Predicting and Understanding the Enzymatic Inhibition of Human Peroxiredoxin 5 by 4-Substituted Pyrocatechols by Combining Funnel Metadynamics, Solution NMR, and Steady-State Kinetics.

Funnel metadynamics is a kind of computational simulation used to enhance the sampling of protein-ligand binding events in solution. By characterization of the binding interaction events, an estimated absolute binding free energy can be calculated. Nuclear magnetic resonance and funnel metadynamics were used to evaluate the binding of pyrocatechol derivatives (catechol, 4-methylcatechol, and 4-tert-butylcatechol) to human peroxiredoxin 5. Human peroxiredoxins are peroxidases involved in cellular peroxide homeostasis. Recently, overexpressed or suppressed peroxiredoxin levels have been linked to various diseases. Here, the catechol derivatives were found to be inhibitors against human peroxiredoxin 5 through a partial mixed type noncompetitive mechanism. Funnel metadynamics provided a microscopic model for interpreting the inhibition mechanism. Correlations were observed between the inhibition constants and the absolute binding free energy. Overall, this study showcases the fact that funnel metadynamics simulations can be employed as a preliminary approach to gain an in-depth understanding of potential enzyme inhibitors.

Funnel-metadynamics and solution NMR to estimate protein-ligand affinities.

One of the intrinsic properties of proteins is their capacity to interact selectively with other molecules in their environment, inducing many chemical equilibria each differentiated by the mutual affinities of the components. A comprehensive understanding of these molecular binding processes at atomistic resolution requires formally the complete description of the system dynamics and statistics at the relevant time scales. While solution NMR observables are averaged over different time scales, from picosecond to second, recent new molecular dynamics protocols accelerated considerably the simulation time of realistic model systems. Based on known ligands recently discovered either by crystallography or NMR for the human peroxiredoxin 5, their affinities were for the first time accurately evaluated at atomistic resolution comparing absolute binding free-energy estimated by funnel-metadynamics simulations and solution NMR experiments. In particular, free-energy calculations are demonstrated to discriminate two closely related ligands as pyrocatechol and 4-methylpyrocathecol separated just by 1 kcal/mol in aqueous solution. The results provide a new experimental and theoretical basis for the estimation of ligand-protein affinities.

Competitive binding of UBPY and ubiquitin to the STAM2 SH3 domain revealed by NMR.

To date, the signal transducing adaptor molecule 2 (STAM2) was shown to harbour two ubiquitin binding domains (UBDs) known as the VHS and UIM domains, while the SH3 domain of STAM2 was reported to interact with deubiquitinating enzymes (DUBs) like UBPY and AMSH. In the present study, NMR evidences the interaction of the STAM2 SH3 domain with ubiquitin, demonstrating that SH3 constitutes the third UBD of STAM2. Furthermore, we show that a UBPY-derived peptide can outcompete ubiquitin for SH3 binding and vice versa. These results suggest that the SH3 domain of STAM2 plays versatile roles in the context of ubiquitin mediated receptor sorting.

Dynamical properties of the loop 320s of substrate-free and substrate-bound muscle creatine kinase by NMR: evidence for independent subunits.

Muscle creatine kinase (MCK; EC2.7.3.2) is a 86 kDa homodimer that belongs to the family of guanidino kinases. MCK has been intensively studied for several decades, but it is still not known why it is a dimer because this quaternary structure does not translate into obvious structural or functional advantages over the homologous monomeric arginine kinase. In particular, it remains to be demonstrated whether MCK subunits are independent. Here, we describe NMR chemical-shift perturbation and relaxation experiments designed to study the active site 320s flexible loop of this enzyme. The analysis was performed with the enzyme in its ligand-free and MgADP-complexed forms, as well as with the transition-state analogue abortive complex (MCK-Mg-ADP-creatine-nitrate ion). Our data indicate that each subunit can bind substrates independently.

Evidence for cooperative and domain-specific binding of the signal transducing adaptor molecule 2 (STAM2) to Lys63-linked diubiquitin.

