Reconstitution of a native-like SH2 domain from disordered peptide fragments examined by multidimensional heteronuclear NMR

The N-terminal SH2 domain from the p85alpha subunit of phosphatidylinositol 3' kinase is cleaved specifically into 9- and 5-kD fragments by limited proteolytic digestion with trypsin. The noncovalent SH2 domain complex and its constituent tryptic peptides have been investigated using high-resolution heteronuclear magnetic resonance (NMR). These studies have established the viability of the SH2 domain as a fragment complementation system. The individual peptide fragments are predominantly unstructured in solution. In contrast, the noncovalent 9-kD + 5-kD complex shows a native-like (1)H-(15)N HSQC spectrum, demonstrating that the two fragments fold into a native-like structure on binding. Chemical shift analysis of the noncovalent complex compared to the native SH2 domain reveals that the highest degree of perturbation in the structure occurs at the cleavage site within a flexible loop and along the hydrophobic interface between the two peptide fragments. Mapping of these chemical shift changes on the structure of the domain reveals changes consistent with the reduction in affinity for the target peptide ligand observed in the noncovalent complex relative to the intact protein. The 5-kD fragment of the homologous Src protein is incapable of structurally complementing the p85 9-kD fragment, either in complex formation or in the context of the full-length protein. These high-resolution structural studies of the SH2 domain fragment complementation features establish the suitability of the system for further protein-folding and design studies.     

Binding similarity network of ligand

The protein and ligand interaction takes an important part in protein function. Both ligand and its binding site are essential components for understanding how the protein-ligand complex functions. Until now, there have been many studies about protein function and evolution, but they usually lacked ligand information. Accordingly, in this study, we tried to answer the following questions: how much ligand and binding site are associated with protein function, and how ligands themselves are related to each other in terms of binding site. To answer the questions, we presented binding similarity network of ligand. Through the network analysis, we attempted to reveal systematic relationship between the ligand and binding site. The results showed that ligand binding site and function were closely related (conservation ratio, 81%). We also showed conservative tendency of function in line with ligand structure similarity with some exceptional cases. In addition, the binding similarity network of ligand revealed scale-free property to some degree like other biological networks. Since most nodes formed highly connected cluster, a clustering coefficient was very high compared with random. All the highly connected ligands (hubs) were involved in various functions forming large cluster and tended to act as a bridge between modular clusters in the network.     

Interaction of Hoechst 33258 and echinomycin with nucleosomal DNA fragments containing isolated ligand binding sites

We have examined the interaction of Hoechst 33258 and echinomycin with nucleosomal DNA fragments which contain isolated ligand binding sites. A 145 base pair fragment was prepared on the basis of the sequence of tyrT DNA, which contained no CpG or (A/T)(4) binding sites for these ligands. Isolated binding sites were introduced into this fragment at discrete locations where the minor groove is known to face toward or away from the protein core when reconstituted onto nucleosome core particles. The interaction of ligands with target sites on these nucleosomal DNA fragments was assessed by DNase I footprinting. We find that Hoechst 33258 can bind to single nucleosomal sites which face both toward and away from the protein core, without affecting the nucleosome structure. Hoechst binding is also observed on nucleosomal fragments which contain two or more drug binding sites, though in these cases the footprints are accompanied by the presence of new cleavage products in positions which suggest that the ligand has caused a proportion of the DNA molecules to adopt a new rotational positioning on the protein surface. Hoechst 33258 does not affect nucleosome reconstitution with any of these fragments. In contrast, the bifunctional intercalating antibiotic echinomycin is not able to bind to single nucleosomal CpG sites. Echinomycin footprints are observed on nucleosomal fragments containing two or more CpG sites, but there are no changes in the cleavage patterns in the remainder of the fragment. Echinomycin abolishes nucleosome reconstitution when included in the reconstitution mixture.     

