![Show Menu](styles/mobile-menu.png)
![Page Background](./../common/page-substrates/page0155.jpg)
Conformational Ensembles from Experimental Data
and Computer Simulations
Poster Abstracts
151
114-POS
Board 34
Protein Evolution Under a Computational Microscope
Huafeng Xu
1
, David E. Shaw
1
,
1
D. E. Shaw Research, New York, NY, USA,
2
Columbia University, New York, NY, USA.
Protein evolution is often driven by changes in the amino acid sequence that alter the affinity and
specificity of the protein for its binding partners. Mutations that do not appear to affect the
specific interactions between the protein and its binding partner can have a surprisingly large
effect on the binding affinity. We have used long-timescale molecular dynamics (MD)
simulations to study how such mutations affect the binding affinities in two examples of protein
evolution: 1) the affinity maturation of a broadly neutralizing antibody lineage against influenza
hemagglutinin, in which two divergent maturation pathways have led to mature antibodies that
potently neutralize a broad range of H1 influenza viruses, and 2) the adaptation of influenza
hemagglutinin, in which hemagglutinins of avian strains have accrued mutations that favor
binding to human receptors over avian receptors (a requirement for human transmission). In the
affinity maturation study, our simulations contributed to the finding that the affinity increase in
the mature antibodies was primarily attributable to the stabilization of the CDR H3 loop in the
antigen-binding conformation. In the hemagglutinin adaptation study, our simulations suggested
that the human receptor adopts a dynamic ensemble of diverse conformations in binding to
hemagglutinin, including a novel conformation not yet seen in crystallography. We identified
mutations in an avian hemagglutinin that we predicted would favor the formation of new specific
interactions with the human receptor in the novel binding conformation, and then experimentally
verified increased affinity of the mutant hemagglutinin for the human receptor. Results from
these studies suggest that conformations generated by MD simulations, including some which
have not been previously identified by experimental structural determination, may help unravel
the structural mechanism underlying the evolution of proteins, and serve as starting points for
engineering biomolecular complexes with enhanced affinity and specificity.