Room-temperature crystallography using a microfluidic protein crystal array device and its application to protein–ligand complex structure analysis

Room-temperature (RT) protein crystallography provides significant information to elucidate protein function under physiological conditions. In particular, contrary to typical binding assays, X-ray crystal structure analysis of a protein–ligand complex can determine the three-dimensional (3D) configuration of its binding site. This allows the development of effective drugs by structure-based and fragment-based (FBDD) drug design. However, RT crystallography and RT crystallography-based protein–ligand complex analyses require the preparation and measurement of numerous crystals to avoid the X-ray radiation damage. Thus, for the application of RT crystallography to protein–ligand complex analysis, the simultaneous preparation of protein–ligand complex crystals and sequential X-ray diffraction measurement remain challenging. Here, we report an RT crystallography technique using a microfluidic protein crystal array device for protein–ligand complex structure analysis. We demonstrate the microfluidic sorting of protein crystals into microwells without any complicated procedures and apparatus, whereby the sorted protein crystals are fixed into microwells and sequentially measured to collect X-ray diffraction data. This is followed by automatic data processing to calculate the 3D protein structure. The microfluidic device allows the high-throughput preparation of the protein–ligand complex solely by the replacement of the microchannel content with the required ligand solution. We determined eight trypsin–ligand complex structures for the proof of concept experiment and found differences in the ligand coordination of the corresponding RT and conventional cryogenic structures. This methodology can be applied to easily obtain more natural structures. Moreover, drug development by FBDD could be more effective using the proposed methodology.

Room temperature protein crystallography and its application to protein–ligand complex structure analysis was demonstrated using a microfluidic protein crystal array device.     

Generating 3D molecules conditional on receptor binding sites with deep generative models

The goal of structure-based drug discovery is to find small molecules that bind to a given target protein. Deep learning has been used to generate drug-like molecules with certain cheminformatic properties, but has not yet been applied to generating 3D molecules predicted to bind to proteins by sampling the conditional distribution of protein–ligand binding interactions. In this work, we describe for the first time a deep learning system for generating 3D molecular structures conditioned on a receptor binding site. We approach the problem using a conditional variational autoencoder trained on an atomic density grid representation of cross-docked protein–ligand structures. We apply atom fitting and bond inference procedures to construct valid molecular conformations from generated atomic densities. We evaluate the properties of the generated molecules and demonstrate that they change significantly when conditioned on mutated receptors. We also explore the latent space learned by our generative model using sampling and interpolation techniques. This work opens the door for end-to-end prediction of stable bioactive molecules from protein structures with deep learning.

We generate 3D molecules conditioned on receptor binding sites by training a deep generative model on protein–ligand complexes. Our model uses the conditional receptor information to make chemically relevant changes to the generated molecules.     

Large scale relative protein ligand binding affinities using non-equilibrium alchemy

Ligand binding affinity calculations based on molecular dynamics (MD) simulations and non-physical (alchemical) thermodynamic cycles have shown great promise for structure-based drug design. However, their broad uptake and impact is held back by the notoriously complex setup of the calculations. Only a few tools other than the free energy perturbation approach by Schrödinger Inc. (referred to as FEP+) currently enable end-to-end application. Here, we present for the first time an approach based on the open-source software pmx that allows to easily set up and run alchemical calculations for diverse sets of small molecules using the GROMACS MD engine. The method relies on theoretically rigorous non-equilibrium thermodynamic integration (TI) foundations, and its flexibility allows calculations with multiple force fields. In this study, results from the Amber and Charmm force fields were combined to yield a consensus outcome performing on par with the commercial FEP+ approach. A large dataset of 482 perturbations from 13 different protein–ligand datasets led to an average unsigned error (AUE) of 3.64 ± 0.14 kJ mol⁻¹, equivalent to Schrödinger''s FEP+ AUE of 3.66 ± 0.14 kJ mol⁻¹. For the first time, a setup is presented for overall high precision and high accuracy relative protein–ligand alchemical free energy calculations based on open-source software.

Relative ligand binding affinity calculations based on molecular dynamics (MD) simulations and non-physical (alchemical) thermodynamic cycles have shown great promise for structure-based drug design.     

