Sequence‐based prediction of protein–peptide binding sites using support vector machine

Protein–peptide interactions are essential for all cellular processes including DNA repair, replication, gene‐expression, and metabolism. As most protein – peptide interactions are uncharacterized, it is cost effective to investigate them computationally as the first step. All existing approaches for predicting protein – peptide binding sites, however, are based on protein structures despite the fact that the structures for most proteins are not yet solved. This article proposes the first machine‐learning method called SPRINT to make Sequence‐based prediction of Protein – peptide Residue‐level Interactions. SPRINT yields a robust and consistent performance for 10‐fold cross validations and independent test. The most important feature is evolution‐generated sequence profiles. For the test set (1056 binding and non‐binding residues), it yields a Matthews’ Correlation Coefficient of 0.326 with a sensitivity of 64% and a specificity of 68%. This sequence‐based technique shows comparable or more accurate than structure‐based methods for peptide‐binding site prediction. SPRINT is available as an online server at: http://sparks-lab.org/ . © 2016 Wiley Periodicals, Inc.     

A Synthetic Adenylation‐Domain‐Based tRNA‐Aminoacylation Catalyst

The incorporation of non‐proteinogenic amino acids represents a major challenge for the creation of functionalized proteins. The ribosomal pathway is limited to the 20–22 proteinogenic amino acids while nonribosomal peptide synthetases (NRPSs) are able to select from hundreds of different monomers. Introduced herein is a fusion‐protein‐based design for synthetic tRNA‐aminoacylation catalysts based on combining NRPS adenylation domains and a small eukaryotic tRNA‐binding domain (Arc1p‐C). Using rational design, guided by structural insights and molecular modeling, the adenylation domain PheA was fused with Arc1p‐C using flexible linkers and achieved tRNA‐aminoacylation with both proteinogenic and non‐proteinogenic amino acids. The resulting aminoacyl‐tRNAs were functionally validated and the catalysts showed broad substrate specificity towards the acceptor tRNA. Our strategy shows how functional tRNA‐aminoacylation catalysts can be created for bridging the ribosomal and nonribosomal worlds. This opens up new avenues for the aminoacylation of tRNAs with functional non‐proteinogenic amino acids.     

Capillary electrophoretic methods applied to the investigation of peptide complexes

This article gives an overview of the applications of capillary electrophoretic methods to investigate the non‐covalent interactions of peptides (peptide complexes) with variable middle‐ and high‐molecular‐mass receptors (ligands) as well as with small ions and molecules in the period 2007–2014. Different modes of capillary electrophoretic methods, such as mobility shift (vacancy) affinity capillary electrophoresis, multiple injection affinity capillary electrophoresis, partial filling affinity capillary electrophoresis, Hummel–Dryer method, vacancy peak method and (continuous) frontal analysis capillary electrophoresis, are briefly described and their applicability to determination of binding constants of peptide complexes is discussed. In addition, the detailed experimental conditions of individual applications and the values of binding constants of the particular peptide complexes are presented.     

Adding energy minimization strategy to peptide‐design algorithm enables better search for RNA‐binding peptides: Redesigned λ N peptide binds boxB RNA

Our previously developed peptide‐design algorithm was improved by adding an energy minimization strategy which allows the amino acid sidechains to move in a broad configuration space during sequence evolution. In this work, the new algorithm was used to generate a library of 21‐mer peptides which could substitute for λ N peptide in binding to boxB RNA. Six potential peptides were obtained from the algorithm, all of which exhibited good binding capability with boxB RNA. Atomistic molecular dynamics simulations were then conducted to examine the ability of the λ N peptide and three best evolved peptides, viz. Pept01, Pept26, and Pept28, to bind to boxB RNA. Simulation results demonstrated that our evolved peptides are better at binding to boxB RNA than the λ N peptide. Sequence searches using the old (without energy minimization strategy) and new (with energy minimization strategy) algorithms confirm that the new algorithm is more effective at finding good RNA‐binding peptides than the old algorithm. © 2016 Wiley Periodicals, Inc.     

