De novo synthetic antimicrobial peptide design with a recurrent neural network.

Antibiotic resistance is recognized as an imminent and growing global health threat. New antimicrobial drugs are urgently needed due to the decreasing effectiveness of conventional small-molecule antibiotics. Antimicrobial peptides (AMPs), a class of host defense peptides, are emerging as promising candidates to address this need. The potential sequence space of amino acids is combinatorially vast, making it possible to extend the current arsenal of antimicrobial agents with a practically infinite number of new peptide-based candidates. However, mining naturally occurring AMPs, whether directly by wet lab screening methods or aided by bioinformatics prediction tools, has its theoretical limit regarding the number of samples or genomic/transcriptomic resources researchers have access to. Further, manually designing novel synthetic AMPs requires prior field knowledge, restricting its throughput. In silico sequence generation methods are gaining interest as a high-throughput solution to the problem. Here, we introduce AMPd-Up, a recurrent neural network based tool for de novo AMP design, and demonstrate its utility over existing methods. Validation of candidates designed by AMPd-Up through antimicrobial susceptibility testing revealed that 40 of the 58 generated sequences possessed antimicrobial activity against Escherichia coli and/or Staphylococcus aureus. These results illustrate that AMPd-Up can be used to design novel synthetic AMPs with potent activities.

Adaptive Supramolecular Networks: Emergent Sensing from Complex Systems.

Molecular differentiation by supramolecular sensors is typically achieved through sensor arrays, relying on the pattern recognition responses of large panels of isolated sensing elements. Here we report a new one-pot systems chemistry approach to differential sensing in biological solutions. We constructed an adaptive network of three cross-assembling sensor elements with diverse analyte-binding and photophysical properties. This robust sensing approach exploits complex interconnected sensor-sensor and sensor-analyte equilibria, producing emergent supramolecular and photophysical responses unique to each analyte. We characterize the basic mechanisms by which an adaptive network responds to analytes. The inherently data-rich responses of an adaptive network discriminate among very closely related proteins and protein mixtures without relying on designed protein recognition elements. We show that a single adaptive sensing solution provides better analyte discrimination using fewer response observations than a sensor array built from the same components. We also show the network's ability to adapt and respond to changing biological solutions over time.

Associating Biological Activity and Predicted Structure of Antimicrobial Peptides from Amphibians and Insects.

Antimicrobial peptides (AMPs) are a diverse class of short, often cationic biological molecules that present promising opportunities in the development of new therapeutics to combat antimicrobial resistance. Newly developed in silico methods offer the ability to rapidly discover numerous novel AMPs with a variety of physiochemical properties. Herein, using the rAMPage AMP discovery pipeline, we bioinformatically identified 51 AMP candidates from amphibia and insect RNA-seq data and present their in-depth characterization. The studied AMPs demonstrate activity against a panel of bacterial pathogens and have undetected or low toxicity to red blood cells and human cultured cells. Amino acid sequence analysis revealed that 30 of these bioactive peptides belong to either the Brevinin-1, Brevinin-2, Nigrocin-2, or Apidaecin AMP families. Prediction of three-dimensional structures using ColabFold indicated an association between peptides predicted to adopt a helical structure and broad-spectrum antibacterial activity against the Gram-negative and Gram-positive species tested in our panel. These findings highlight the utility of associating the diverse sequences of novel AMPs with their estimated peptide structures in categorizing AMPs and predicting their antimicrobial activity.

Molecular recognition of methylated amino acids and peptides by Pillar[6]MaxQ.

We report the molecular recognition properties of Pillar[n]MaxQ (P[n]MQ) toward a series of (methylated) amino acids, amino acid amides, and post-translationally modified peptides by a combination of 1H NMR, isothermal titration calorimetry, indicator displacement assays, and molecular dynamics simulations. We find that P6MQ is a potent receptor for N-methylated amino acid side chains. P6MQ recognized the H3K4Me3 peptide with Kd = 16 nM in phosphate buffered saline.

Binding Methylarginines and Methyllysines as Free Amino Acids: A Comparative Study of Multiple Host Classes.