As the upstream component of the ESCRT (endosomal sorting complexes required for transport) machinery, the ESCRT-0 complex is responsible for directing ubiquitinated membrane proteins to the multivesicular body pathway. ESCRT-0 is formed by two subunits known as Hrs (hepatocyte growth factor-regulated substrate) and STAM (signal transducing adaptor molecule), both of which harbor multiple ubiquitin-binding domains (UBDs). In particular, STAM2 possesses two UBDs, the VHS (Vps27/Hrs/Stam) and UIM (ubiquitin interacting motif) domains, connected by a 20-amino acid flexible linker. In the present study, we report the interactions of the UIM domain and VHS-UIM construct of STAM2 with monoubiquitin (Ub), Lys(48)- and Lys(63)-linked diubiquitins. Our results demonstrate that the UIM domain alone binds monoubiquitin, Lys(48)- and Lys(63)-linked diubiquitins with the same affinity and in the same binding mode. Interestingly, binding of VHS-UIM to Lys(63)-linked diubiquitin is not only avid, but also cooperative. We also show that the distal domain of Lys(63)-linked diubiquitin stabilizes the helical structure of the UIM domain and that the corresponding complex adopts a specific structural organization responsible for its greater affinity. In contrast, binding of VHS-UIM to Lys(48)-linked diubiquitin and monoubiquitin is not cooperative and does not show any avidity. These results may explain the better sorting efficiency of some cargoes polyubiquitinated with Lys(63)-linked chains over monoubiquitinated cargoes or those tagged with Lys(48)-linked chains.

Binding evaluation of fragment-based scaffolds for probing allosteric enzymes.

Fragment-based drug discovery has become a powerful method for the generation of drug leads against therapeutic targets. Beyond the identification of novel and effective starting points for drug design, fragments have emerged as reliable tools for assessing protein druggability and identifying protein hot spots. Here, we have examined fragments resulting from the deconstruction of known inhibitors from the glycogen phosphorylase enzyme, a therapeutic target against type 2 diabetes, with two motivations. First, we have analyzed the fragment binding to the multiple binding sites of the glycogen phosphorylase, and then we have investigated the use of fragments to study allosteric enzymes. The work we report illustrates the power of fragmentlike ligands not only for probing the various binding pockets of proteins, but also for uncovering cooperativity between these various binding sites.

The Arabidopsis peptide kiss of death is an inducer of programmed cell death.

Programmed cell death (PCD) has a key role in defence and development of all multicellular organisms. In plants, there is a large gap in our knowledge of the molecular machinery involved at the various stages of PCD, especially the early steps. Here, we identify kiss of death (KOD) encoding a 25-amino-acid peptide that activates a PCD pathway in Arabidopsis thaliana. Two mutant alleles of KOD exhibited a reduced PCD of the suspensor, a single file of cells that support embryo development, and a reduced PCD of root hairs after a 55°C heat shock. KOD expression was found to be inducible by biotic and abiotic stresses. Furthermore, KOD expression was sufficient to cause death in leaves or seedlings and to activate caspase-like activities. In addition, KOD-induced PCD required light in leaves and was repressed by the PCD-suppressor genes AtBax inhibitor 1 and p35. KOD expression resulted in depolarization of the mitochondrial membrane, placing KOD above mitochondria dysfunction, an early step in plant PCD. A KOD∷GFP fusion, however, localized in the cytosol of cells and not mitochondria.

NMR reveals a different mode of binding of the Stam2 VHS domain to ubiquitin and diubiquitin.

The VHS domain of the Stam2 protein is a ubiquitin binding domain involved in the recognition of ubiquitinated proteins committed to lysosomal degradation. Among all VHS domains, the VHS domain of Stam proteins is the strongest binder to monoubiqiuitin and exhibits preferences for K63-linked chains. In the present paper, we report the solution NMR structure of the Stam2-VHS domain in complex with monoubiquitin by means of chemical shift perturbations, spin relaxation, and paramagnetic relaxation enhancements. We also characterize the interaction of Stam2-VHS with K48- and K63-linked diubiquitin chains and report the first evidence that VHS binds differently to these two chains. Our data reveal that VHS enters the hydrophobic pocket of K48-linked diubiquitin and binds the two ubiquitin subunits with different affinities. In contrast, VHS interacts with K63-linked diubiquitin in a mode similar to its interaction with monoubiquitin. We also suggest possible structural models for both K48- and K63-linked diubiquitin in interaction with VHS. Our results, which demonstrate a different mode of binding of VHS for K48- and K63-linked diubiquitin, may explain the preference of VHS for K63- over K48-linked diubiquitin chains and monoubiquitin.