The statistical significance of selected sense–antisense peptide interactions

Sense and antisense peptides, encoded by sense and corresponding antisense DNA strands, are capable of specific interactions that could be a driving force to mediate protein–protein or protein–peptide binding associations. The complementary residue hypothesis suggests that these interactions are founded upon the sum of pairwise interactions between amino acids encoded by corresponding sense and antisense codons. Despite many successful experimental results obtained with the hypothesis, however, the physicochemical basis for these interactions is poorly understood. We examined the potential of the hypothesis for general identification of protein–protein interaction sites, and the possible role of the hypothesis in determining folding in a broad set of protein structures. In addition, we performed a structural study to investigate the binding of a complementary peptide to IL‐1F2. Our results suggest that complementary residue pairs are no more frequent or conserved than average in protein–protein interfaces, and are statistically under‐represented amongst contacting residue pairs in folded protein structures. Although our structural results matched experimental observations of binding between the peptide and IL‐1F2, complementary residue interactions do not appear to be dominant in the bound structure. Overall, our data do not allow us to conclude that the complementary residue hypothesis accounts for specific sense–antisense peptide interactions. © 2012 Wiley Periodicals, Inc.     

New Modalities for Challenging Targets in Drug Discovery

Our ever‐increasing understanding of biological systems is providing a range of exciting novel biological targets, whose modulation may enable novel therapeutic options for many diseases. These targets include protein–protein and protein–nucleic acid interactions, which are, however, often refractory to classical small‐molecule approaches. Other types of molecules, or modalities, are therefore required to address these targets, which has led several academic research groups and pharmaceutical companies to increasingly use the concept of so‐called “new modalities”. This Review defines for the first time the scope of this term, which includes novel peptidic scaffolds, oligonucleotides, hybrids, molecular conjugates, as well as new uses of classical small molecules. We provide the most representative examples of these modalities to target large binding surface areas such as those found in protein–protein interactions and for biological processes at the center of cell regulation.     

A Cyclized Helix‐Loop‐Helix Peptide as a Molecular Scaffold for the Design of Inhibitors of Intracellular Protein–Protein Interactions by Epitope and Arginine Grafting

The design of inhibitors of intracellular protein–protein interactions (PPIs) remains a challenge in chemical biology and drug discovery. We propose a cyclized helix‐loop‐helix (cHLH) peptide as a scaffold for generating cell‐permeable PPI inhibitors through bifunctional grafting: epitope grafting to provide binding activity, and arginine grafting to endow cell‐permeability. To inhibit p53–HDM2 interactions, the p53 epitope was grafted onto the C‐terminal helix and six Arg residues were grafted onto another helix. The designed peptide cHLHp53‐R showed high inhibitory activity for this interaction, and computational analysis suggested a binding mode for HDM2. Confocal microscopy of cells treated with fluorescently labeled cHLHp53‐R revealed cell membrane penetration and cytosolic localization. The peptide inhibited the growth of HCT116 and LnCap cancer cells. This strategy of bifunctional grafting onto a well‐structured peptide scaffold could facilitate the generation of inhibitors for intracellular PPIs.     

Double Strain‐Promoted Macrocyclization for the Rapid Selection of Cell‐Active Stapled Peptides

Peptide stapling is a method for designing macrocyclic alpha‐helical inhibitors of protein–protein interactions. However, obtaining a cell‐active inhibitor can require significant optimization. We report a novel stapling technique based on a double strain‐promoted azide–alkyne reaction, and exploit its biocompatibility to accelerate the discovery of cell‐active stapled peptides. As a proof of concept, MDM2‐binding peptides were stapled in parallel, directly in cell culture medium in 96‐well plates, and simultaneously evaluated in a p53 reporter assay. This in situ stapling/screening process gave an optimal candidate that showed improved proteolytic stability and nanomolar binding to MDM2 in subsequent biophysical assays. α‐Helicity was confirmed by a crystal structure of the MDM2‐peptide complex. This work introduces in situ stapling as a versatile biocompatible technique with many other potential high‐throughput biological applications.     

Dissecting Hofmeister Effects: Direct Anion–Amide Interactions Are Weaker than Cation–Amide Binding

Whereas there is increasing evidence for ion‐induced protein destabilization through direct ion–protein interactions, the strength of the binding of anions to proteins relative to cation–protein binding has remained elusive. In this work, the rotational mobility of a model amide in aqueous solution was used as a reporter for the interactions of different anions with the amide group. Protein‐stabilizing salts such as KCl and KNO₃ do not affect the rotational mobility of the amide. Conversely, protein denaturants such as KSCN and KI markedly reduce the orientational freedom of the amide group. Thus these results provide evidence for a direct denaturation mechanism through ion–protein interactions. Comparing the present findings with results for cations shows that in contrast to common belief, anion–amide binding is weaker than cation–amide binding.     