Rapid and accurate estimation of protein–ligand relative binding affinities using site-identification by ligand competitive saturation

Predicting relative protein–ligand binding affinities is a central pillar of lead optimization efforts in structure-based drug design. The site identification by ligand competitive saturation (SILCS) methodology is based on functional group affinity patterns in the form of free energy maps that may be used to compute protein–ligand binding poses and affinities. Presented are results obtained from the SILCS methodology for a set of eight target proteins as reported originally in Wang et al. (J. Am. Chem. Soc., 2015, 137, 2695–2703) using free energy perturbation (FEP) methods in conjunction with enhanced sampling and cycle closure corrections. These eight targets have been subsequently studied by many other authors to compare the efficacy of their method while comparing with the outcomes of Wang et al. In this work, we present results for a total of 407 ligands on the eight targets and include specific analysis on the subset of 199 ligands considered previously. Using the SILCS methodology we can achieve an average accuracy of up to 77% and 74% when considering the eight targets with their 199 and 407 ligands, respectively, for rank-ordering ligand affinities as calculated by the percent correct metric. This accuracy increases to 82% and 80%, respectively, when the SILCS atomic free energy contributions are optimized using a Bayesian Markov-chain Monte Carlo approach. We also report other metrics including Pearson''s correlation coefficient, Pearlman''s predictive index, mean unsigned error, and root mean square error for both sets of ligands. The results obtained for the 199 ligands are compared with the outcomes of Wang et al. and other published works. Overall, the SILCS methodology yields similar or better-quality predictions without a priori need for known ligand orientations in terms of the different metrics when compared to current FEP approaches with significant computational savings while additionally offering quantitative estimates of individual atomic contributions to binding free energies. These results further validate the SILCS methodology as an accurate, computationally efficient tool to support lead optimization and drug discovery.

Predicting relative protein–ligand binding affinities is a central pillar of lead optimization efforts in structure-based drug design.     

Alchemical absolute protein–ligand binding free energies for drug design

The recent advances in relative protein–ligand binding free energy calculations have shown the value of alchemical methods in drug discovery. Accurately assessing absolute binding free energies, although highly desired, remains a challenging endeavour, mostly limited to small model cases. Here, we demonstrate accurate first principles based absolute binding free energy estimates for 128 pharmaceutically relevant targets. We use a novel rigorous method to generate protein–ligand ensembles for the ligand in its decoupled state. Not only do the calculations deliver accurate protein–ligand binding affinity estimates, but they also provide detailed physical insight into the structural determinants of binding. We identify subtle rotamer rearrangements between apo and holo states of a protein that are crucial for binding. When compared to relative binding free energy calculations, obtaining absolute binding free energies is considerably more challenging in large part due to the need to explicitly account for the protein in its apo state. In this work we present several approaches to obtain apo state ensembles for accurate absolute ΔG calculations, thus outlining protocols for prospective application of the methods for drug discovery.

Molecular dynamics based absolute protein–ligand binding free energies can be calculated accurately and at large scale to facilitate drug discovery.     

Proteomimetic surface fragments distinguish targets by function

The fragment-centric design promises a means to develop complex xenobiotic protein surface mimetics, but it is challenging to find locally biomimetic structures. To address this issue, foldameric local surface mimetic (LSM) libraries were constructed. Protein affinity patterns, ligand promiscuity and protein druggability were evaluated using pull-down data for targets with various interaction tendencies and levels of homology. LSM probes based on H14 helices exhibited sufficient binding affinities for the detection of both orthosteric and non-orthosteric spots, and overall binding tendencies correlated with the magnitude of the target interactome. Binding was driven by two proteinogenic side chains and LSM probes could distinguish structurally similar proteins with different functions, indicating limited promiscuity. Binding patterns displayed similar side chain enrichment values to those for native protein–protein interfaces implying locally biomimetic behavior. These analyses suggest that in a fragment-centric approach foldameric LSMs can serve as useful probes and building blocks for undruggable protein interfaces.

Foldameric local surface mimetics (LSMs) detect spots at protein surfaces and are promising building blocks in a fragment-centric design of xenobiotic structures and protein–protein interaction inhibitors.     

