High-confidence 3D template matching for cryo-electron tomography.

Visual proteomics attempts to build atlases of the molecular content of cells but the automated annotation of cryo electron tomograms remains challenging. Template matching (TM) and methods based on machine learning detect structural signatures of macromolecules. However, their applicability remains limited in terms of both the abundance and size of the molecular targets. Here we show that the performance of TM is greatly improved by using template-specific search parameter optimization and by including higher-resolution information. We establish a TM pipeline with systematically tuned parameters for the automated, objective and comprehensive identification of structures with confidence 10 to 100-fold above the noise level. We demonstrate high-fidelity and high-confidence localizations of nuclear pore complexes, vaults, ribosomes, proteasomes, fatty acid synthases, lipid membranes and microtubules, and individual subunits inside crowded eukaryotic cells. We provide software tools for the generic implementation of our method that is broadly applicable towards realizing visual proteomics.

Thinner is not always better: Optimizing cryo-lamellae for subtomogram averaging.

Cryo-electron tomography (cryo-ET) is a powerful method to elucidate subcellular architecture and to structurally analyze biomolecules in situ by subtomogram averaging, yet data quality critically depends on specimen thickness. Cells that are too thick for transmission imaging can be thinned into lamellae by cryo-focused ion beam (cryo-FIB) milling. Despite being a crucial parameter directly affecting attainable resolution, optimal lamella thickness has not been systematically investigated nor the extent of structural damage caused by gallium ions used for FIB milling. We thus systematically determined how resolution is affected by these parameters. We find that ion-induced damage does not affect regions more than 30 nanometers from either lamella surface and that up to ~180-nanometer lamella thickness does not negatively affect resolution. This shows that there is no need to generate very thin lamellae and lamella thickness can be chosen such that it captures cellular features of interest, thereby opening cryo-ET also for studies of large complexes.

RNA Captures More Cations than DNA: Insights from Molecular Dynamics Simulations.

The distribution of cations around nucleic acids is essential for a broad variety of processes ranging from DNA condensation and RNA folding to the detection of biomolecules in biosensors. Predicting the exact distribution of ions remains challenging since the distribution and, hence, a broad variety of nucleic acid properties depend on the salt concentration, the valency of the ions, and the ion type. Despite the importance, a general theory to quantify ion-specific effects for highly charged biomolecules is still lacking. Moreover, recent experiments reveal that despite their similar building blocks, DNA and RNA duplexes can react differently to the same ionic conditions. The aim of our current work is to provide a comprehensive set of molecular dynamics simulations using more than 180 µs of simulation time. For the mono- and divalent cations Li+, Na+, K+, Cs+, Ca2+, Sr2+, and Ba2+, the simulations allow us to reveal the ion-specific distributions and binding patterns for DNA and RNA duplexes. The microscopic insights from the simulations display the origin of ion-specificity and shed light on the question of why DNA and RNA show opposing behavior in the same ionic conditions. Finally, the detailed binding patterns from the simulations reveal why RNA can capture more cations than DNA.

Twisting DNA by salt.

The structure and properties of DNA depend on the environment, in particular the ion atmosphere. Here, we investigate how DNA twist -one of the central properties of DNA- changes with concentration and identity of the surrounding ions. To resolve how cations influence the twist, we combine single-molecule magnetic tweezer experiments and extensive all-atom molecular dynamics simulations. Two interconnected trends are observed for monovalent alkali and divalent alkaline earth cations. First, DNA twist increases monotonously with increasing concentration for all ions investigated. Second, for a given salt concentration, DNA twist strongly depends on cation identity. At 100 mM concentration, DNA twist increases as Na+ < K+ < Rb+ < Ba2+ < Li+ ≈ Cs+ < Sr2+ < Mg2+ < Ca2+. Our molecular dynamics simulations reveal that preferential binding of the cations to the DNA backbone or the nucleobases has opposing effects on DNA twist and provides the microscopic explanation of the observed ion specificity. However, the simulations also reveal shortcomings of existing force field parameters for Cs+ and Sr2+. The comprehensive view gained from our combined approach provides a foundation for understanding and predicting cation-induced structural changes both in nature and in DNA nanotechnology.

Extended magnesium and calcium force field parameters for accurate ion-nucleic acid interactions in biomolecular simulations.

