Analysis of ribosomal protein block structure: Functional characterization, evolutionary implications and distant homology search using discrete state models.

摘要

The ribosome, a very complex molecular machine, plays a fundamental role in all living organisms and exhibits extraordinary engineering design concepts. This investigation examined the complexity of the translational apparatus, seeking to understand how its components evolved to their present configuration. It includes detailed sequence and structural comparative analyses of the ribosome and its associated proteins with functional characterization of these components across the three phylogenetic domains.; Amino acid sequence alignments of ribosomal proteins revealed an unusual taxon-specific block structure, with some blocks universally conserved and others specific to one or two phylogenetic domains. Statistical and phylogenetic analyses of the universal blocks imply that modern Bacteria, Archaea and Eukarya clearly have a common ancestor, while the phylodomain-specific blocks suggest that these groups also share more recent, taxon-specific cenancestors. Major evolutionary implications of the observed block structure are: (i) the crenarchaeal, endosymbiotic origin of the modern eukaryotic translational apparatus; and (ii) the occurrence of a prokaryotic bottleneck that drastically reduced the diversity of modern species progenitors about 2.2 billion years ago.; Surprisingly, the highly conserved blocks identified in most of the translation-related proteins do not associate consistently with any identifiable particular function or structural feature, or even with rRNA contacts. A comprehensive investigation of the rRNA-ribosomal protein interactions, however, demonstrated a major role of ribosomal proteins in constraining the rRNA conformational space and stabilizing its correct, universally conserved core fold.; In order to identify possible evolutionary relationships between the taxon-specific block structure and other proteins, a new stochastic tool for the identification of distant homologous domains in single-, repeated- and multi-domain contexts was implemented. The approach uses sequence and structure information embedded in Discrete State Models, and a Markov threading technique to estimate the compatibility of any query sequence with the models under consideration. The method was successfully applied to a variety of cases, including the ribosomal blocks, the WD40-repeat domain and the very diverse ubiquitin-like family.