity (40%) inside the atomic force microscope (AFM) chamber for at least 2 h at 26.1°C at 1.1 mbarg. Conductive atomic force microscopy was performed using an Oxford Instruments/Asylum Research Cypher ES atomic force microscope. All topographic and current imaging was performed with a Pt/Ir-coated Arrow-ContPT tip with a 0.36701-N/m spring constant (NanoWorld AG, Neuchâtel, Switzerland). Topographic imaging was performed at a force of 0.1 nN. The conductive tip was attached to an ORCA dual-gain transimpedance amplifier and held at ground to serve as a translatable top electrode. A 300-mV bias was applied to the HOPG, and the locally detected current response of the archaellum was identified. Point-mode current-voltage (I-V) spectroscopy was performed by applying the conducting AFM tip at a force of 1 nN to the top of the archaellum and performing a voltage sweep at a frequency of 0.99 Hz. Continual force and current responses were collected for each I-V curve (Fig. S3) to ensure good consistent contact with the sample and avoid archaellum damage. The HOPG was periodically touched between samples to ensure the correct I-V response and tip quality (Fig. S4). Conductance was calculated from the linear portion of each I-V curve (−0.2 V to 0.2 V) as previously described (2). Average conductance and standard deviation were calculated for each of the three independent points on the three independent archaella.