Bioinformatics

Something Old, Something New

+ See all authors and affiliations

Science  09 Aug 2002:
Vol. 297, Issue 5583, pp. 899
DOI: 10.1126/science.297.5583.899c

The difficulty of determining the crystal structures of macromolecular complexes, especially those that may exist only transiently, has prompted recourse to other methods, such as the widely used comparison of amino acid sequences as well as the newer mapping of protein-protein interactions.

Aloy et al. have borrowed from both toolkits in building a model of the structure of the yeast exosome, a complex of 11 distinct subunits that furnishes exonuclease activity during the processing of RNA molecules. Sequence analysis of the yeast proteins and of bacterial polynucleotide phosphorylase (PNPase) delineated core and peripheral domains representing ribonuclease and RNA-binding domains. Electron microscopy of the exosome and comparison with the crystal structure of PNPase established an overall size and shape similarity. Placement and ordering of the six core domains within the trimeric envelope was accomplished on the basis of functional site analysis and a comparison to the structure of the trimeric proliferating cell nuclear antigen. How accurate is this map? Some interaction data are consistent, some not. — GJC

EMBO Rep.3, 628 (2002).

Related Content

Navigate This Article