Atomic Structure of Nearly Indestructible Pili from a Hyperthermophilic Acidophile

Abstract

The mechanism by which viruses infect hyperthermophilic archaeal acidophiles has been poorly understood. Quemin et al. (2013) showed by transmission electron microscopy and whole-cell electron cryo-tomography that SIRV2 rod-shaped virions specifically recognize the tips of highly abundant pilus-like filaments on the surface of S. islandicus. However, the identity of these pili was unknown. All approaches to biochemically analyze the filaments failed, as the filaments could never be degraded or depolymerized under a variety of extreme conditions: boiling in SDS, boiling in 5M guanidinium chloride, or incubating with a high concentration of pepsin for a week. We therefore were never able to analyze the filaments by mass spectrometry to determine the identity of the protein composition of the pili. We have been able to use cryo-EM to generate a 4.1 Å resolution reconstruction of the pili, which, combined with bioinformatics, has allowed us to determine its amino acid sequence and build an atomic model de novo. The N-terminal portion of the protein is a homolog of the bacterial Type 4 pilin (T4P) N-terminal domain, but in contrast to bacterial T4P, the helix does not partially melt and is packed in the pili in a very different manner. The globular domain is mainly β-stranded, with a largely hydrophobic surface. Glycosylation of several surface residues can be seen in the cryo-EM map, and STEM mass analysis suggests a significant amount of additional glycosylation that must be largely disordered and therefore not seen in the reconstruction. The factors that cause these pili to be nearly indestructible will be discussed.