Atomistic simulation of carbohydrate-protein complex formation: Hevein-32 domain.

Interactions between proteins and their small molecule ligands are of great importance for the process of drug design. Here we report an unbiased molecular dynamics simulation of systems containing hevein domain (HEV32) with N-acetylglucosamine mono-, di- or trisaccharide. Carbohydrate molecules were placed outside the binding site. Three of six simulations (6 × 2 µs) led to binding of a carbohydrate ligand into the binding mode in agreement with the experimentally determined structure. Unbinding was observed in one simulation (monosaccharide). There were no remarkable intermediates of binding for mono and disaccharide. Trisaccharide binding was initiated by formation of carbohydrate-aromatic CH/π interactions. Our results indicate that binding of ligands followed the model of conformational selection because the conformation of the protein ready for ligand binding was observed before the binding. This study extends the concept of docking by dynamics on carbohydrate-protein interactions.

Anncolvar: Approximation of Complex Collective Variables by Artificial Neural Networks for Analysis and Biasing of Molecular Simulations.

The state of a molecular system can be described in terms of collective variables. These low-dimensional descriptors of molecular structure can be used to monitor the state of the simulation, to calculate free energy profiles or to accelerate rare events by a bias potential or a bias force. Frequent calculation of some complex collective variables may slow down the simulation or analysis of trajectories. Moreover, many collective variables cannot be explicitly calculated for newly sampled structures. In order to address this problem, we developed a new package called anncolvar. This package makes it possible to build and train an artificial neural network model that approximates a collective variable. It can be used to generate an input for the open-source enhanced sampling simulation PLUMED package, so the collective variable can be monitored and biased by methods available in this program. The computational efficiency and the accuracy of anncolvar are demonstrated on selected molecular systems (cyclooctane derivative, Trp-cage miniprotein) and selected collective variables (Isomap, molecular surface area).

The C-type lectin-like receptor Nkrp1b: Structural proteomics reveals features affecting protein conformation and interactions.

The cytotoxicity of mouse natural killer (NK) cells in response to pathological changes in target cells is regulated via the Nkrp1b receptor. Here, we characterized the Nkrp1b structure and structural features (stalk, loop, and oligomerization state) that affect its interactions. To study the Nkrp1b protein structure and the functional importance of its stalk, two Nkrp1b protein variants differing by the presence of the stalk were prepared. These variants were studied using a combination of structural mass spectrometry approaches with computational modeling to derive structural models. In addition, information about biological activity and localization in mammalian cells was acquired using scanning microscopy techniques and western blotting. Based on these methods, we obtained the structure of Nkrp1b ectodomain in its monomeric and dimeric conformations, identified the dimerization interface, and determined disulfide connections within the molecule. We found that Nkrp1b occurs as a mixture of monomers and homodimers, both in vitro and in vivo. SIGNIFICANCE: Despite the long-standing assumption that Nkrp1 proteins are homodimers connected by disulfide bonds in the stalk region, our data showed that both Nkrp1b protein variants form monomers and homodimers irrespective of the presence of the stalk. We demonstrated that the stalk is not crucial for protein dimerization or ligand binding and that Nkrp1b interacts with its natural ligands only in its monomeric conformation; therefore, dimers may have another regulatory function. Using a unique combination of computational, biochemical, and biological methods, we revealed the structural conformation and behavior of Nkrp1b in its native state. In addition, it is a first report utilizing the intermolecular chemical cross-linking of light- and heavy-labeled protein chains together with ion mobility-mass spectrometry to design the structural models of protein homodimers.

Molecular simulations of hevein/(GlcNAc)3 complex with weakened OH/O and CH/π hydrogen bonds: implications for their role in complex stabilization.

Carbohydrate-protein complexes are often characterized by interactions via aromatic amino acid residues. Several mechanisms have been proposed to explain these stacking-like interactions between pyranose sugars and aromatic moieties. The physical basis of these interactions is being explained as either dispersion CH/π or hydrophobic. In order to elucidate the nature of these interactions, we performed a series of molecular dynamics simulation of hevein domain (HEV32) in complex with (ß-D-GlcNAc)3. Selected OH/O and CH/π hydrogen bonds involved in carbohydrate recognition were artificially weakened in 100 ns molecular dynamics simulations. Separate weakening of either OH/O or CH/π hydrogen bonds was not sufficient to destabilize the complex. This indicates that other effects, not solely CH/π dispersion interactions, contribute significantly to the stability of the complex. Significant destabilization of complexes was reached only by simultaneous weakening of OH/O and CH/π hydrogen bonds. This also shows that classical hydrogen bonds and CH/π interactions are working in concert to stabilize this carbohydrate-protein test case.