Detecting Protein-Ligand Interactions with Nitroxide Based Paramagnetic Cosolutes.

NMR spectroscopy techniques can provide important information about protein-ligand interactions. Here we tested an NMR approach which relies on the measurement of paramagnetic relaxation enhancements (PREs) arising from analogous cationic, anionic or neutral soluble nitroxide molecules, which distribute around the protein-ligand complex depending on near-surface electrostatic potentials. We applied this approach to two protein-ligand systems, interleukin-8 interacting with highly charged glycosaminoglycans and the SH2 domain of Grb2 interacting with less charged phospho-tyrosine tripeptides. The electrostatic potential around interleukin-8 and its changes upon binding of glycosaminoglycans could be derived from the PRE data and confirmed by theoretical predictions from Poisson-Boltzmann calculations. The ligand influence on the PREs and NMR-derived electrostatic potentials of Grb2 SH2 was localized to a narrow protein region which allowed the localization of the peptide binding pocket. Our analysis suggests that experiments with nitroxide cosolutes can be useful for investigating protein-ligand electrostatic interactions and mapping ligand binding sites.

Role for Carboxylic Acid Moiety in NSAIDs: Favoring the Binding at Site II of Bovine Serum Albumin.

The molecular structures of nonsteroidal anti-inflammatory drugs (NSAIDs) vary, but most contain a carboxylic acid functional group (RCOOH). This functional group is known to be related to the mechanism of cyclooxygenase inhibition and also causes side effects, such as gastrointestinal bleeding. This study proposes a new role for RCOOH in NSAIDs: facilitating the interaction at the binding site II of serum albumins. We used bovine serum albumin (BSA) as a model to investigate the interactions with ligands at site II. Using dansyl-proline (DP) as a fluorescent site II marker, we demonstrated that only negatively charged NSAIDs such as ibuprofen (IBP), naproxen (NPX), diflunisal (DFS), and ketoprofen (KTP) can efficiently displace DP from the albumin binding site. We confirmed the importance of RCOO by neutralizing IBP and NPX through esterification, which reduced the displacement of DP. The competition was also monitored by stopped-flow experiments. While IBP and NPX displaced DP in less than 1 s, the ester derivatives were ineffective. We also observed a higher affinity of negatively charged NSAIDs using DFS as a probe and ultrafiltration experiments. Molecular docking simulations showed an essential salt bridge between the positively charged residues Arg409 and Lys413 with RCOO-, consistent with the experimental findings. We performed a ligand dissociation pathway and corresponding energy analysis by applying molecular dynamics. The dissociation of NPX showed a higher free energy barrier than its ester. Apart from BSA, we conducted some experimental studies with human serum albumin, and similar results were obtained, suggesting a general effect for other mammalian serum albumins. Our findings support that the RCOOH moiety affects not only the mechanism of action and side effects but also the pharmacokinetics of NSAIDs.

Benchmarking Water Models in Molecular Dynamics of Protein-Glycosaminoglycan Complexes.

Glycosaminoglycans (GAGs) made of repeating disaccharide units intricately engage with proteins, playing a crucial role in the spatial organization of the extracellular matrix (ECM) and the transduction of biological signals in cells to modulate a number of biochemical processes. Exploring protein-GAG interactions reveals several challenges for their analysis, namely, the highly charged and periodic nature of GAGs, their multipose binding, and the abundance of the interfacial water molecules in the protein-GAG complexes. Most of the studies on protein-GAG interactions are conducted using the TIP3P water model, and there are no data on the effect of various water models on the results obtained in molecular dynamics (MD) simulations of protein-GAG complexes. Hence, it is essential to perform a systematic analysis of different water models in MD simulations for these systems. In this work, we aim to evaluate the properties of the protein-GAG complexes in MD simulations using different explicit: TIP3P, SPC/E, TIP4P, TIP4PEw, OPC, and TIP5P and implicit: IGB = 1, 2, 5, 7, and 8 water models to find out which of them are best suited to study the dynamics of protein-GAG complexes. The FF14SB and GLYCAM06 force fields were used for the proteins and GAGs, respectively. The interactions of several GAG types, such as heparin, chondroitin sulfate, and hyaluronic acid with basic fibroblast growth factor, cathepsin K, and CD44 receptor, respectively, are investigated. The observed variations in different descriptors used to study the binding in these complexes emphasize the relevance of the choice of water models for the MD simulation of these complexes.

