Intermediate Filaments Supporting Cell Shape and Growth in Bacteria.

For years intermediate filaments (IF), belonging to the third class of filamentous cytoskeletal proteins alongside microtubules and actin filaments, were thought to be exclusive to metazoan cells. Structurally these eukaryote IFs are very well defined, consisting of globular head and tail domains, which flank the central rod-domain. This central domain is dominated by an α-helical secondary structure predisposed to form the characteristic coiled-coil, parallel homo-dimer. These elementary dimers can further associate, both laterally and longitudinally, generating a variety of filament-networks built from filaments in the range of 10 nm in diameter. The general role of these filaments with their characteristic mechano-elastic properties both in the cytoplasm and in the nucleus of eukaryote cells is to provide mechanical strength and a scaffold supporting diverse shapes and cellular functions.Since 2003, after the first bacterial IF-like protein, crescentin was identified, it has been evident that bacteria also employ filamentous networks, other than those built from bacterial tubulin or actin homologues, in order to support their cell shape, growth and, in some cases, division. Intriguingly, compared to their eukaryote counterparts, the group of bacterial IF-like proteins shows much wider structural diversity. The sizes of both the head and tail domains are markedly reduced and there is great variation in the length of the central rod-domain. Furthermore, bacterial rod-domains often lack the sub-domain organisation of eukaryote IFs that is the defining feature of the IF-family. However, the fascinating display of filamentous assemblies, including rope, striated cables and hexagonal laces together with the conditions required for their formation both in vitro and in vivo strongly resemble that of eukaryote IFs suggesting that these bacterial proteins are deservedly classified as part of the IF-family and that the current definition should be relaxed slightly to allow their inclusion. The lack of extensive head and tail domains may well make the bacterial proteins more amenable for structural characterisation, which will be essential for establishing the mechanism for their association into filaments. What is more, the well-developed tools for bacterial manipulations provide an excellent opportunity of studying the bacterial systems with the prospect of making significant progress in our understanding of the general underlying principles of intermediate filament assemblies.