From: ncbi-seminar-admin@ncbi.nlm.nih.gov on behalf of John Wilbur
[wilbur@ncbi.nlm.nih.gov]
Sent: Friday, September 20, 2002 12:59 PM
To: CBBSearch@ncbi.nlm.nih.gov; ncbi-seminar@ncbi.nlm.nih.gov
Subject: Aravind seminar - Monday

L. Aravind Seminar Sept. 23, 2 pm, Natcher Center, 6th floor, South
Conference Room
-------------------------------------------------------------------------

The origin and evolution of the protein universe
L Aravind, PhD
Computational Biology Branch,
National Center for Biotechnology Information

Ever since the pioneering work of Zuckerkandl and Pauling, protein
sequences and structures have been used extensively to infer organismal
phylogeny and to predict biochemical functions. However, the origin and
early evolution of proteins, and the diversification of protein families
during the divergence of the major lineages of life, are poorly understood.
The recent explosion of complete genome sequences and protein structures
provides the necessary material to construct and test hypothesis regarding
these issues in protein evolution. As these models cannot be easily
reconstructed using conventional phylogenetic methods alone, diverse
methodologies were applied to glean the requisite information. These
include: 1) Detection of distant relationships between proteins through a
combination of sequence and structural analysis, and classification of the
protein world into monophyletic assemblages of domains. 2) Identification
of sequence or structural features that are shared derived characters for
particular clades within these monophyletic assemblages of domains. 3)
Comparisons of the phyletic distributions of domains to derive the domain
complements in the last common ancestors of the organismal lineages under
consideration. 4) Detection of anomalies in tree topologies and phyletic
distributions of particular proteins to decipher evolutionary forces
involved, such as, lateral transfer and gene loss.

As a result, the most conserved sets of proteins and constituent domains,
traceable to the Last Universal Common Ancestor (LUCA) of all life forms
were identified. Through a comparison of the homologous domains in this
set, the pre-LUCA stages of protein evolution were reconstructed. One of
the conclusions that became apparent was that even before the extant
translation apparatus was in place, complex protein domains, resembling
extant forms, were already being synthesized. These investigations also
lead to the hypothesis that early enzymatic domains with specific
activities emerged from generalized RNA (ribozyme) interacting modules. The
ribozymes were eventually displaced through the acquisition of
strategically placed residues that allowed the protein to acquire catalytic
activity.

The subsequent evolution of proteins was explored by identifying certain
general principles that explain the origin of several distinct classes of
domains. These include, the emergence of lineage specific alpha-helical
domains through the duplication and collapse of simple helical segments,
fusion of pre-existing domains into single folding units, and stabilization
of incipient protein folds through metal chelation and disulfide bond
formation. Additionally, various tendencies in the adaptive radiation of
protein domains, such as lineage specific expansions, colonization of new
functional niches, and emergence of novel domain architectures following
massive lateral transfer events were identified. Use of these tendencies to
explain the origin of eukaryotes and particular biological systems during
the subsequent diversification of the eukaryotes is illustrated.
Application of these studies in identifying and predicting functions of
previously uncharacterized components of the eukaryotic chromatin, RNA
metabolism- and signal transduction- systems is demonstrated.