Combining evolutionary and structural information for local protein structure prediction

We study the effects of various factors in representing and combining evolutionary and structural information for local protein structural prediction based on fragment selection. We prepare databases of fragments from a set of non-redundant protein domains. For each fragment, evolutionary information is derived from homologous sequences and represented as estimated effective counts and frequencies of amino acids (evolutionary frequencies) at each position. Position-specific amino acid preferences called structural frequencies are derived from statistical analysis of discrete local structural environments in database structures. Our method for local structure prediction is based on ranking and selecting database fragments that are most similar to a target fragment. Using secondary structure type as a local structural property, we test our method in a number of settings. The major findings are: (1) the COMPASS-type scoring function for fragment similarity comparison gives better prediction accuracy than three other tested scoring functions for profile-profile comparison. We show that the COMPASS-type scoring function can be derived both in the probabilistic framework and in the framework of statistical potentials. (2) Using the evolutionary frequencies of database fragments gives better prediction accuracy than using structural frequencies. (3) Finer definition of local environments, such as including more side-chain solvent accessibility classes and considering the backbone conformations of neighboring residues, gives increasingly better prediction accuracy using structural frequencies. (4) Combining evolutionary and structural frequencies of database fragments, either in a linear fashion or using a pseudocount mixture formula, results in improvement of prediction accuracy. Combination at the log-odds score level is not as effective as combination at the frequency level. This suggests that there might be better ways of combining sequence and structural information than the commonly used linear combination of log-odds scores. Our method of fragment selection and frequency combination gives reasonable results of secondary structure prediction tested on 56 CASP5 targets (average SOV score 0.77), suggesting that it is a valid method for local protein structure prediction. Mixture of predicted structural frequencies and evolutionary frequencies improve the quality of local profile-to-profile alignment by COMPASS.     

Structure-activity relationships by mass spectrometry: identification of novel MMP-3 inhibitors

A novel class of nonpeptide inhibitors of stromelysin (MMP-3) has been discovered with the use of mass spectrometry. The method relies on the development of structure-activity relationships by mass spectrometry (SAR by MS) and utilizes information derived from the binding of known inhibitors to identify novel inhibitors of a target protein with a minimum of synthetic effort. Noncovalent complexes of known inhibitors with a target protein are analyzed; these inhibitors are deconstructed into sets of fragments which compete for common or overlapping binding sites on the target protein. The binding of each fragment set can be studied independently. With the use of competition studies, novel members of each fragment set are identified from compound libraries that bind to the same site on the target protein. A novel inhibitor of the target protein was then constructed by chemically linking a combination of members of each fragment set in a manner guided by the proximity and orientation of the fragments derived from the known inhibitors. In the case of stromelysin, a novel inhibitor composed of favorably linked fragments was observed to form a 1:1 complex with stromelysin. Compounds that were not linked appropriately formed higher order complexes with stoichiometries of 2:1 or greater. These linked molecules were subsequently assessed for their ability to block stromelysin function in a chromogenic substrate assay.     

Automatic identification and representation of protein binding sites for molecular docking.

Molecular docking is a popular way to screen for novel drug compounds. The method involves aligning small molecules to a protein structure and estimating their binding affinity. To do this rapidly for tens of thousands of molecules requires an effective representation of the binding region of the target protein. This paper presents an algorithm for representing a protein's binding site in a way that is specifically suited to molecular docking applications. Initially the protein's surface is coated with a collection of molecular fragments that could potentially interact with the protein. Each fragment, or probe, serves as a potential alignment point for atoms in a ligand, and is scored to represent that probe's affinity for the protein. Probes are then clustered by accumulating their affinities, where high affinity clusters are identified as being the "stickiest" portions of the protein surface. The stickiest cluster is used as a computational binding "pocket" for docking. This method of site identification was tested on a number of ligand-protein complexes; in each case the pocket constructed by the algorithm coincided with the known ligand binding site. Successful docking experiments demonstrated the effectiveness of the probe representation.     

