 |
 |
Scientific Report 2006
Molecular Biology
Predicting Protein Structure, Association, and Inhibitors
R. Abagyan, J. An,* W.
Bisson, A. Cheltsov, K. Hyun, J. Kovacs, I. Kufareva, P. Lam,** G. Nicola, A.
Saldanha
* Genome Sciences Centre, Vancouver, British Columbia
** Molsoft
L.L.C., La Jolla, California
Today
the Protein Data Bank contains more than 37,000 structures and is growing at a rate
of 20 per day. These structures provide a unique opportunity for functional studies
and rational design of therapeutic agents. We continue to focus on annotating and
characterizing the protein structures in terms of their interaction interfaces and
flexibility; predicting protein associations; modeling homologous structures and
membrane proteins; predicting conformational rearrangements; and, finally, using
ligand docking and virtual screening to detect inhibitors of specific molecular
targets. This past year, our efforts in the last area led to new or improved inhibitors
against the receptor for epidermal growth factor (EGFR), anthrax lethal factor,
dynamin, α1-antitrypsin,
and the androgen receptor.
Bioinformatics and Cheminformatics
We helped G. Siuzdak and his group,
Department of Molecular Biology, build a cheminformatics system for characterizing
metabolites on the basis of liquid chromatography–mass spectrometry data. The
resulting software (XCMS) incorporates novel nonlinear retention time alignment,
matched filtration, peak detection, and peak matching and is freely available from
http://metlin.scripps.edu/download/. The software helps identify changes in specific
endogenous metabolites, such as potential biomarkers.
We have proposed a method for sharing
chemical information in conjunction with data on experimental compounds without
revealing the identity of the compounds. Privacy of chemical structure is of paramount
importance in the industrial sector, and the proposed solution opens a way to transfer
a rich knowledge base from the pharmaceutical industry to academia.
Finally, we collaborated with scientists
at the Structural Genomics Consortium, Oxford, England, to improve the way the new
structures are annotated, distributed, and animated by using internal coordinates–based
methods.
Ligand Discovery
Small-molecule therapeutic agents
can be discovered by using docking and virtual chemical library screening. The docking
technology can also help in understanding structural mechanisms of action of small
molecules and rational design of better molecules. However, modeling protein flexibility
and ligand-induced conformational changes is a major challenge. We modeled the induced
receptor rearrangements at several levels, including relevant normal modes combined
with full side-chain sampling and “minus-one” calculations. In particular,
we used the developed ligand-induced receptor simulation techniques to identify
new antagonists of the androgen receptor and the first small-molecule inhibitors
of α1-antitrypsin
polymerization.
Our docking-based in silico chemical
library screening against the EGFR tyrosine kinase and the consequent experimental validation
allowed identification of several compounds with antiproliferative effects on cancer
cells. Among them, a C(4)-N(1)-substituted pyrazolo[3,4-d]pyrimidine inhibits EGFR
tyrosine kinase activity at micromolar concentrations.
We screened a library of tyrphostins
against the GTPase activity of dynamin I and performed optimization of discovered
compounds. The results yielded a number of promising inhibitors that are effective
at micromolar concentrations.
Using a fragment-based approach,
we developed inhibitors of the lethal factor metalloproteinase of Bacillus anthracis.
The discovered compounds are highly potent and selective against lethal factor in
in vitro assays, including cell-based assays.
Peptide Docking and Structure Prediction
Predicting partial protein structure
or molecular association is a critical task of computational biology, which remains
a focus of our research. In particular, we developed a method for ab initio prediction
of peptide-MHC binding geometry for diverse class I MHC allotypes. Such models are
useful for predicting specific ternary complexes with T-cell receptors and for designing
new molecules that interact with these complexes. The surprisingly accurate prediction
(0.75-Å backbone root mean square deviation) that we achieved by using our
method for cross-docking of a highly flexible decapeptide, dissimilar to the original
bound peptide, and docking predictions with homology models for 2 allotypes with
mean backbone root mean square deviations of less than 1.0 Å illustrate the
effectiveness of the method.
Predicting Functional Sites
Functional annotation of protein
structures involves identifying and characterizing protein-protein interfaces, oligomerization
states, and binding sites for small ligands. We developed a method called protein
interface recognition that can be used to predict interfaces on the basis of an
isolated protein structure and does not depend on evolutionary information. The
method was benchmarked by using a diverse set of 748 protein interfaces. The accuracy
and efficiency make the method a suitable tool for automated high-throughput annotation
of protein structures discovered in structural proteomics studies (Fig. 1).
 |
| Fig. 1. Predicting
protein oligomerization geometry by using protein interface recognition. |
Some protein interfaces can safely
be targeted for drug discovery. We developed a systematic approach to assessing
the “druggability” of a protein interface. The approach includes detecting
a suitable ligand-binding pocket with maximal confidence in a functionally sensitive
location on the biomolecule and assessing the reliability of the local structure.