Magnesium and calcium play an essential role in the folding and function of nucleic acids. To correctly describe their interactions with DNA and RNA in biomolecular simulations, an accurate parameterization is crucial. In most cases, the ion parameters are optimized based on a set of experimental solution properties such as solvation free energies, radial distribution functions, water exchange rates, and activity coefficient derivatives. However, the transferability of such bulk-optimized ion parameters to quantitatively describe biomolecular systems is limited. Here, we extend the applicability of our previous bulk-optimized parameters by including experimental binding affinities toward the phosphate oxygen on nucleic acids. In particular, we systematically adjust the combination rules that are an integral part of the pairwise interaction potentials of classical force fields. This allows us to quantitatively describe specific ion binding to nucleic acids without changing the solution properties in the most simple and efficient way. We show the advancement of the optimized Lorentz combination rule for two representative nucleic acid systems. For double-stranded DNA, the optimized combination rule for Ca2+ significantly improves the agreement with experiments, while the standard combination rule leads to unrealistically distorted DNA structures. For the add A-riboswitch, the optimized combination rule for Mg2+ improves the structure of two specifically bound Mg2+ ions as judged by the experimental distance to the binding site. Including experimental binding affinities toward specific ion binding sites on biomolecules, therefore, provides a promising perspective to develop a more accurate description of metal cations for biomolecular simulations.

Optimized Magnesium Force Field Parameters for Biomolecular Simulations with Accurate Solvation, Ion-Binding, and Water-Exchange Properties.

Magnesium ions play an essential role in many vital processes. To correctly describe their interactions in molecular dynamics simulations, an accurate parametrization is crucial. Despite the importance and considerable scientific effort, current force fields based on the commonly used 12-6 Lennard-Jones interaction potential fail to reproduce a variety of experimental solution properties. In particular, no parametrization exists so far that simultaneously reproduces the solvation free energy and the distance to the water oxygens in the first hydration shell. Moreover, current Mg2+ force fields significantly underestimate the rate of water exchange leading to unrealistically slow exchange kinetics. In order to make progress in the development of improved models, we systematically optimize the Mg2+ parameters in combination with the TIP3P water model in a much larger parameter space than previously done. The results show that a long-ranged interaction potential and modified Lorentz-Berthelot combination rules allow us to accurately reproduce multiple experimental properties including the solvation free energy, the distances to the oxygens of the first hydration shell, the hydration number, the activity coefficient derivative in MgCl2 solutions, the self-diffusion coefficient, and the binding affinity to the phosphate oxygen of RNA. Matching this broad range of thermodynamic properties, we present two sets of optimal parameters: MicroMg yields water exchange on the microsecond timescale in agreement with experiments. NanoMg yields water exchange on the nanosecond timescale facilitating the direct observation of ion-binding events. As shown for the example of the add A-riboswitch, the optimized parameters correctly reproduce the structure of specifically bound ions and permit the de novo prediction of Mg2+-binding sites in biomolecular simulations.

Hofmeister Series for Metal-Cation-RNA Interactions: The Interplay of Binding Affinity and Exchange Kinetics.

A large variety of physicochemical properties involving RNA depends on the type of metal cation present in solution. In order to gain microscopic insight into the origin of these ion specific effects, we apply molecular dynamics simulations to describe the interactions of metal cations and RNA. For the three most common ion binding sites on RNA, we calculate the binding affinities and exchange rates of eight different mono- and divalent metal cations. Our results reveal that binding sites involving phosphate groups preferentially bind metal cations with high charge density (such as Mg2+) in inner-sphere conformations while binding sites involving N7 or O6 atoms preferentially bind cations with low charge density (such as K+). The binding affinity therefore follows a direct Hofmeister series at the backbone but is reversed at the nucleobases leading to a high selectivity of ion binding sites on RNA. In addition, the exchange rates for cation binding cover almost 5 orders of magnitude, leading to a vastly different time scale for the lifetimes of contact pairs. Taken together, the site-specific binding affinities and the specific lifetime of contact pairs provide the microscopic explanation of ion specific effects observed in a wide variety of macroscopic RNA properties. Finally, combining the results from atomistic simulations with extended Poisson-Boltzmann theory allows us to predict the distribution of metal cations around double-stranded RNA at finite concentrations and to reproduce the results of ion counting experiments with good accuracy.

Coarse-Grained Double-Stranded RNA Model from Quantum-Mechanical Calculations.

A coarse-grained model for simulating structural properties of double-stranded RNA is developed with parameters obtained from quantum-mechanical calculations. This model follows previous parametrization for double-stranded DNA, which is based on mapping the all-atom picture to a coarse-grained four-bead scheme. Chemical and structural differences between RNA and DNA have been taken into account for the model development. The parametrization is based on simulations using density functional theory (DFT) on separate units of the RNA molecule without implementing experimental data. The total energy is decomposed into four terms of physical significance: hydrogen bonding interaction, stacking interactions, backbone interactions, and electrostatic interactions. The first three interactions are treated within DFT, whereas the last one is included within a mean field approximation. Our double-stranded RNA coarse-grained model predicts stable helical structures for RNA. Other characteristics, such as structural or mechanical properties are reproduced with a very good accuracy. The development of the coarse-grained model for RNA allows extending the spatial and temporal length scales accessed by computer simulations and being able to model RNA-related biophysical processes, as well as novel RNA nanostructures.