Synthetic DNA-bindings ligands containing reaction centers specific for AT- and GC-pairs in DNA

In the present communication design, synthesis and DNA binding activities of three bis-netropsins and two netropsin analogs containing two N-propylpyrrolecarboxamide fragments linked covalently to peptides Gly-Gly-(analog I) and Val-Val-Val-Gly-Gly-(analog II) are reported. Each bis-netropsin consists of two netropsin-like fragments attached to peptides -Gly-Cys-Gly-NH2 (compound IIIa), H-Gly-Cys-Gly-Gly-Gly-(compound IV) or Gly-Cys-Sar-NH2 (compound IIIb) which are linked symmetrically via S-S bonds. Physico-chemical studies show that each bis-netropsin carries 6 AT-specific reaction centers and covers approximately 10 base pairs upon binding to poly(dA).poly(dT). This indicates that two netropsin-like fragments of the bis-netropsin molecule are implicated in specific interaction with DNA base pairs. The peptide fragments of bis-netropsins IIIa and IV form small beta-sheets containing two-GC-specific reaction centers. The DNase I cleavage patterns of bis-netropsin-DNA complexes visualized by high resolution gel electrophoresis show that the preferred binding sites for bis-netropsins IIIa and IV are identical and contain two runs of three or more AT pairs separated by two GC pairs. Specificity determinants of netropsin analog II binding in the beta-associated dimeric form are identical to those of bis-netropsin IIIa thereby indicating that there is a similarity in the structure of complexes formed by these ligands with DNA. In the monomeric form analog II exhibits binding specificity identical to that of analog I. Replacement of C-terminal glycine residues by sarcosines in the peptide fragments of bis-netropsin IIIa leads to a decrease in the affinity of ligand for DNA.     

A multiple-start Monte Carlo docking method.

We present a method to search for possible binding modes of molecular fragments at a specific site of a potential drug target of known structure. Our method is based on a Monte Carlo (MC) algorithm applied to the translational and rotational degrees of freedom of the probe fragment. Starting from a randomly generated initial configuration, favorable binding modes are generated using a two-step process. An MC run is first performed in which the energy in the Metropolis algorithm is substituted by a score function that measures the average distance of the probe to the target surface. This has the effect of making buried probes move toward the target surface and also allows enhanced sampling of deep pockets. In a second MC run, a pairwise atom potential function is used, and the temperature parameter is slowly lowered during the run (Simulated Annealing). We repeat this procedure starting from a large number of different randomly generated initial configurations in order to find all energetically favorable docking modes in a specified region around the target. We test this method using two inhibitor-receptor systems: Streptomyces griseus proteinase B in complex with the third domain of the ovomucoid inhibitor from turkey, and dihydrofolate reductase from E. coli in complex with methotrexate. The method could consistently reproduce the complex found in the crystal structure searching from random initial positions in cubes ranging from 25 A to 50 A about the binding site. In the case of SGPB, we were also successful in docking to the native structure. In addition, we were successful in docking small probes in a search that included the entire protein surface.     

Knowledge-based Fragment Binding Prediction

Target-based drug discovery must assess many drug-like compounds for potential activity. Focusing on low-molecular-weight compounds (fragments) can dramatically reduce the chemical search space. However, approaches for determining protein-fragment interactions have limitations. Experimental assays are time-consuming, expensive, and not always applicable. At the same time, computational approaches using physics-based methods have limited accuracy. With increasing high-resolution structural data for protein-ligand complexes, there is now an opportunity for data-driven approaches to fragment binding prediction. We present FragFEATURE, a machine learning approach to predict small molecule fragments preferred by a target protein structure. We first create a knowledge base of protein structural environments annotated with the small molecule substructures they bind. These substructures have low-molecular weight and serve as a proxy for fragments. FragFEATURE then compares the structural environments within a target protein to those in the knowledge base to retrieve statistically preferred fragments. It merges information across diverse ligands with shared substructures to generate predictions. Our results demonstrate FragFEATURE''s ability to rediscover fragments corresponding to the ligand bound with 74% precision and 82% recall on average. For many protein targets, it identifies high scoring fragments that are substructures of known inhibitors. FragFEATURE thus predicts fragments that can serve as inputs to fragment-based drug design or serve as refinement criteria for creating target-specific compound libraries for experimental or computational screening.     