Characterization of caged compounds binding to proteins by NMR spectroscopy.

Photolysable caged ligands are used to investigate protein function and activity. Here, we investigate the binding properties of caged nucleotides and their photo released products to well established but evolutionary and structurally unrelated nucleotide-binding proteins, rabbit muscle creatine kinase (RMCK) and human annexin A6 (hAnxA6), using saturation transfer difference NMR spectroscopy. We detect the binding of the caged nucleotides and discuss the general implications on interpreting data collected with photolysable caged ligands using different techniques. Strategies to avoid non-specific binding of caged compound to certain proteins are also suggested.

Ligand specificity in fragment-based drug design.

Fragment-based drug design consists of identifying low-molecular weight compounds that weakly bind to a target macromolecule and will then be modified or linked to yield potent inhibitors. The specificity of these low-complexity and low-affinity molecules has rarely been discussed in the literature. To address this question, NMR spectroscopy was used to investigate the interactions of 150 fragments with five proteins: three proteins from the Bcl-2 family (Bcl-x(L), Bcl-w, and Mcl-1), human peroxiredoxin 5, for which very few ligands have been reported, and human serum albumin, which is known to bind a large number of ligands. Our results show that the fragments are rather versatile binders and able to identify binding hot spots in very different targets. Despite the different hit rates observed related to the druggability of the proteins, two scaffolds appear as preferred binders for all proteins. Low specificity was observed between homologous proteins or unrelated poorly druggable proteins, while higher specificity could be achieved with highly druggable targets.

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 deconstruction of Bcl-xL inhibitors.

Fragment-based drug design consists of screening low-molecular-weight compounds in order to identify low-affinity ligands that are then modified or linked to yield potent inhibitors. The method thus attempts to build bioactive molecules in a modular way and relies on the hypothesis that the fragment binding mode will be conserved upon elaboration of the active molecule. If the inverse process is considered, do the fragments resulting from the deconstruction of high-affinity inhibitors recapitulate their binding mode in the large molecule? Few studies deal with this issue. Here, we report the analysis of 22 fragments resulting from the dissection of 9 inhibitors of the antiapoptotic protein Bcl-x(L). To determine if the fragments retained affinity toward the protein and identify their binding site, ligand-observed and protein-observed NMR experiments were used. The analysis of the fragments behavior illustrates the complexity of low-affinity protein-ligand interactions involved in the fragment-based construction of bioactive molecules.

Hyaluronidase binds differently DPPC, DPPS or GlcNAc-bearing glycolipid biomimetic monolayers.

Bovine testis hyaluronidase (btHyal) had been shown to have direct effects on cancer cells and to be a useful adjuvant in several medicines. Furthermore this enzyme had been found to be membrane-associated. Thus, in this work, the interactions between btHyal and membranes were analyzed by using lipid monolayers at the air-water interface as a biomimetic membrane system. This allowed us to define the btHyal interactions with two residues of hyaluronic acid (a btHyal substrate), GlcNAc and carboxylic group, which are present in cholesteryl-triethoxy-N-acetylglucosamine (Chol-E3-GlcNAc) and in DPPS, respectively. btHyal bound preferentially Chol-E3-GlcNAc monolayers and showed a decreasing affinity for Chol-E3-GlcNAc-DPPC monolayers containing decreasing amount of glycolipid, suggesting a crucial role of the glycolipid GlcNAc. Furthermore the significant btHyal binding to DPPS was not affected by the presence of free GlcNAc in the subphase. These results and the absence of significant binding of btHyal to pure DPPC monolayer suggest that the protein interacts with the lipid monolayer by mimicking the enzyme-substrate interactions or by electrostatic interactions.