Solvent Model Benchmark for Molecular Dynamics of Glycosaminoglycans.

In computational studies of glycosaminoglycans (GAGs), a group of anionic, periodic linear polysaccharides, so far there has been very little discussion about the role of solvent models in the molecular dynamics simulations of these molecules. Predominantly, the TIP3P water model is commonly used as one of the most popular explicit water models in general. However, there are numerous alternative explicit and implicit water models that are neglected in the computational research of GAGs. Since solvent-mediated interactions are particularly important for GAG dynamic and structural properties, it would be of great interest for the GAG community to establish the solvent model that is suited the best in terms of the quality of theoretically obtained GAG parameters and, at the same time, would be reasonably demanding in terms of computational resources required. In this study, heparin (HP) was simulated using five implicit and six explicit solvent models with the aim to find out how different solvent models influence HP's molecular descriptors in the molecular dynamics simulations. Here, we initiate the search for the most appropriate solvent representation for GAG systems and we hope to encourage other groups to contribute to this highly relevant subject.

Ligand binding of interleukin-8: a comparison of glycosaminoglycans and acidic peptides.

Recognition and binding of regulatory proteins to glycosaminoglycans (GAGs) from the extracellular matrix is a process of high biological importance. The interaction between negatively charged sulfate or carboxyl groups of the GAGs and clusters of basic amino acids on the protein is crucial in this binding process and it is believed that electrostatics represent the key factor for this interaction. However, given the rather undirected nature of electrostatics, it is important to achieve a clear understanding of its role in protein-GAG interactions and how specificity and selectivity in these systems can be achieved, when the classical key-lock binding motif is not applicable. Here, we compare protein binding of a highly charged heparin (HP) hexasaccharide with four de novo designed decapeptides of varying negative net charge. The charge density of these peptides was comparable to typical GAGs of the extracellular matrix. We used the regulatory protein interleukin-8 (IL-8) because its interactions with GAGs are well described. All four peptide ligands bind to the same epitope of IL-8 but show much weaker binding affinity as revealed in 1H-15N HSQC NMR titration experiments. Complementary molecular docking and molecular dynamics simulations revealed further atomistic details of the interaction mode of GAG versus peptide ligands. Overall, similar contributions to the binding energy and hydrogen bond formation are determined for HP and the highly charged peptides, suggesting that the entropic loss of the peptides upon binding likely account for the remarkably different affinity of GAG versus peptide ligands to IL-8.

Studying specificity in protein-glycosaminoglycan recognition with umbrella sampling.

In the past few decades, glycosaminoglycan (GAG) research has been crucial for gaining insights into various physiological, pathological, and therapeutic aspects mediated by the direct interactions between the GAG molecules and diverse proteins. The structural and functional heterogeneities of GAGs as well as their ability to bind specific proteins are determined by the sugar composition of the GAG, the size of the GAG chains, and the degree and pattern of sulfation. A deep understanding of the interactions in protein-GAG complexes is essential to explain their biological functions. In this study, the umbrella sampling (US) approach is used to pull away a GAG ligand from the binding site and then pull it back in. We analyze the binding interactions between GAGs of three types (heparin, desulfated heparan sulfate, and chondroitin sulfate) with three different proteins (basic fibroblast growth factor, acidic fibroblast growth factor, and cathepsin K). The main focus of our study was to evaluate whether the US approach is able to reproduce experimentally obtained structures, and how useful it can be for getting a deeper understanding of GAG properties, especially protein recognition specificity and multipose binding. We found that the binding free energy landscape in the proximity of the GAG native binding pose is complex and implies the co-existence of several binding poses. The sliding of a GAG chain along a protein surface could be a potential mechanism of GAG particular sequence recognition by proteins.

Rapamycin-induced autophagy in osteosarcoma cells is mediated via the biglycan/Wnt/ß-catenin signaling axis.