Inhibition of Ras Signaling by Blocking Ras–Effector Interactions with Cyclic Peptides

Ras genes are frequently activated in human cancers, but the mutant Ras proteins remain largely “undruggable” through the conventional small‐molecule approach owing to the absence of any obvious binding pockets on their surfaces. By screening a combinatorial peptide library, followed by structure–activity relationship (SAR) analysis, we discovered a family of cyclic peptides possessing both Ras‐binding and cell‐penetrating properties. These cell‐permeable cyclic peptides inhibit Ras signaling by binding to Ras‐GTP and blocking its interaction with downstream proteins and they induce apoptosis of cancer cells. Our results demonstrate the feasibility of developing cyclic peptides for the inhibition of intracellular protein–protein interactions and of direct Ras inhibitors as a novel class of anticancer agents.     

Simultaneous prediction of binding free energy and specificity for PDZ domain–peptide interactions

Interactions between protein domains and linear peptides underlie many biological processes. Among these interactions, the recognition of C-terminal peptides by PDZ domains is one of the most ubiquitous. In this work, we present a mathematical model for PDZ domain–peptide interactions capable of predicting both affinity and specificity of binding based on X-ray crystal structures and comparative modeling with Rosetta. We developed our mathematical model using a large phage display dataset describing binding specificity for a wild type PDZ domain and 91 single mutants, as well as binding affinity data for a wild type PDZ domain binding to 28 different peptides. Structural refinement was carried out through several Rosetta protocols, the most accurate of which included flexible peptide docking and several iterations of side chain repacking and backbone minimization. Our findings emphasize the importance of backbone flexibility and the energetic contributions of side chain-side chain hydrogen bonds in accurately predicting interactions. We also determined that predicting PDZ domain–peptide interactions became increasingly challenging as the length of the peptide increased in the N-terminal direction. In the training dataset, predicted binding energies correlated with those derived through calorimetry and specificity switches introduced through single mutations at interface positions were recapitulated. In independent tests, our best performing protocol was capable of predicting dissociation constants well within one order of magnitude of the experimental values and specificity profiles at the level of accuracy of previous studies. To our knowledge, this approach represents the first integrated protocol for predicting both affinity and specificity for PDZ domain–peptide interactions.     

Assessment of a Large Enzyme–Drug Complex by Proton‐Detected Solid‐State NMR Spectroscopy without Deuteration

Solid‐state NMR spectroscopy has recently enabled structural biology with small amounts of non‐deuterated proteins, largely alleviating the classical sample production demands. Still, despite the benefits for sample preparation, successful and comprehensive characterization of complex spin systems in the few cases of higher‐molecular‐weight proteins has thus far relied on traditional ¹³C‐detected methodology or sample deuteration. Herein we show for a 29 kDa carbonic anhydrase:acetazolamide complex that different aspects of solid‐state NMR assessment of a complex spin system can be successfully accessed using a non‐deuterated, 500 μg sample in combination with adequate spectroscopic tools. The shown access to protein structure, protein dynamics, as well as biochemical parameters in amino acid sidechains, such as histidine protonation states, will be transferable to proteins that are not expressible in E. coli.     

Ester Carbonyl Vibration as a Sensitive Probe of Protein Local Electric Field

The ability to quantify the local electrostatic environment of proteins and protein/peptide assemblies is key to gaining a microscopic understanding of many biological interactions and processes. Herein, we show that the ester carbonyl stretching vibration of two non‐natural amino acids, L ‐aspartic acid 4‐methyl ester and L ‐glutamic acid 5‐methyl ester, is a convenient and sensitive probe in this regard, since its frequency correlates linearly with the local electrostatic field for both hydrogen‐bonding and non‐hydrogen‐bonding environments. We expect that the resultant frequency–electric‐field map will find use in various applications. Furthermore, we show that, when situated in a non‐hydrogen‐bonding environment, this probe can also be used to measure the local dielectric constant (ε). For example, its application to amyloid fibrils formed by Aβ_16–22 revealed that the interior of such β‐sheet assemblies has an ε value of approximately 5.6.     

Do Intrinsically Disordered Proteins Possess High Specificity in Protein–Protein Interactions?