Characterization of protein–ligand interactions by SABRE

Nuclear spin hyperpolarization through signal amplification by reversible exchange (SABRE), the non-hydrogenative version of para-hydrogen induced polarization, is demonstrated to enhance sensitivity for the detection of biomacromolecular interactions. A target ligand for the enzyme trypsin includes the binding motif for the protein, and at a distant location a heterocyclic nitrogen atom for interacting with a SABRE polarization transfer catalyst. This molecule, 4-amidinopyridine, is hyperpolarized with 50% para-hydrogen to yield enhancement values ranging from −87 and −34 in the ortho and meta positions of the heterocyclic nitrogen, to −230 and −110, for different solution conditions. Ligand binding is identified by flow-NMR, in a two-step process that separately optimizes the polarization transfer in methanol while detecting the interaction in a predominantly aqueous medium. A single scan Carr–Purcell–Meiboom–Gill (CPMG) experiment identifies binding by the change in R₂ relaxation rate. The SABRE hyperpolarization technique provides a cost effective means to enhance NMR of biological systems, for the identification of protein–ligand interactions and other applications.

Protein–ligand binding interactions are characterized by the para-H₂ based hyperpolarization technique SABRE and flow-NMR. Binding to the protein is identified by R₂ change of a ligand first interacting with the Ir polarization transfer catalyst.     

Combining Molecular Dynamic Information and an Aspherical-Atom Data Bank in the Evaluation of the Electrostatic Interaction Energy in Multimeric Protein-Ligand Complex: A Case Study for HIV-1 Protease

Computational analysis of protein–ligand interactions is of crucial importance for drug discovery. Assessment of ligand binding energy allows us to have a glimpse of the potential of a small organic molecule to be a ligand to the binding site of a protein target. Available scoring functions, such as in docking programs, all rely on equations that sum each type of protein–ligand interactions in order to predict the binding affinity. Most of the scoring functions consider electrostatic interactions involving the protein and the ligand. Electrostatic interactions constitute one of the most important part of total interactions between macromolecules. Unlike dispersion forces, they are highly directional and therefore dominate the nature of molecular packing in crystals and in biological complexes and contribute significantly to differences in inhibition strength among related enzyme inhibitors. In this study, complexes of HIV-1 protease with inhibitor molecules (JE-2147 and darunavir) were analyzed by using charge densities from the transferable aspherical-atom University at Buffalo Databank (UBDB). Moreover, we analyzed the electrostatic interaction energy for an ensemble of structures, using molecular dynamic simulations to highlight the main features of electrostatic interactions important for binding affinity.     

Density-functional tight-binding for phosphine-stabilized nanoscale gold clusters

We report a parameterization of the second-order density-functional tight-binding (DFTB2) method for the quantum chemical simulation of phosphine-ligated nanoscale gold clusters, metalloids, and gold surfaces. Our parameterization extends the previously released DFTB2 “auorg” parameter set by connecting it to the electronic parameter of phosphorus in the “mio” parameter set. Although this connection could technically simply be accomplished by creating only the required additional Au–P repulsive potential, we found that the Au 6p and P 3d virtual atomic orbital energy levels exert a strong influence on the overall performance of the combined parameter set. Our optimized parameters are validated against density functional theory (DFT) geometries, ligand binding and cluster isomerization energies, ligand dissociation potential energy curves, and molecular orbital energies for relevant phosphine-ligated Au_n clusters (n = 2–70), as well as selected experimental X-ray structures from the Cambridge Structural Database. In addition, we validate DFTB simulated far-IR spectra for several phosphine- and thiolate-ligated gold clusters against experimental and DFT spectra. The transferability of the parameter set is evaluated using DFT and DFTB potential energy surfaces resulting from the chemisorption of a PH₃ molecule on the gold (111) surface. To demonstrate the potential of the DFTB method for quantum chemical simulations of metalloid gold clusters that are challenging for traditional DFT calculations, we report the predicted molecular geometry, electronic structure, ligand binding energy, and IR spectrum of Au₁₀₈S₂₄(PPh₃)₁₆.

We report a parameterization of the density-functional tight-binding (DFTB) method for the accurate prediction of molecular, electronic and vibrational structure of phosphine-ligated nanoscale gold clusters, metalloids, and gold surfaces.     