Biglycan is a class I secreted small leucine-rich proteoglycan (SLRP), which regulates signaling pathways connected to bone pathologies. Autophagy is a vital catabolic process with a dual role in cancer progression. Here, we show that biglycan inhibits autophagy in two osteosarcoma cell lines (P ≤ 0.001), while rapamycin-induced autophagy decreases biglycan expression in MG63 osteosarcoma cells and abrogates the biglycan-induced cell growth increase (P ≤ 0.001). Rapamycin also inhibits ß-catenin translocation to the nucleus, inhibiting the Wnt pathway (P ≤ 0.001) and reducing biglycan's colocalization with the Wnt coreceptor LRP6 (P ≤ 0.05). Furthermore, biglycan exhibits protective effects against the chemotherapeutic drug doxorubicin in MG63 OS cells through an autophagy-dependent manner (P ≤ 0.05). Cotreatment of these cells with rapamycin and doxorubicin enhances cells response to doxorubicin by decreasing biglycan (P ≤ 0.001) and ß-catenin (P ≤ 0.05) expression. Biglycan deficiency leads to increased caspase-3 activation (P ≤ 0.05), suggesting increased apoptosis of biglycan-deficient cells treated with doxorubicin. Computational models of LRP6 and biglycan complexes suggest that biglycan changes the receptor's ability to interact with other signaling molecules by affecting the interdomain bending angles in the receptor structure. Biglycan binding to LRP6 activates the Wnt pathway and ß-catenin nuclear translocation by disrupting ß-catenin degradation complex (P ≤ 0.01 and P ≤ 0.05). Interestingly, this mechanism is not followed in moderately differentiated, biglycan-nonexpressing U-2OS OS cells. To sum up, biglycan exhibits protective effects against the doxorubicin in MG63 OS cells by activating the Wnt signaling pathway and inhibiting autophagy.

Explicit solvent repulsive scaling replica exchange molecular dynamics (RS-REMD) in molecular modeling of protein-glycosaminoglycan complexes.

Glycosaminoglcyans (GAGs), linear anionic periodic polysaccharides, are crucial for many biologically relevant functions in the extracellular matrix. By interacting with proteins GAGs mediate processes such as cancer development, cell proliferation and the onset of neurodegenerative diseases. Despite this eminent importance of GAGs, they still represent a limited focus for the computational community in comparison to other classes of biomolecules. Therefore, there is a lack of modeling tools designed specifically for docking GAGs. One has to rely on existing docking software developed mostly for small drug molecules substantially differing from GAGs in their basic physico-chemical properties. In this study, we present an updated protocol for docking GAGs based on the Repulsive Scaling Replica Exchange Molecular Dynamics (RS-REMD) that includes explicit solvent description. The use of this water model improved docking performance both in terms of its accuracy and speed. This method represents a significant computational progress in GAG-related research.

Molecular Dynamics Approaches Dissect Molecular Mechanisms Underlying Methylene Blue-Glycosaminoglycan Interactions.

Glycosaminoglycans (GAGs) are a class of periodic anionic linear polysaccharides involved in a number of biologically relevant processes in the extracellular matrix via interactions with various types of molecules including proteins, peptides and small organic molecules. The metachromatic dye methylene blue (MB) is a GAG binding agent. This molecule possesses a tricyclic, monocationic phenothiazine ring system, while the terminal methyl groups attached to the nitrogen atoms bear the most positive charges of the cation and, therefore, represent potential binding sites for negatively charged GAGs. In this study, we rigorously explored molecular mechanisms underlying these interactions for several GAG types: heparin, heparan and chondroitin sulfates. We found that GAG-MB interactions are predominantly electrostatically driven, with the particularly important role of sulfate groups. MB oligomeric stack formation was favored in the presence of GAGs. Furthermore, the impact of MB binding on the conformation of GAGs was also evaluated. The novel results allow for better quantitative analytics of GAG composition in the studied biochemical systems using MB dye as a GAG-specific marker. Our data add to the knowledge on small molecule-GAG interactions and could be potentially useful for novel developments in drug design and putative disease therapies in which GAGs are involved.

Further analyses of APRIL/APRIL-receptor/glycosaminoglycan interactions by biochemical assays linked to computational studies.

A proliferation-inducing ligand (APRIL) is a member of the tumor necrosis factor superfamily. APRIL is quite unique in this superfamily for at least for two reasons: (i) it binds to glycosaminoglycans (GAGs) via its positively charged N-terminus; (ii) one of its signaling receptor, the transmembrane activator and CAML interactor (TACI), was also reported to bind GAGs. Here, as provided by biochemical evidences with the use of an APRIL deletion mutant linked to computational studies, APRIL-GAG interaction involved other regions than the APRIL N-terminus. Preferential interaction of APRIL with heparin followed by chondroitin sulfate E was confirmed by in silico analysis. Both computational and experimental approaches did not reveal the heparan sulfate binding to TACI. Together, computational results corroborated experiments contributing with atomistic details to the knowledge on this biologically relevant trimolecular system. Additionally, a high-throughput rigorous analysis of the free energy calculations data was performed to critically evaluate the applied computational methodologies.