Specific protein–protein interactions are critical to cellular function. Structural flexibility and disorder‐to‐order transitions upon binding enable intrinsically disordered proteins (IDPs) to overcome steric restrictions and form complementary binding interfaces, and thus, IDPs are widely considered to have high specificity and low affinity for molecular recognition. However, flexibility may also enable IDPs to form complementary binding interfaces with misbinding partners, resulting in a great number of nonspecific interactions. Consequently, it is questionable whether IDPs really possess high specificity. In this work, we investigated this question from a thermodynamic viewpoint. We collected mutant thermodynamic data for 35 ordered protein complexes and 43 disordered protein complexes. We found that the enthalpy–entropy compensation for disordered protein complexes was more complete than that for ordered protein complexes. We further simulated the binding processes of ordered and disordered protein complexes under mutations. Simulation data confirmed the observation of experimental data analyses and further revealed that disordered protein complexes possessed smaller changes in binding free energy than ordered protein complexes under the same mutation perturbations. Therefore, interactions of IDPs are more malleable than those of ordered proteins due to their structural flexibility in the complex. Our results provide new clues for exploring the relationship between protein flexibility, adaptability, and specificity.     

FRET in a Synthetic Flavin‐ and Bilin‐binding Protein

The last decade has seen development and application of a large number of novel fluorescence‐based techniques that have revolutionized fluorescence microscopy in life sciences. Preferred tags for such applications are genetically encoded fluorescent proteins (FP), mostly derivatives of the green fluorescent protein (GFP). Combinations of FPs with wavelength‐separated absorption/fluorescence properties serve as excellent tools for molecular interaction studies, for example, protein–protein complexes or enzyme–substrate interactions, based on the FRET phenomenon (Förster resonance energy transfer). However, alternatives are requested for experimental conditions where FP proteins or FP couples are not or less efficiently applicable. We here report as a “proof of principle” a specially designed, non‐naturally occurring protein (LG1) carrying a combination of a flavin‐binding LOV‐ and a photochromic bilin‐binding GAF domain and demonstrate a FRET process between both chromophores.     

Native Top‐Down Mass Spectrometry of TAR RNA in Complexes with a Wild‐Type tat Peptide for Binding Site Mapping

Ribonucleic acids (RNA) frequently associate with proteins in many biological processes to form more or less stable complex structures. The characterization of RNA–protein complex structures and binding interfaces by nuclear magnetic resonance (NMR) spectroscopy, X‐ray crystallography, or strategies based on chemical crosslinking, however, can be quite challenging. Herein, we have explored the use of an alternative method, native top‐down mass spectrometry (MS), for probing of complex stoichiometry and protein binding sites at the single‐residue level of RNA. Our data show that the electrostatic interactions between HIV‐1 TAR RNA and a peptide comprising the arginine‐rich binding region of tat protein are sufficiently strong in the gas phase to survive phosphodiester backbone cleavage of RNA by collisionally activated dissociation (CAD), thus allowing its use for probing tat binding sites in TAR RNA by top‐down MS. Moreover, the MS data reveal time‐dependent 1:2 and 1:1 stoichiometries of the TAR–tat complexes and suggest structural rearrangements of TAR RNA induced by binding of tat peptide.     

Structural Basis of Sirtuin 6 Activation by Synthetic Small Molecules

Sirtuins are protein deacylases regulating metabolism and stress responses, and are implicated in aging‐related diseases. Small molecule activators for the human sirtuins Sirt1‐7 are sought as chemical tools and potential therapeutics, such as for cancer. Activators are available for Sirt1 and exploit its unique N‐terminus, whereas drug‐like activators for Sirt2–7 are lacking. We synthesized and screened pyrrolo[1,2‐a]quinoxaline derivatives, yielding the first synthetic Sirt6 activators. Biochemical assays show direct, substrate‐independent compound binding to the Sirt6 catalytic core and potent activation of Sirt6‐dependent deacetylation of peptide substrates and complete nucleosomes. Crystal structures of Sirt6/activator complexes reveal that the compounds bind to a Sirt6‐specific acyl channel pocket and identify key interactions. Our results establish potent Sirt6 activation with small molecules and provide a structural basis for further development of Sirt6 activators as tools and therapeutics.     

A Synthetic Loop Replacement Peptide That Blocks Canonical NF‐κB Signaling

Aberrant canonical NF‐κB signaling is implicated in diseases from autoimmune disorders to cancer. A major therapeutic challenge is the need for selective inhibition of the canonical pathway without impacting the many non‐canonical NF‐κB functions. Here we show that a selective peptide‐based inhibitor of canonical NF‐κB signaling, in which a hydrogen bond in the NBD peptide is synthetically replaced by a non‐labile bond, shows an about 10‐fold increased potency relative to the original inhibitor. Not only is this molecule, NBD2, a powerful tool for dissection of canonical NF‐κB signaling in disease models and healthy tissues, the success of the synthetic loop replacement suggests that the general strategy could be useful for discovering modulators of the many protein–protein interactions mediated by such structures.