HR: 0800h
AN: B11C-0623 [Abstracts]
TI: Predicting Structure and Function for Novel Proteins of an Extremophilic Iron Oxidizing Bacterium
AU: * Wheeler, K
EM: korin@llnl.gov
AF: Lawrence Livermore National Laboratory, 7000 East Avenue, Llivermore, CA 94610, United
States
AU: Zemla, A
EM: adamz@llnl.gov
AF: Lawrence Livermore National Laboratory, 7000 East Avenue, Llivermore, CA 94610, United
States
AU: Banfield, J
EM: jbanfield@berkeley.edu
AF: University of California Berkeley, Dept fo Earth and Planetary Sciences
307 McCone Hall 94720-4767, Berkeley, CA 94720-4767, United States
AU: Thelen, M
EM: mthelen@llnl.gov
AF: Lawrence Livermore National Laboratory, 7000 East Avenue, Llivermore, CA 94610, United
States
AB:
Proteins isolated from uncultivated microbial populations represent the functional components of microbial
processes and contribute directly to community fitness under natural conditions. Investigations into proteins in
the environment are hindered by the lack of genome data, or where available, the high proportion of proteins of
unknown function. We have identified thousands of proteins from biofilms in the extremely acidic drainage outflow
of an iron mine ecosystem (1). With an extensive genomic and proteomic foundation, we have focused directly on
the problem of several hundred proteins of unknown function within this well-defined model system.
Here we describe the geobiological insights gained by using a high throughput computational approach for
predicting structure and function of 421 novel proteins from the biofilm community. We used a homology based
modeling system to compare these proteins to those of known structure (AS2TS) (2). This approach has resulted
in the assignment of structures to 360 proteins (85%) and provided functional information for up to 75% of the
modeled proteins.
Detailed examination of the modeling results enables confident, high-throughput prediction of the roles of many
of the novel proteins within the microbial community. For instance, one prediction places a protein in the
phosphoenolpyruvate/pyruvate domain superfamily as a carboxylase that fills in a gap in an otherwise complete
carbon cycle. Particularly important for a community in such a metal rich environment is the evolution of over 25%
of the novel proteins that contain a metal cofactor; of these, one third are likely Fe containing proteins. Two of the
most abundant proteins in biofilm samples are unusual c-type cytochromes. Both of these proteins catalyze iron-
oxidation, a key metabolic reaction supporting the energy requirements of this community. Structural models of
these cytochromes verify our experimental results on heme binding and electron transfer reactivity, and provide
details for a working hypothesis of electron flow within the biofilm's major bacterium. Nearly 7% of the novel
proteins contain tetratrico peptide repeat (TPR) modules, a protein-protein interaction domain that participates in
signal transduction and a wide variety of other cellular functions. Like many biofilms, the various organisms in
this community use unknown mechanisms to communicate, relying upon each other for survival. Especially
interesting is evidence that most of these novel TPR proteins are located in the extracellular or membrane
fractions, suggesting their role in intracellular communication.
(1) Ram et al, 2005, Science 308:1915-20, "Community Proteomics of a Natural Microbial Biofilm"
(2) Zemla et al, 2005, Nucleic Acids Res 33 (Web Server issue):W111-5, "AS2TS system for protein structure
modeling and analysis"
This work was funded by the DOE Genomics: GTL Program and was performed under the auspices of the DOE
by the University of California, Lawrence Livermore National Laboratory under contract W-7405-Eng-48.
DE: 0430 Computational methods and data processing
DE: 0448 Geomicrobiology
DE: 0456 Life in extreme environments
DE: 0465 Microbiology: ecology, physiology and genomics (4840)
SC: Biogeosciences [B]
MN: 2007 Fall Meeting