Interaction domain on thioredoxin for Pseudomonas aeruginosa 5'-adenylylsulfate reductase.

NMR spectroscopy has been used to map the interaction domain on Escherichia coli thioredoxin for the thioredoxin- dependent 5'-adenylylsulfate reductase from Pseudomonas aeruginosa (PaAPR). Seventeen thioredoxin amino acids, all clustered around Cys-32 (the more surface-exposed of the two active-site cysteines), have been located at the PaAPR binding site. The center of the binding domain is dominated by nonpolar amino acids, with a smaller number of charged and polar amino acids located on the periphery of the site. Twelve of the amino acids detected by NMR have non-polar, hydrophobic side chains, including one aromatic amino acid (Trp-31). Four of the thioredoxin amino acids at the PaAPR binding site have polar side chains (Lys-36, Asp-61, Gln-62 and Arg-73), with three of the four having charged side chains. Site-directed mutagenesis experiments have shown that replacement of Lys-36, Asp-61, and Arg-73 and of the absolutely conserved Trp-31 significantly decreases the V(max) for the PaAPR-catalyzed reduction of 5'-adenylylsulfate, with E. coli thioredoxin serving as the electron donor. The most dramatic effect was observed with the W31A variant, which showed no activity as a donor to PaAPR. Although the thiol of the active-site Cys-256 of PaAPR is the point of the initial nucleophilic attack by reduced thioredoxin, mutagenic replacement of Cys-256 by serine has no effect on thioredoxin binding to PaAPR.

NMR measure of translational diffusion and fractal dimension. Application to molecular mass measurement.

Experimental NMR diffusion measure on polymers and on globular proteins are presented. These results, complemented with results found in the literature, enable a general description of effective fractal dimension for objects such as small organic molecules, sugars, polymers, DNA, and proteins. Results are compared to computational simulations as well as to theoretical values. A global picture of the diffusion phenomenon emerges from this description. A power law relating molecular mass with diffusion coefficients is described and found to be valid over 4 orders of magnitude. From this law, the fractal dimension of the molecular family can be measured, with experimental values ranging from 1.41 to 2.56 in full agreement with theoretical approaches. Finally, a method for evaluating the molecular mass of unknown solutes is described and implemented as a Web page.

NMR screening applied to the fragment-based generation of inhibitors of creatine kinase exploiting a new interaction proximate to the ATP binding site.

Using an in-house fragment NMR library, we identified a set of ligands that bind rabbit muscular creatine kinase, an enzyme involved in key ATP-dependent processes. The ligands docked to the crystal structures of creatine kinase indicated that a phenylfuroic acid could enter into a pocket adjacent to the nucleotide binding site. This fragment served as an anchor to develop in silico a series of potential inhibitors which could partly access the nucleotide binding site. The short synthesis of only four derivatives provided entirely novel hit compounds that reversibly inhibit creatine kinase at micromolar concentrations with a mixed ATP-competitive/noncompetitive mechanism in agreement with the structural model of the inhibited enzyme. These initial biologically active compounds are novel and modular and exploit a new interaction proximate to the ATP binding site.

Structure and dynamics of surfactin studied by NMR in micellar media.

The NMR structure of the cyclic lipopeptide surfactin from Bacillus subtilis was determined in sodium dodecyl sulfate (SDS) micellar solution. The two negatively charged side chains of surfactin form a polar head opposite to most hydrophobic side chains, accounting for its amphiphilic nature and its strong surfactant properties. Disorder was observed around the fatty acid chain, and 15N relaxation studies were performed to investigate whether it originates from a dynamic phenomenon. A very large exchange contribution to transverse relaxation rate R(2) was effectively observed in this region, indicating slow conformational exchange. Temperature variation and Carr-Purcell-Meiboom-Gill (CPMG) delay variation relaxation studies provided an estimation of the apparent activation energy around 35-43 kJ x mol(-1) and an exchange rate of about 200 ms(-1) for this conformational exchange. 15N relaxation parameters were also recorded in dodecylphosphocholine (DPC) micelles and DMSO. Similar chemical exchange around the fatty acid was found in DPC but not in DMSO, which demonstrates that this phenomenon only occurs in micellar media. Consequently, it may either reflect the disorder observed in our structures determined in SDS or originate from an interaction of the lipopeptide with the detergent, which would be qualitatively similar with an anionic (SDS) or a zwitterionic (DPC) detergent. These structural and dynamics results on surfactin are the first NMR characterization of a lipopeptide incorporated in micelles. Moreover, they provide a model of surfactin determined in a more biomimetic environment than an organic solvent, which could be useful for understanding the molecular mechanism of its biological activity.