Crystal Structure of the Tetrameric Fibrinogen-like Recognition Domain of Fibrinogen C Domain Containing 1 (FIBCD1) Protein

The high resolution crystal structures of a recombinant fragment of the C-terminal fibrinogen-like recognition domain of FIBCD1, a vertebrate receptor that binds chitin, have been determined. The overall tetrameric structure shows similarity in structure and aggregation to the horseshoe crab innate immune protein tachylectin 5A. The high affinity ligand N-acetylmannosamine (ManNAc) binds in the S1 site, predominantly via the acetyl group with the oxygen and acetamide nitrogen hydrogen-bonded to the protein and the methyl group inserted into a hydrophobic pocket. The binding of the ManNAc pyranose ring differs markedly between the two independent subunits, but in all structures the binding of the N-acetyl group is conserved. In the native structure, a crystal contact results in one of the independent protomers binding the first GlcNAc of the Asn³⁴⁰ N-linked glycan on the other independent protomer. In the ligand-bound structure this GlcNAc is replaced by the higher affinity ligand ManNAc. In addition, a sulfate ion has been modeled into the electron density at a location similar to the S3 binding site in L-ficolin, whereas in the native structure an acetate ion has been placed in the S1 N-acetyl binding site, and a sulfate ion has been placed adjacent to this site. These ion binding sites are ideally placed to receive the N-acetyl and sulfate groups of sulfated GalNAc residues of glycosaminoglycans such as chondroitin and dermatan sulfate. Together, these structures give insight into important determinants of ligand selectivity, demonstrating versatility in recognition and binding while maintaining conservation in N-acetyl and calcium binding.     

On the pathway of forming enzymatically productive ligand-protein complexes in lactate dehydrogenase

We have carried out a series of studies on the binding of a substrate mimic to the enzyme lactate dehydrogenase (LDH) using advanced kinetic approaches, which begin to provide a molecular picture of the dynamics of ligand binding for this protein. Binding proceeds via a binding-competent subpopulation of the nonligated form of the protein (the LDH/NADH binary complex) to form a protein-ligand encounter complex. The work here describes the collapse of the encounter complex to form the catalytically competent Michaelis complex. Isotope-edited static Fourier transform infrared studies on the bound oxamate protein complex reveal two kinds of oxamate environments: 1), a major populated structure wherein all significant hydrogen-bonding patterns are formed at the active site between protein and bound ligand necessary for the catalytically productive Michaelis complex and 2), a minor structure in a configuration of the active site that is unfavorable to carry out catalyzed chemistry. This latter structure likely simulates a dead-end complex in the reaction mixture. Temperature jump isotope-edited transient infrared studies on the binding of oxamate with LDH/NADH suggest that the evolution of the encounter complex between LDH/NADH and oxamate collapses via a branched reaction pathway to form the major and minor bound species. The production of the catalytically competent protein-substrate complex has strong similarities to kinetic pathways found in two-state protein folding processes. Once the encounter complex is formed between LDH/NADH and substrate, the ternary protein-ligand complex appears to “fold” to form a compact productive complex in an all or nothing like fashion with all the important molecular interactions coming together at the same time.     

A novel computational tool for automated structure-based drug design

The computer program LUDI for automated structure-based drug design is described. The program constructs possible new ligands for a given protein of known three-dimensional structure. This novel approach is based upon rules about energetically favourable non-bonded contact geometries between functional groups of the protein and the ligand which are derived from a statistical analysis of crystal packings of organic molecules. In a first step small fragments are docked into the protein binding site in such a way that hydrogen bonds and ionic interactions can be formed with the protein and hydrophobic pockets are filled with lipophilic groups of the ligands. The program can then append further fragments onto a previously positioned fragments or onto an already existing ligand (e.g., a lead structure that one seeks to improve). It is also possible to link several fragments together by bridge fragments to form a complete molecule. All putative ligands retrieved or constructed by LUDI are scored. We use a simple scoring function that was fitted to experimentally determined binding constants of protein–ligand complexes. LUCI is a very fast program with typical execution times of 1–5 min on a work station and is therefore suitable for interactive usage.     