This approach was applied to the Skp1-Cullin-F-box protein ubiquitin ligase interface.
It can be used before high-throughput or virtual library screening.
CD59
is a membrane glycoprotein with therapeutic potential for treatment of inflammatory
conditions. Using scanning mutagenesis, refined nuclear magnetic resonance models,
and additional site-specific mutations, we identified a binding interface on CD59
that is much broader than previously thought. We identified substitutions that decreased
CD59 activity and a surprising number of substitutions that enhanced it. On the
basis of these findings, we prepared clinically relevant soluble mutant CD59-based
proteins that had up to a 3-fold increase in complement inhibitory activity.
Publications
Abagyan, R. Problems
in computational structural genomics. In: Structural Proteomics. Sundstrom,
M., Norin, M., Edwards, A. (Eds.). CRC Press, Boca Raton, FL, 2006, p. 223.
Abagyan, R., Lee, W.H., Raush,
E., Budagyan, L., Totrov, M., Sundstrom, M., Marsden, B.D. Disseminating
structural genomics data to the public: from a data dump to an animated story. Trends
Biochem. Sci. 31:76, 2006.
Bordner, A., Abagyan, R.A.
Ab initio prediction of peptide-MHC binding geometry
for diverse class I MHC allotypes. Proteins 63:512, 2006.
Cardozo, T., Abagyan, R.
Druggability of SCF ubiquitin ligase-protein interfaces. Methods Enzymol. 399:634,
2005.
Cavasotto, C.N., Orry, A.J.W.,
Abagyan, R. Receptor flexibility in ligand docking.
In: Handbook of Theoretical and Computational Nanotechnology. Rieth, M.,
Schommers, W. (Eds.). American Scientific Publishers, Stevenson Ranch, CA, 2006,
Vol.6, p. 217.
Cavasotto, C.N., Orry, A.J.W.,
Abagyan, R.A. The challenge of considering receptor
flexibility in ligand docking and virtual screening. Curr. Comput. Aided Drug Des.
1:423, 2005.
Cavasotto, C.N., Ortiz, M.A.,
Abagyan, R.A., Piedrafita, F.J. In silico identification
of novel EGFR inhibitors with antiproliferative activity against cancer cells. Bioorg.
Med. Chem. Lett. 16:1969, 2006.
Forino, M., Johnson, S., Wong,
T.Y., Rozanov, D.V., Savinov, A.Y., Li, W., Fattorusso, R., Becattini, B., Orry,
A.J., Jung, D., Abagyan, R.A., Smith, J.W., Alibek, K., Liddington, R.C., Strongin,
A.Y., Pellecchia, M. Efficient synthetic inhibitors
of anthrax lethal factor. Proc. Natl. Acad. Sci. U. S. A. 102:9499, 2005.
Hill, T., Odell, L.R., Edwards,
J.K., Graham, M.E., McGeachie, A.B., Rusak, J., Quan, A., Abagyan, R., Scott, J.L.,
Robinson, P.J., McCluskey, A. Small molecule inhibitors
of dynamin I GTPase activity: development of dimeric tyrphostins. J. Med. Chem.
48:7781, 2005.
Huang, Y., Smith, C.A., Song,
H., Morgan, B.P., Abagyan, R., Tomlinson, S. Insights
into the human CD59 complement binding interface toward engineering new therapeutics.
J. Biol. Chem. 280:34073, 2005.
Kovacs, J.A., Cavasotto, C.N.,
Abagyan, R.A. Conformational sampling of protein
flexibility in generalized coordinates: application to ligand docking. J. Comput.
Theor. Nanosci. 2:354, 2005.
Kufareva, I., Budagyan, L.,
Raush, E., Totrov, M., Abagyan, R. PIER: protein
interface recognition for structural proteomics. Proteins, in press.
Orry, A.J., Abagyan, R.A.,
Cavasotto, C.N. Structure-based development of target-specific
compound libraries. Drug Discov. Today 11:261, 2006.
Smith, C.A., O’Maille,
G., Want, E.J., Qin, C., Trauger, S.A., Brandon, T.R., Custodio, D.E., Abagyan,
R., Siuzdak, G. METLIN: a metabolite mass spectral
database. Ther. Drug Monit. 27:747, 2005.
Smith, C.A., Want, E.J., O’Maille,
G., Abagyan, R., Siuzdak, G. XCMS: processing mass
spectrometry data for metabolite profiling using nonlinear peak alignment, matching,
and identification. Anal. Chem. 78:779, 2006.
Tetko, I.V., Abagyan, R., Oprea,
T.I. Surrogate data: a secure way to share corporate
data. J. Comput. Aided Mol. Des. 19:749, 2005.
|
|