In situ K-edge X-ray absorption spectroscopy of the ligand environment of single-site Au/C catalysts during acetylene hydrochlorination

The replacement of HgCl₂/C with Au/C as a catalyst for acetylene hydrochlorination represents a significant reduction in the environmental impact of this industrial process. Under reaction conditions atomically dispersed cationic Au species are the catalytic active site, representing a large-scale application of heterogeneous single-site catalysts. While the metal nuclearity and oxidation state under operating conditions has been investigated in catalysts prepared from aqua regia and thiosulphate, limited studies have focused on the ligand environment surrounding the metal centre. We now report K-edge soft X-ray absorption spectroscopy of the Cl and S ligand species used to stabilise these isolated cationic Au centres in the harsh reaction conditions. We demonstrate the presence of three distinct Cl species in the materials; inorganic Cl⁻, Au–Cl, and C–Cl and how these species evolve during reaction. Direct evidence of Au–S interactions is confirmed in catalysts prepared using thiosulfate precursors which show high stability towards reduction to inactive metal nanoparticles. This stability was clear during gas switching experiments, where exposure to C₂H₂ alone did not dramatically alter the Au electronic structure and consequently did not deactivate the thiosulfate catalyst.

In situ chlorine and sulphur XAS shows a dynamic ligand environment around cationic Au single-sites during acetylene hydrochlorination.     

Mechanochemical synthesis of glycine oligomers in a virtual rotational diamond anvil cell

Mechanochemistry of glycine under compression and shear at room temperature is predicted using quantum-based molecular dynamics (QMD) and a simulation design based on rotational diamond anvil cell (RDAC) experiments. Ensembles of high throughput semiempirical density functional tight binding (DFTB) simulations are used to identify chemical trends and bounds for glycine chemistry during rapid shear under compressive loads of up to 15.6 GPa. Significant chemistry is found to occur during compressive shear above 10 GPa. Recovered products consist of small molecules such as water, structural analogs to glycine, heterocyclic molecules, large oligomers, and polypeptides including the simplest polypeptide glycylglycine at up to 4% mass fraction. The population and size of oligomers generally increases with pressure. A number of oligomeric polypeptide precursors and intermediates are also identified that consist of two or three glycine monomers linked together through C–C, C–N, and/or C–O bridges. Even larger oligomers also form that contain peptide C–N bonds and exhibit branched structures. Many of the product molecules exhibit one or more chiral centers. Our simulations demonstrate that athermal mechanical compressive shearing of glycine is a plausible prebiotic route to forming polypeptides.

Compressive shearing forces can induce mechanochemical oligomerization reactions in glycine.     

Metal–ligand–Lewis acid multi-cooperative catalysis: a step forward in the Conia-ene reaction

An original multi-cooperative catalytic approach was developed by combining metal–ligand cooperation and Lewis acid activation. The [(SCS)Pd]₂ complex featuring a non-innocent indenediide-based ligand was found to be a very efficient and versatile catalyst for the Conia-ene reaction, when associated with Mg(OTf)₂. The reaction operates at low catalytic loadings under mild conditions with HFIP as a co-solvent. It works with a variety of substrates, including those bearing internal alkynes. It displays complete 5-exo vs. 6-endo regio-selectivity. In addition, except for the highly congested ^tBu-substituent, the reaction occurs with high Z vs. E stereo-selectivity, making it synthetically useful and complementary to known catalysts.

An original multi-cooperative catalytic approach was developed by combining metal–ligand cooperation and Lewis acid activation.     

Metal–ligand cooperativity in the soluble hydrogenase-1 from Pyrococcus furiosus

Metal–ligand cooperativity is an essential feature of bioinorganic catalysis. The design principles of such cooperativity in metalloenzymes are underexplored, but are critical to understand for developing efficient catalysts designed with earth abundant metals for small molecule activation. The simple substrate requirements of reversible proton reduction by the [NiFe]-hydrogenases make them a model bioinorganic system. A highly conserved arginine residue (R355) directly above the exogenous ligand binding position of the [NiFe]-catalytic core is known to be essential for optimal function because mutation to a lysine results in lower catalytic rates. To expand on our studies of soluble hydrogenase-1 from Pyrococcus furiosus (Pf SH1), we investigated the role of R355 by site-directed-mutagenesis to a lysine (R355K) using infrared and electron paramagnetic resonance spectroscopic probes sensitive to active site redox and protonation events. It was found the mutation resulted in an altered ligand binding environment at the [NiFe] centre. A key observation was destabilization of the Ni_a³⁺–C state, which contains a bridging hydride. Instead, the tautomeric Ni_a⁺–L states were observed. Overall, the results provided insight into complex metal–ligand cooperativity between the active site and protein scaffold that modulates the bridging hydride stability and the proton inventory, which should prove valuable to design principles for efficient bioinspired catalysts.