NMR structure determination of a synthetic analogue of bacillomycin Lc reveals the strategic role of L-Asn1 in the natural iturinic antibiotics.

Iturins are a group of antifungal produced by Bacillus subtilis. All are cyclic lipopeptides with seven alpha-amino acids of configuration LDDLLDL and one beta-amino fatty acid. The bacillomycin L is a member of this family and its NMR structure was previously resolved using the sequence Asp-Tyr-Asn-Ser-Gln-Ser-Thr. In this work, we carefully examined the NMR spectra of this compound and detected an error in the sequence. In fact, Asp1 and Gln5 need to be changed into Asn1 and Glu5, which therefore makes it identical to bacillomycin Lc. As a consequence, it now appears that all iturinic peptides with antibiotic activity share the common beta-amino fatty acid 8-L-Asn1-D-Tyr2-D-Asn3 sequence. To better understand the conformational influence of the acidic residue L-Asp1, present, for example in the inactive iturin C, the NMR structure of the synthetic analogue SCP [cyclo (L-Asp1-D-Tyr2-D-Asn3-L-Ser4-L-Gln5-D-Ser6-L-Thr7-beta-Ala8)] was determined and compared with bacillomycin Lc recalculated with the corrected sequence. In both cases, the conformers obtained were separated into two families of similar energy which essentially differ in the number and type of turns. A detailed analysis of both cyclopeptide structures is presented here. In addition, CD and FTIR spectra were performed and confirmed the conformational differences observed by NMR between both cyclopeptides.

Protein-protein interactions within peroxiredoxin systems.

Peroxiredoxin systems in plants were demonstrated involved in crucial roles related to reactive oxygenated species (ROS) metabolism and the linked cell signalling to ROS. Peroxiredoxins function as peroxidasic systems that combine at least a reactivating reductant agent like thioredoxins, and sometimes glutaredoxins and glutathion. In the past three years a number of peroxiredoxin structures were solved by crystallography in different experimental crystallisation conditions. The structures have revealed a significant propensity of peroxiredoxins for oligomerism that was confirmed by biophysical studies in solution using NMR and other methods as analytical ultra-centrifugation. These studies showed that quaternary structures of peroxiredoxins involve specific protein-protein interaction interfaces that rely upon the peroxiredoxin types and/or their redox conditions. The protein-protein interactions with the reactivating redoxins essentially lead to transient unstable complexes. We review herein the different protein-protein interactions characterized or deduced from those reports.

Glutathionylation induces the dissociation of 1-Cys D-peroxiredoxin non-covalent homodimer.

1-Cys peroxiredoxins (1-Cys Prxs) are antioxidant enzymes that catalyze the reduction of hydroperoxides into alcohols using a strictly conserved cysteine. 1-Cys B-Prxs, homologous to human PrxVI, were recently shown to be reactivated by glutathione S-transferase (GST) pi via the formation of a GST-Prx heterodimer and Prx glutathionylation. In contrast, 1-Cys D-Prxs, homologous to human PrxV, are reactivated by the glutaredoxin-glutathione system through an unknown mechanism. To investigate the mechanistic events that mediate the 1-Cys D-Prx regeneration, interaction of the Prx with glutathione was studied by mass spectrometry and NMR. This work reveals that the Prx can be glutathionylated on its active site cysteine. Evidences are reported that the glutathionylation of 1-Cys D-Prx induces the dissociation of the Prx non-covalent homodimer, which can be recovered by reduction with dithiothreitol. This work demonstrates for the first time the existence of a redox-dependent dimer-monomer switch in the Prx family, similar to the decamer-dimer switch for the 2-Cys Prxs.