Insights into the roles of charged residues in substrate binding and mode of action of mannuronan C-5 epimerase AlgE4.

Mannuronan C-5 epimerases catalyze the epimerization of monomer residues in the polysaccharide alginate, changing the physical properties of the biopolymer. The enzymes are utilized to tailor alginate to numerous biological functions by alginate-producing organisms. The underlying molecular mechanism that control the processive movement of the epimerase along the substrate chain is still elusive. To study this, we have used an interdisciplinary approach combining molecular dynamics simulations with experimental methods from mutant studies of AlgE4, where initial epimerase activity and product formation were addressed with nuclear magnetic resonance spectroscopy, and characteristics of enzyme-substrate interactions were obtained with isothermal titration calorimetry and optical tweezers. Positive charges lining the substrate-binding groove of AlgE4 appear to control the initial binding of poly-mannuronate, and binding also seems to be mediated by both electrostatic and hydrophobic interactions. After the catalytic reaction, negatively charged enzyme residues might facilitate dissociation of alginate from the positive residues, working like electrostatic switches, allowing the substrate to translocate in the binding groove. Molecular simulations show translocation increments of two monosaccharide units before the next productive binding event resulting in mannuronate and guluronate (MG)-block formation, with the epimerase moving with its N-terminus towards the reducing end of the alginate chain. Our results indicate that the charge pair R343-D345 might be directly involved in conformational changes of a loop that can be important for binding and dissociation. The computational and experimental approaches used in this study complement each other, allowing for a better understanding of individual residues' roles in binding and movement along the alginate chains.

Evaluation of replica exchange with repulsive scaling approach for docking glycosaminoglycans.

Glycosaminoglycans (GAGs), long linear periodic anionic polysaccharides, are key molecules in the extracellular matrix (ECM). Therefore, deciphering their role in the biologically relevant context is important for fundamental understanding of the processes ongoing in ECM and for establishing new strategies in the regenerative medicine. Although GAGs represent a number of computational challenges, molecular docking is a powerful tool for analysis of their interactions. Despite the recent development of GAG-specific docking approaches, there is plenty of room for improvement. Here, replica exchange molecular dynamics with repulsive scaling (REMD-RS) recently proved to be a successful approach for protein-protein complexes, was applied to dock GAGs. In this method, effective pairwise radii are increased in different Hamiltonian replicas. REMD-RS is shown to be an attractive alternative to classical docking approaches for GAGs. This work contributes to setting up of GAG-specific computational protocols and provides new insights into the nature of these biological systems.

Investigation of the structure of regulatory proteins interacting with glycosaminoglycans by combining NMR spectroscopy and molecular modeling - the beginning of a wonderful friendship.

The interaction of regulatory proteins with extracellular matrix or cell surface-anchored glycosaminoglycans (GAGs) plays important roles in molecular recognition, wound healing, growth, inflammation and many other processes. In spite of their high biological relevance, protein-GAG complexes are significantly underrepresented in structural databases because standard tools for structure determination experience difficulties in studying these complexes. Co-crystallization with subsequent X-ray analysis is hampered by the high flexibility of GAGs. NMR spectroscopy experiences difficulties related to the periodic nature of the GAGs and the sparse proton network between protein and GAG with distances that typically exceed the detection limit of nuclear Overhauser enhancement spectroscopy. In contrast, computer modeling tools have advanced over the last years delivering specific protein-GAG docking approaches successfully complemented with molecular dynamics (MD)-based analysis. Especially the combination of NMR spectroscopy in solution providing sparse structural constraints with molecular docking and MD simulations represents a useful synergy of forces to describe the structure of protein-GAG complexes. Here we review recent methodological progress in this field and bring up examples where the combination of new NMR methods along with cutting-edge modeling has yielded detailed structural information on complexes of highly relevant cytokines with GAGs.

Role of Oligosaccharide Chain Polarity in Protein-Glycosaminoglycan Interactions.