Metal–ligand cooperativity is an essential feature of bioinorganic catalysis.     

Identification of β-strand mediated protein–protein interaction inhibitors using ligand-directed fragment ligation

β-Strand mediated protein–protein interactions (PPIs) represent underexploited targets for chemical probe development despite representing a significant proportion of known and therapeutically relevant PPI targets. β-Strand mimicry is challenging given that both amino acid side-chains and backbone hydrogen-bonds are typically required for molecular recognition, yet these are oriented along perpendicular vectors. This paper describes an alternative approach, using GKAP/SHANK1 PDZ as a model and dynamic ligation screening to identify small-molecule replacements for tranches of peptide sequence. A peptide truncation of GKAP functionalized at the N- and C-termini with acylhydrazone groups was used as an anchor. Reversible acylhydrazone bond exchange with a library of aldehyde fragments in the presence of the protein as template and in situ screening using a fluorescence anisotropy (FA) assay identified peptide hybrid hits with comparable affinity to the GKAP peptide binding sequence. Identified hits were validated using FA, ITC, NMR and X-ray crystallography to confirm selective inhibition of the target PDZ-mediated PPI and mode of binding. These analyses together with molecular dynamics simulations demonstrated the ligands make transient interactions with an unoccupied basic patch through electrostatic interactions, establishing proof-of-concept that this unbiased approach to ligand discovery represents a powerful addition to the armory of tools that can be used to identify PPI modulators.

Dynamic ligation screening is used to identify acylhydrazone-linked peptide-fragment hybrids which bind to the SHANK1 PDZ domain with comparable affinity to the native GKAP peptide as shown by biophysical and structural analyses.     

Towards the rational design of Pt-based alloy catalysts for the low-temperature water-gas shift reaction: from extended surfaces to single atom alloys

The rational design of Pt-based catalysts for the low-temperature water-gas-shift (LT-WGS) reaction is an active research field because of its important role played in the fuel cell-based hydrogen economy, especially in mobile applications. Previous theoretical analyses have suggested that Pt alloys, leading to a weaker CO binding affinity than the Pt metal, could help alleviate CO poisoning and thus should be promising catalysts of the LT-WGS reaction. However, experimental research along this line was rather ineffective in the past decade. In the present work, we employed the state-of-the-art kinetic Monte Carlo (KMC) simulations to examine the influences of the electronic effect by introducing sub-surface alloys and/or core–shell structures, and the synergetic effect by introducing single atom alloys on the catalytic performance of Pt-alloy catalysts. Our KMC simulations have highlighted the importance of the OH binding affinity on the catalyst surfaces to reduce the barrier of water dissociation as the rate determining step, instead of the CO binding affinity as has been emphasized before in conventional mean-field kinetic models. Along this new direction of catalyst design, we found that Pt–Ru synergetic effects can significantly increase the activity of the Pt metal, leading to Ru_1–3@Pt alloys with a tetrahedron site of one surface-three subsurface Ru atoms on the Pt host, showing a turnover frequency of about five orders of magnitude higher than the Pt metal.

KMC simulations show that decreasing the barrier of H₂O decomposition is more beneficial than decreasing the CO binding affinity in LT-WGS, while the latter was overemphasized by MF-MKM. Here Ru_1–3@Pt alloy is proposed as a promising catalyst.     

Prediction of binding constants of protein ligands: A fast method for the prioritization of hits obtained from de novo design or 3D database search programs

A dataset of 82 protein–ligand complexes of known 3D structure and binding constant K_i was analysed to elucidate the important factors that determine the strength of protein–ligand interactions. The following parameters were investigated: the number and geometry of hydrogen bonds and ionic interactions between the protein and the ligand, the size of the lipophilic contact surface, the flexibility of the ligand, the electrostatic potential in the binding site, water molecules in the binding site, cavities along the protein–ligand interface and specific interactions between aromatic rings. Based on these parameters, a new empirical scoring function is presented that estimates the free energy of binding for a protein–ligand complex of known 3D structure. The function distinguishes between buried and solvent accessible hydrogen bonds. It tolerates deviations in the hydrogen bond geometry of up to 0.25 Å in the length and up to 30 °Cs in the hydrogen bond angle without penalizing the score. The new energy function reproduces the binding constants (ranging from 3.7 × 10^-2 M to 1 × 10^-14 M, corresponding to binding energies between -8 and -80 kJ/mol) of the dataset with a standard deviation of 7.3 kJ/mol corresponding to 1.3 orders of magnitude in binding affinity. The function can be evaluated very fast and is therefore also suitable for the application in a 3D database search or de novo ligand design program such as LUDI. The physical significance of the individual contributions is discussed.     

