Technical Highlight - September 2014
Short description: Computational and crystallographic analyses identify new hybrid folds that fill gaps in existing structural space.
Proteins are grouped into families and superfamilies according to their sequence and structural similarities. These classifications serve as the basis for protein function annotation and structural prediction, among other tasks. Similarities have also been assumed to reflect evolutionary links between superfamilies, but the relationships across superfamilies have been more difficult to define, due in part to a lack of appropriate computational tools.
Höcker and colleagues now employ improved homology search methods to query the Structural Classification of Proteins database, looking for connections within and between the (βα)8-barrel and flavodoxin-like folds. The (βα)8-barrel fold superfamilies were highly interconnected, while the flavodoxin-like fold proteins were only minimally linked.
Surprisingly, two of the flavodoxin-like fold superfamilies could be connected to twelve of the (βα)8-barrel superfamilies. To explore this sequence space further, the authors searched amongst existing genomic data for proteins that were equally related to both (βα)8-barrel and flavodoxin-like superfamilies. One of the identified sequences was N-TM0182, the N-terminal region of a hypothetic protein from Thermotoga maritima, the structure of which the authors solved (PDB 4Q37).
N-TM0182 crystallized as a β-strand swapped dimer of unclear physiological relevance. The topology of the protein follows an alternating helix-strand pattern similar to both the (βα)8-barrel and flavodoxin-like folds, with an additional strand at the C-terminus to complete a six-stranded sheet that displays similar curvature to that found in (βα)8-barrel proteins, but sandwiched between helices in a layer architecture similar to flavodoxin-like structures.
The authors further identified an (αβ)2 fragment of N-TM0182 that displayed the highest sequence identity— 37%— with representative superfamily counterparts, a flavodoxin-like cobalamin-binding protein from Desulfitobacterium hafniense (PDB 4JGI) and a (βα)8-barrel class I KDPG aldolase from T. maritima (PDB 1WA3). The authors further noted that the (αβ)2 structure could be superimposed multiple times on that of the KDPG aldolase due to its repetitive curved sheet, suggesting that this fragment served as an evolutionary nucleus for all three folds. These newly defined relationships help to fill a gap in protein structural space and provide a method by which scientists can search for sequences likely to have new structures.
J.A. Farías-Rico, S. Schmidt, and B. Höcker. Evolutionary relationship of two ancient protein superfolds.
Nat. Chem. Biol. 10, 710-715 (2014). doi:10.1038/nchembio.1579