Glycosaminoglycans (GAGs) are long unbranched anionic polysaccharides made up of repetitive disaccharide units involved in biologically relevant processes in the extracellular matrix such as cell proliferation and communication. A GAG can be bound in antiparallel energetically comparable orientations on the protein surface, and these orientations are, therefore, difficult to distinguish both experimentally and computationally. In this study, for the first time we analyzed the impact of the GAG chain polarity on the interactions with Fibroblast Growth Factors-1 and -2 (FGF-1 and FGF-2). We performed a series of 1 µs molecular dynamics simulations of the FGF-1 and FGF-2 complexes with heparin (HP), a GAG representative, of different length. We analyzed the relationship between the HP orientation, energetic, and conformational space characteristics of FGF-1-HP and FGF-2-HP complexes. We concluded that HP can be bound by these proteins in the same binding sites but in different orientations, while the orientation present in the experimental structure might be favorable. Our data presented in this study provide a novel view on the impact of GAG polarity on the specificity of protein-GAG complex formation, which is an essential aspect for the proper understanding of the intermolecular interactions in these systems.

Modeling Protein-Glycosaminoglycan Complexes: Does the Size Matter?

Docking glycosaminoglycans (GAGs) has been challenging because of the complex nature of these long periodic linear and negatively charged polysaccharides. Although standard docking tools like Autodock3 are successful when docking GAGs up to hexameric length, they experience challenges to properly dock longer GAGs. Similar limitations concern other docking approaches typically developed for docking ligands of limited size to proteins. At the same time, most of more advanced docking approaches are challenging for a user who is inexperienced with complex in silico methodologies. In this work, we evaluate the binding energies of complexes with different lengths of GAGs using all-atom molecular dynamics simulations. Based on this analysis, we propose a new docking protocol for long GAGs that consists of conventional docking of short GAGs and further elongation with the use of a coarse-grained representation of the GAG parts not being in direct contact with its protein receptor. This method automated by a simple script is straightforward to use within the Autodock3 framework but also useful in combination with other standard docking tools. We believe that this method with some minor case-specific modifications could also be used for docking other linear charged polymers.

Computational insights into the role of calcium ions in protein-glycosaminoglycan systems.

Glycosaminoglycans (GAGs) are anionic, periodic, linear polysaccharides which are composed of periodic disaccharide units. They play a vital role in many biological processes ongoing in the extracellular matrix. In terms of computational approaches, GAGs are very challenging molecules due to their high flexibility, periodicity, predominantly electrostatic-driven nature of interactions with their protein counterparts and potential multipose binding. Furthermore, the molecular mechanisms underlying GAG-mediated interactions are not fully known yet, and experimental techniques alone are not always sufficient to gain insights into them. The aim of this study was to characterize protein-ion-GAG complexes for the systems where ions are directly involved in GAG binding. Molecular docking, molecular dynamics and free energy calculation approaches were applied to model and rigorously analyse the interactions between annexins (II and V), calcium ions (Ca2+) and heparin (HP). The computational data were examined and discussed in the context of the structural data previously reported by the crystallographic studies. The computational results confirm that the presence of Ca2+ has a tremendous impact on the annexin-HP binding site. This study provides a general computational pipeline to discover the complexity of protein-GAG interactions and helps to understand the role of ions involved at the atomic level. The limitations of the applied protocols are described and discussed pointing at the challenges persisting in the state-of-the-art in silico tools to study protein-ion-GAG systems.

Effect of Tetraphenylborate on Physicochemical Properties of Bovine Serum Albumin.

The binding interactions of bovine serum albumin (BSA) with tetraphenylborate ions ([B(Ph)4]-) have been investigated by a set of experimental methods (isothermal titration calorimetry, steady-state fluorescence spectroscopy, differential scanning calorimetry and circular dichroism spectroscopy) and molecular dynamics-based computational approaches. Two sets of structurally distinctive binding sites in BSA were found under the experimental conditions (10 mM cacodylate buffer, pH 7, 298.15 K). The obtained results, supported by the competitive interactions experiments of SDS with [B(Ph)4]- for BSA, enabled us to find the potential binding sites in BSA. The first site is located in the subdomain I A of the protein and binds two [B(Ph)4]- ions (logK(ITC)1 = 7.09 ± 0.10; ΔG(ITC)1 = -9.67 ± 0.14 kcal mol-1; ΔH(ITC)1 = -3.14 ± 0.12 kcal mol-1; TΔS(ITC)1 = -6.53 kcal mol-1), whereas the second site is localized in the subdomain III A and binds five ions (logK(ITC)2 = 5.39 ± 0.06; ΔG(ITC)2 = -7.35 ± 0.09 kcal mol-1; ΔH(ITC)2 = 4.00 ± 0.14 kcal mol-1; TΔS(ITC)2 = 11.3 kcal mol-1). The formation of the {[B(Ph)4]-}-BSA complex results in an increase in the thermal stability of the alfa-helical content, correlating with the saturation of the particular BSA binding sites, thus hindering its thermal unfolding.

