Atomic Force Microscopy-assisted modeling of macromolecular assembly – AFM-Assembly

1- Scientific background and objectives: - Post-genomic era has accelerated structure determinations of proteins from a large variety of organisms. There are more than 86000 known protein structural domains. The challenge in facing biology is now to understand how these domains assemble to form large multi-domain protein machinery. Current biophysical methodologies face increasing difficulties in studying multi-domain proteins and their assemblies with partners, such as to work with complicate and hard to crystallize large systems. Recent advances in Atomic Force Microscopy (AFM) encourage us to think that AFM is well-adapted to modern challenges in structural biology mostly due to the ease of use concerning the biological samples. AFM does not need lenses to form images. Instead, they use a sharply pointed sensor tip to explore properties of the sample surface. It does not require crystallized samples and it is also well adapted for working on small sample quantities even at a single molecule level. The objective of this proposal is to demonstrate that the surface representation of molecules obtained using AFM, known as topographic envelops, will allow us to reconstruct large multi-domain proteins as well as assemble protein-protein complexes. - - - 2- Description of the project, methodology: - The project contains two complementary aspects: collecting high-resolution topographic envelops of proteins using AFM and in silico reconstructing these macromolecules using their isolated components for which we have three-dimensional (3D) coordinates. For this purpose, we are developing in silico modeling techniques that incorporate the AFM topographic envelops of macromolecules as experimental constraints. In this proposal we will demonstrate the applicability of our approach to three different but complementary systems: a large multi-domain protein (antibody) will be reconstructed using their isolated fragments (Fab and Fc), a protein-protein complex made of the blood coagulation factor Va and its associated enzyme the factor Xa will be assembled, and the assembly of the quaternary structure of an unknown protein structure using a recently build molecular model. The first challenge is to generate high-resolution images for single macromolecules. We will make use of the latest advances in image processing techniques such as tip-broadening deconvolution algorithms and ultra-thin carbon nanotubes in order to increase the lateral resolution of our topographic envelops. The second challenge concerns reconstruction algorithms to assemble molecules in real space. The strategy we are developing is to transform the AFM-based topographic envelope into a molecular volume. Such volumes are used as a confined space where exhaustive search for the best fitting orientation of isolated protein domains can be performed with six degrees of freedom, a step called self-docking. - Our laboratory is certified with the label ISO9001:2000 concerning project management in scientific research. According to our guide, the timetable of the project spans over three years and will identify appropriate milestones. - - - 3- Expected results: - The goal of this three-year proposal is to build a computational workflow that takes as input high-resolution topographic envelops and atomic coordinates of a molecule and produces as output a predicted orientation of the molecule within the experimental constraints using the volume defined by envelopes. Expected application of the reconstruction of multi-domain proteins can be used in the study of conformational flexibility of large macromolecules, an essential property of antibodies, for instance. In addition, it will allow the investigation of large macromolecules containing partially unfolded regions or domains, a property that is extremely difficult to study with classical structural biology techniques. Finally, using experimental constraints in protein-protein docking offers new possibilities in the construction of large s...