The protein S-binding site localized to the central core of C4b-binding protein

Human C4b-binding protein (C4BP) is a regulator of the classical pathway of the complement system. It appears in two forms in plasma, as free protein and in a noncovalent complex with the vitamin K-dependent coagulation protein, protein S. In the electron microscope C4BP has a spider-like structure with a central core and seven extended tentacles, each of which has a binding site for C4b, although the protein S-binding site has not been unequivocally pinpointed. C4BP was subjected to chymotrypsin digestion which yielded two major fragments, one of 160 kDa representing the central core, and one of 48 kDa representing the cleaved-off tentacles. We have now localized the protein S-binding site to the 160-kDa central core fragment. Using immunoblotting with a panel of polyclonal antisera, the isolated central core was shown to be completely devoid of 48-kDa fragments. The protein S-binding site was susceptible to proteolysis by chymotrypsin, but was protected by a molar excess of protein S included during the proteolysis. The 160-kDa central core fragment consisted of identical, disulfide-linked 25-kDa peptides and a proper disulfide bond arrangement was crucial to protein S binding. Using a direct binding assay it was shown that the isolated central core had the same affinity for protein S as intact C4BP.     

Ligand Binding Site Similarity Identification Based on Chemical and Geometric Similarity

The similarity comparison of binding sites based on amino acid between different proteins can facilitate protein function identification. However, Binding site usually consists of several crucial amino acids which are frequently dispersed among different regions of a protein and consequently make the comparison of binding sites difficult. In this study, we introduce a new method, named as chemical and geometric similarity of binding site (CGS-BSite), to compute the ligand binding site similarity based on discrete amino acids with maximum-weight bipartite matching algorithm. The principle of computing the similarity is to find a Euclidean Transformation which makes the similar amino acids approximate to each other in a geometry space, and vice versa. CGS-BSite permits site and ligand flexibilities, provides a stable prediction performance on the flexible ligand binding sites. Binding site prediction on three test datasets with CGS-BSite method has similar performance to Patch-Surfer method but outperforms other five tested methods, reaching to 0.80, 0.71 and 0.85 based on the area under the receiver operating characteristic curve, respectively. It performs a marginally better than Patch-Surfer on the binding sites with small volume and higher hydrophobicity, and presents good robustness to the variance of the volume and hydrophobicity of ligand binding sites. Overall, our method provides an alternative approach to compute the ligand binding site similarity and predict potential special ligand binding sites from the existing ligand targets based on the target similarity.     

Fragment screening using capillary electrophoresis (CEfrag) for hit identification of heat shock protein 90 ATPase inhibitors

CEfrag is a new fragment screening technology based on affinity capillary electrophoresis (ACE). Here we report on the development of a mobility shift competition assay using full-length human heat shock protein 90α (Hsp90α), radicicol as the competitor probe ligand, and successful screening of the Selcia fragment library. The CEfrag assay was able to detect weaker affinity (IC(50) >500 μM) fragments than were detected by a fluorescence polarization competition assay using FITC-labeled geldanamycin. The binding site of selected fragments was determined by co-crystallization with recombinant Hsp90α N-terminal domain and X-ray analysis. The results of this study confirm that CEfrag is a sensitive microscale technique enabling detection of fragments binding to the biological target in near-physiological solution.     

Induced bending of plasmid pLS1 DNA by the plasmid-encoded protein RepA

The broad host range streptococcal plasmid pLS1 encodes for a 5.1-kDa repressor protein, RepA. This protein has affinity for DNA (linear or supercoiled) and is translated from the same mRNA as the replication initiator protein RepB. By gel retardation assays, we observed that RepA shows specificity for binding to the plasmid HinfID fragment, which includes the target of the protein. The target of RepA within the plasmid DNA molecule has been located around the plasmid single site ApaLI. This site is included in a region that contains the promoter for the repA and repB genes and is contiguous to the plasmid ori(+). A complex sequence-directed DNA curvature is observed in this region of pLS1. Upon addition of RepA to plasmid linear DNA or to circularly permuted restriction fragments, this intrinsic curvature was greatly enhanced. Thus, a strong RepA-induced bending could be located in the vicinity of the ApaLI site. Visualization of the bent DNA was achieved by electron microscopy of complexes between RepA and plasmid DNA fragments containing the RepA target.     

The change of DNA structure by specific binding of the cAMP receptor protein from rotation diffusion and dichroism measurements.

