Scaffolding protein functional sites using deep learning

The binding and catalytic functions of proteins are generally mediated by a small number of functional residues held in place by the overall protein structure. Here, we describe deep learning approaches for scaffolding such functional sites without needing to prespecify the fold or secondary structure of the scaffold. The first approach, "constrained hallucination," optimizes sequences such that their predicted structures contain the desired functional site. The second approach, "inpainting," starts from the functional site and fills in additional sequence and structure to create a viable protein scaffold in a single forward pass through a specifically trained RoseTTAFold network. We use these two methods to design candidate immunogens, receptor traps, metalloproteins, enzymes, and protein-binding proteins and validate the designs using a combination of in silico and experimental tests.

Molecular Modeling of Bacterial Nanomachineries

Matteo Thomas Degiacomi

Proteins have the ability to assemble in multimeric states to perform their specific biological function. Unfortunately, characterizing experimentally these structures at atomistic resolution is usually difficult. For this reason, in silico methodologies aiming at predicting how multiple protein copies arrange to forma multimeric complex would be desirable. We present Parallel OptimizationWorkbench (POW), a swarm intelligence based optimization framework able to deal, in principle, with any optimization problem. We show that POW can be applied to biologically relevant problems such as prediction of protein assemblies and the parameterization of a Coarse-Grained force field for proteins. By combining POW optimizations, Molecular Dynamics simulations, Poisson-Boltzmann calculations and a variety of experiments, we subsequently study two bacterial nanomachieries: Aeromonas hydrophila's pore-forming toxin aerolysin, and Yersinia enterocolitica injectisome. These structures are challenging both for their size, and for the timescales involved in their functioning. Aerolysin is a pore-forming toxin secreted as an hydrophilic monomer. By means of large conformational changes, the protein heptamerizes on the target cell's surface, and finally inserts β-barrel into its lipid bilayer, causing cell death. The main hurdle in the study of this structure is the complexity of the mode of action, which spans timescales currently unreachable by classical molecular dynamics. We show that aerolysin C-terminal region has the dual role of preventing premature oligomerization and helping the folding of tertiary structure, qualifying therefore as an intramolecular chaperone. We study the transmembrane β-barrel properties and compare them with those of the homologous protein α-hemolysin. We show that aerolysin's barrel is more rigid than α-hemolysin's, and should be anion selective. We present models for aerolysin heptamer both in prepore and, for the first time, in membrane-inserted conformation. Our results are validated experimentally, and are consistent with known biochemical and structural data. The injectisome is an example of a type III secretion system. Its most striking feature is probably its size: hundreds of proteins assemble in a unique structure spanning the Gram-negative bacterial double membrane, and protruding outside the cell as a needle for tenth of nanometers. Obtaining an atomistic representation of this massive structure, and therefore some insights about its mode of action, is one of the greatest challenges. We show that the final length of injectisome's needle is determined by the secondary structure content of a ruler protein located inside its cavity during assembly. Using POW, we also produce the first model for Yersinia injectisome's basal body, highlighting the flexibility of this region in adapting between the inner and outer membranes. As a whole, this work demonstrates that a synergy of dry and wet experiments can provide precious insights into macromolecular structure and function.

EPFL2012

Peptide Structure and Radical Chemistry in the Gas Phase Revealed by Product Ion Abundance Analysis in Electron Capture and Transfer Dissociation Mass Spectrometry