Coarse-grained and atomic resolution biomolecular docking with the ATTRACT approach.

The ATTRACT protein-protein docking program has been employed to predict protein-protein complex structures in CAPRI rounds 38-45. For 11 out of 16 targets acceptable or better quality solutions have been submitted (~70%). It includes also several cases of peptide-protein docking and the successful prediction of the geometry of carbohydrate-protein interactions. The option of combining rigid body minimization and simultaneous optimization in collective degrees of freedom based on elastic network modes was employed and systematically evaluated. Application to a large benchmark set indicates a modest improvement in docking performance compared to rigid docking. Possible further improvements of the docking approach in particular at the scoring and the flexible refinement steps are discussed.

Role of Glycosaminoglycans in Procathepsin B Maturation: Molecular Mechanism Elucidated by a Computational Study.

Procathepsins are an inactive, immature form of cathepsins, predominantly cysteine proteases present in the extracellular matrix (ECM) and in lysosomes that play a key role in various biological processes such as bone resorption or intracellular proteolysis. The enzymatic activity of cathepsins can be mediated by glycosaminoglycans (GAGs), long unbranched periodic negatively charged polysaccharides found in ECM that take part in many biological processes such as anticoagulation, angiogenesis, and tissue regeneration. In addition to the known effects on mature cathepsins, GAGs can mediate the maturation process of procathepsins, in particular, procathepsin B. However, the detailed mechanism of this mediation at the molecular level is still unknown. In this study, for the first time, we aimed to unravel the role of GAGs in this process using computational approaches. We rigorously analyzed procathepsin B-GAG complexes in terms of their dynamics, energetics, and potential allosteric regulation. We revealed that GAGs can stabilize the conformation of the procathepsin B structure with the active site accessible for the substrate and concluded that GAGs most probably bind to procathepsin B once the zymogen adopts the enzymatically active conformation. Our data provided a novel mechanistic view of the maturation process of procathepsin B, while the approaches elaborated here might be useful to study other procathepsins. Furthermore, our data can serve as a rational guide for experimental work on procathepsin-GAG systems that are not characterized in vivo and in vitro yet.

Improved Consensus-Fragment Selection in Template-Assisted Prediction of Protein Structures with the UNRES Force Field in CASP13.

The method for protein-structure prediction, which combines the physics-based coarse-grained UNRES force field with knowledge-based modeling, has been developed further and tested in the 13th Community Wide Experiment on the Critical Assessment of Techniques for Protein Structure Prediction (CASP13). The method implements restraints from the consensus fragments common to server models. In this work, the server models to derive fragments have been chosen on the basis of quality assessment; a fully automatic fragment-selection procedure has been introduced, and Dynamic Fragment Assembly pseudopotentials have been fully implemented. The Global Distance Test Score (GDT_TS), averaged over our "Model 1" predictions, increased by over 10 units with respect to CASP12 for the free-modeling category to reach 40.82. Our "Model 1" predictions ranked 20 and 14 for all and free-modeling targets, respectively (upper 20.2% and 14.3% of all models submitted to CASP13 in these categories, respectively), compared to 27 (upper 21.1%) and 24 (upper 18.9%) in CASP12, respectively. For oligomeric targets, the Interface Patch Similarity (IPS) and Interface Contact Similarity (ICS) averaged over our best oligomer models increased from 0.28 to 0.36 and from 12.4 to 17.8, respectively, from CASP12 to CASP13, and top-ranking models of 2 targets (H0968 and T0997o) were obtained (none in CASP12). The improvement of our method in CASP13 over CASP12 was ascribed to the combined effect of the overall enhancement of server-model quality, our success in selecting server models and fragments to derive restraints, and improvements of the restraint and potential-energy functions.