The structure of complexes formed between cAMP receptor protein (CRP) and various restriction fragments from the promoter region of the lactose operon has been analysed by measurements of electrodichroism. Binding of CRP to a 62-bp fragment containing the major site leads to an increase of the rotation time constant from 0.33 to 0.43 microseconds; addition of cAMP to the complex induces a decrease to 0.25 microseconds. Similar data are obtained for a 80-bp fragment containing the operator site; however, in this case the decrease of the rotation time for the specific complex is only observed when the salt concentration is increased from 3 to 13 mM. A 203-bp fragment containing both sites showed a corresponding change after pre-incubation at 50 mM salt. The salt dependence of the rotation time for the specific complex formed with the 203-bp fragment also indicates that a compact structure is formed at 13 mM salt, whereas the structure is not as compact at 3 mM salt. A 98-bp fragment without specific CRP sites did not reveal changes corresponding to those observed for the specific fragments. The rotation time constants together with the dichroism amplitudes indicate that binding of CRP to specific sites in the presence of cAMP leads to the formation of compact structures, which are consistent with bending of DNA helices. The observed strong salt dependence of the structure is apparently due to electrostatic repulsion between adjoining helix segments.     

Structure of integrin alpha5beta1 in complex with fibronectin

The membrane-distal headpiece of integrins has evolved to specifically bind large extracellular protein ligands, but the molecular architecture of the resulting complexes has not been determined. We used molecular electron microscopy to determine the three-dimensional structure of the ligand-binding headpiece of integrin alpha5beta1 complexed with fragments of its physiological ligand fibronectin. The density map for the unliganded alpha5beta1 headpiece shows a 'closed' conformation similar to that seen in the alphaVbeta3 crystal structure. By contrast, binding to fibronectin induces an 'open' conformation with a dramatic, approximately 80 degrees change in the angle of the hybrid domain of the beta subunit relative to its I-like domain. The fibronectin fragment binds to the interface between the beta-propeller and I-like domains in the integrin headpiece through the RGD-containing module 10, but direct contact of the synergy-region-containing module 9 to integrin is not evident. This finding is corroborated by kinetic analysis of real-time binding data, which shows that the synergy site greatly enhances k(on) but has little effect on the stability or k(off) of the complex.     

Discovery of Fragment Molecules That Bind the Human Peroxiredoxin 5 Active Site

The search for protein ligands is a crucial step in the inhibitor design process. Fragment screening represents an interesting method to rapidly find lead molecules, as it enables the exploration of a larger portion of the chemical space with a smaller number of compounds as compared to screening based on drug-sized molecules. Moreover, fragment screening usually leads to hit molecules that form few but optimal interactions with the target, thus displaying high ligand efficiencies. Here we report the screening of a homemade library composed of 200 highly diverse fragments against the human Peroxiredoxin 5 protein. Peroxiredoxins compose a family of peroxidases that share the ability to reduce peroxides through a conserved cysteine. The three-dimensional structures of these enzymes ubiquitously found throughout evolution have been extensively studied, however, their biological functions are still not well understood and to date few inhibitors have been discovered against these enzymes. Six fragments from the library were shown to bind to the Peroxiredoxin 5 active site and ligand-induced chemical shift changes were used to drive the docking of these small molecules into the protein structure. The orientation of the fragments in the binding pocket was confirmed by the study of fragment homologues, highlighting the role of hydroxyl functions that hang the ligands to the Peroxiredoxin 5 protein. Among the hit fragments, the small catechol molecule was shown to significantly inhibit Peroxiredoxin 5 activity in a thioredoxin peroxidase assay. This study reports novel data about the ligand-Peroxiredoxin interactions that will help considerably the development of potential Peroxiredoxin inhibitors.     

Fragment-based activity space: smaller is better

Fragment-based drug discovery has the potential to supersede traditional high throughput screening based drug discovery for molecular targets amenable to structure determination. This is because the chemical diversity coverage is better accomplished by a fragment collection of reasonable size than by larger HTS collections. Furthermore, fragments have the potential to be efficient target binders with higher probability than more elaborated drug-like compounds. The selection of the fragment screening technique is driven by sensitivity and throughput considerations, and we advocate in the present article the use of high concentration bioassays in conjunction with NMR-based hit confirmation. Subsequent ligand X-ray structure determination of the fragment ligand in complex with the target protein by co-crystallisation or crystal soaking can focus on confirmed binders.