Hisham Ben Hamidane

Peptide and protein ion abundances in mass spectrometry (MS) and tandem mass spectrometry (MS/MS) may provide orthogonal, or complementary, information to their mass measurements. Collision induced dissociation (CID), the most commonly used MS/MS technique, demonstrates high variability of fragmentation pathways and their relative product abundances as a function of experimental parameters. Moreover, a reliable prediction of relative abundances of CID products is possible only for relatively small systems, but not for larger biomolecules. Recently developed radical-induced fragmentation techniques, e.g., electron capture and transfer dissociation (ECD/ETD), appear to demonstrate more repeatable and reproducible fragmentation patterns. However, it has been reported that ECD/ETD product ion abundance distribution for peptides is rather flat and extracting additional peptide sequence information is yet to be done. In the present work, an attempt is made to understand the peptide's ECD/ETD product ion abundance formation fundamentals and to reveal the possible implications of this source of information for improving peptide and protein structural characterization. Specifically, we have shown that ECD/ETD of peptides is sensitive to the chemical environment in terms of peptide primary structure (amino acid nature) and possibly higher order structure (hydrogen bond connectivity). After demonstrating the analytical validity of the ECD/ETD product ion abundance methodology, its parameter dependence and its applicability to different experimental setups, results will be presented and rationalized along two main axes of comprehension: gas phase peptide radical chemistry and structure. Whereas fundamental aspects of ECD/ETD (mechanism, thermodynamics) can be addressed with the former, additional ECD/ETD yield modulations were observed and presumably relate to structural properties of ions in the gas phase. ECD specificity will be demonstrated on the class of amphipathic peptides, where periodicity is revealed in product ion abundance and correlated to peptide primary and secondary structure. To decipher between both contributions, sequence variations of amphipathic peptides were probed by ECD and correlated to gas phase structures suggested by ion mobility mass spectrometry and supported by molecular dynamics computations. After probing amino acid side chain influence, unraveling the chemical environment specificity of ECD/ETD led us to investigate peptide backbone modifications. Peptides containing beta amino acids revealed two main effects in ECD and ETD. We demonstrate that individual substitutions yield c and z type fragments only for amino acids that provide sufficient z• radical stabilization, whereas multiple substitutions shift the preferential fragmentation pathway to the otherwise minor a-type ion formation pathway indicating possible change in peptide protonation. Finally, understanding the partitioning of radical ions between N- and C-terminal fragments led us to investigate ion internal energy dependence. The extent of hydrogen atom transfer, the [c+z] complex lifetime and secondary fragmentation pathways (w-type ion formation) are directly modulated by ion internal energy as demonstrated by decoupling ETD from charge reduced CID. To summarize, ECD/ETD product ion abundance methodology developed in the thesis advances biomolecule radical chemistry understanding and provides a potentially useful methodology for peptide and protein structural characterization.

EPFL2011

PDB-Explorer: a web-based interactive map of the protein data bank in shape space

Daniel Kostro, Luc Patiny, Jean-Louis Reymond, Michaël Giuseppe Zasso

Background: The RCSB Protein Data Bank (PDB) provides public access to experimentally determined 3D-structures of biological macromolecules (proteins, peptides and nucleic acids). While various tools are available to explore the PDB, options to access the global structural diversity of the entire PDB and to perceive relationships between PDB structures remain very limited. Methods: A 136-dimensional atom pair 3D-fingerprint for proteins (3DP) counting categorized atom pairs at increasing through-space distances was designed to represent the molecular shape of PDB-entries. Nearest neighbor searches examples were reported exemplifying the ability of 3DP-similarity to identify closely related biomolecules from small peptides to enzyme and large multiprotein complexes such as virus particles. The principle component analysis was used to obtain the visualization of PDB in 3DP-space. Results: The 3DP property space groups proteins and protein assemblies according to their 3D-shape similarity, yet shows exquisite ability to distinguish between closely related structures. An interactive website called PDB-Explorer is presented featuring a color-coded interactive map of PDB in 3DP-space. Each pixel of the map contains one or more PDB-entries which are directly visualized as ribbon diagrams when the pixel is selected. The PDB-Explorer website allows performing 3DP-nearest neighbor searches of any PDB-entry or of any structure uploaded as protein-type PDB file. All functionalities on the website are implemented in JavaScript in a platform-independent manner and draw data from a server that is updated daily with the latest PDB additions, ensuring complete and up-to-date coverage. The essentially instantaneous 3DP-similarity search with the PDB-Explorer provides results comparable to those of much slower 3D-alignment algorithms, and automatically clusters proteins from the same superfamilies in tight groups. Conclusion: A chemical space classification of PDB based on molecular shape was obtained using a new atom-pair 3D-fingerprint for proteins and implemented in a web-based database exploration tool comprising an interactive color-coded map of the PDB chemical space and a nearest neighbor search tool. The PDB-Explorer website is freely available at www.cheminfo.org/pdbexplorer and represents an unprecedented opportunity to interactively visualize and explore the structural diversity of the PDB.

Biomed Central Ltd2015