How bulky ligands control the chemoselectivity of Pd-catalyzed N-arylation of ammonia

Steric bulk has been recognized as a central design principle for supporting ligands in the widely utilized Buchwald–Hartwig amination. In a recent example, it was shown that a Pd-catalyst carrying a phosphine ligand can successfully aminate aryl halides using ammonia as the nitrogen source. Interestingly, the chemoselectivity of this reaction was found to depend on the steric demand of the phosphine ligand. Whereas a sterically less demanding phosphine affords diphenylamine as the major product, it was shown that the amination reaction can be stopped after the first amination to give aniline if a sterically more encumbering phosphine ligand is used. Density functional theory calculations were carried out to examine the relationship between the steric demand of the phosphine ligand and the chemoselectivity. It was found that the key feature that leads to the chemoselectivity is the ability of the phosphine ligand to rotate the biaryl moiety of the ligand away from the Pd-center upon amine addition to release some of the steric crowding from the Pd-coordination site.

Steric bulk has been recognized as a central design principle for ligands in the widely utilized Buchwald–Hartwig amination. This mechanistic study reveals how this steric effect manipulates the reaction pathway and determines the chemoselectivity.     

Query-guided protein–protein interaction inhibitor discovery

Protein–protein interactions (PPIs) are central to biological mechanisms, and can serve as compelling targets for drug discovery. Yet, the discovery of small molecule inhibitors of PPIs remains challenging given the large and typically shallow topography of the interacting protein surfaces. Here, we describe a general approach to the discovery of orthosteric PPI inhibitors that mimic specific secondary protein structures. Initially, hot residues at protein–protein interfaces are identified in silico or from experimental data, and incorporated into secondary structure-based queries. Virtual libraries of small molecules are then shape-matched against the queries, and promising ligands docked to target proteins. The approach is exemplified experimentally using two unrelated PPIs that are mediated by an α-helix (p53/hDM2) and a β-strand (GKAP/SHANK1-PDZ). In each case, selective PPI inhibitors are discovered with low μM activity as determined by a combination of fluorescence anisotropy and ¹H–¹⁵N HSQC experiments. In addition, hit expansion yields a series of PPI inhibitors with defined structure–activity relationships. It is envisaged that the generality of the approach will enable discovery of inhibitors of a wide range of unrelated secondary structure-mediated PPIs.

Small-molecule protein–protein interaction inhibitors were prioritised on the basis of shape similarity to secondary structure-based queries incorporating hot-spot residues.     

Metal–ligand bond strength determines the fate of organic ligands on the catalyst surface during the electrochemical CO2 reduction reaction

Colloidally synthesised nanocrystals (NCs) are increasingly utilised as catalysts to drive both thermal and electrocatalytic reactions. Their well-defined size and shape, controlled by organic ligands, are ideal to identify the parameters relevant to the activity, selectivity and stability in catalysis. However, the impact of the native surface ligands during catalysis still remains poorly understood, as does their fate. CuNCs are among the state-of-the-art catalysts for the electrochemical CO₂ reduction reaction (CO₂RR). In this work, we study CuNCs that are capped by different organic ligands to investigate their impact on the catalytic properties. We show that the latter desorb from the surface at a cathodic potential that depends on their binding strength with the metal surface, rather than their own electroreduction potentials. By monitoring the evolving surface chemistry in situ, we find that weakly bound ligands desorb very rapidly while strongly bound ligands impact the catalytic performance. This work provides a criterion to select labile ligands versus ligands that will persist on the surface, thus offering opportunity for interface design.

The metal–ligand binding strength is a key parameter in determining the role and fate of the surface ligands on nanoparticle catalysts during the electrochemical CO₂ reduction reaction.