Methylated free amino acids are an important class of targets for host-guest chemistry that have recognition properties distinct from those of methylated peptides and proteins. We present comparative binding studies for three different host classes that are each studied with multiple methylated arginines and lysines to determine fundamental structure-function relationships. The hosts studied are all anionic and include three calixarenes, two acyclic cucurbiturils, and two other cleft-like hosts, a clip and a tweezer. We determined the binding association constants for a panel of methylated amino acids using indicator displacement assays. The acyclic cucurbiturils display stronger binding to the methylated amino acids, and some unique patterns of selectivity. The two other cleft-like hosts follow two different trends, shallow host (clip) following similar trends to the calixarenes, and the other more closed host (tweezer) binding certain less-methylated amino acids stronger than their methylated counterparts. Molecular modelling sheds some light on the different preferences of the various hosts. The results identify hosts with new selectivities and with affinities in a range that could be useful for biomedical applications. The overall selectivity patterns are explained by a common framework that considers the geometry, depth of binding pockets, and functional group participation across all host classes.

Template-Directed Synthesis of Bivalent, Broad-Spectrum Hosts for Neuromuscular Blocking Agents.

We report on the synthesis of bivalent water-soluble calix[4]arene and calix[5]arene hosts, Super-sCx4 and Super-sCx5 as new broad-spectrum supramolecular binders of neuromuscular blocking agents (NMBAs). Synthesis was achieved using the target bisquaternary amine NMBAs as a template to link two highly anionic p-sulfonatocalixarene building blocks in aqueous solution. Bivalent anionic hosts Super-sCx4 and Super-sCx5 bind by engaging both quaternary amines present on a variety of NMBAs. We report low µM binding to structurally diverse alkyl, steroidal, curarine and benzylisoquinoline NMBAs with high selectivity over the neurotransmitter acetylcholine and a variety of other hydrophobic amines.

Mechanism of a Disassembly-Driven Sensing System Studied by Stopped-Flow Kinetics.

We carried out steady-state and stopped-flow photophysical measurements to determine the kinetics of a discrete disassembly driven turn-on fluorescent system. On and off rates for both DimerDye1 assembly and nicotine binding were determined. Relative rates for these competing processes provide insight on how this system can be optimized for sensing applications. Kinetics studies in artificial saliva showed that moving to more complex media has minimal effects on the sensing ability of the system.

Polycomb Paralog Chromodomain Inhibitors Active against Both CBX6 and CBX8*.

Methyllysine reader proteins bind to methylated lysine residues and alter gene transcription by changing either the compaction state of chromatin or by the recruitment of other multiprotein complexes. The polycomb paralog family of methyllysine readers bind to trimethylated lysine on the tail of histone 3 (H3) via a highly conserved aromatic cage located in their chromodomains. Each of the polycomb paralogs are implicated in several disease states. CBX6 and CBX8 are members of the polycomb paralog family with two structurally similar chromodomains. By exploring the structure-activity relationships of a previously reported CBX6 inhibitor we have discovered more potent and cell permeable analogs. Our current report includes potent, dual-selective inhibitors of CBX6 and CBX8. We have shown that the -2 position in our scaffold is an important residue for selectivity amongst the polycomb paralogs. Preliminary cell-based studies show that the new inhibitors impact cell proliferation in a rhabdoid tumor cell line.

An easily accessible, lower rim substituted calix[4]arene selectively binds N,N-dimethyllysine.

Post-translational modifications (PTMs) are critical controllers of protein functions. One set of important PTMs are N-methylated side chains of lysine and arginine, which exist in several functionally distinct forms. Multiple groups have demonstrated the selective binding of the most hydrophobic family member, trimethyllysine (Kme3), using various macrocyclic hosts, but the selective binding of lower methylation states remains challenging. Herein we report that the installation of a sulfonate ester on the lower rim phenol of p-sulfonatocalix[4]arene efficiently generates a potent, N,N-dimethyllysine (Kme2)-selective host in one step from commercially available starting materials. We characterize its binding behaviors in solution, and examine the relationship between its unusual conformational dynamics and its guest-binding properties.

Host-guest binding in water, salty water, and biofluids: general lessons for synthetic, bio-targeted molecular recognition.

Synthetic molecular recognition systems are increasingly being used to solve applied problems in the life sciences, and bio-targeted host-guest chemistry has rapidly arisen as a major field of fundamental research. This tutorial review presents a set of fundamental lessons on how host-guest molecular recognition can be programmed in water. The review uses informative examples of aqueous host-guest chemistry organized around generalizable themes and lessons, building towards lessons focused on molecular recognition in salty solutions and biological fluids. It includes selected examples of macrocyclic host systems that work well, as well as common pitfalls and how to avoid them. The review closes with a survey of the most important and inspirational recent advances, which involve host-guest chemistry in living cells and organisms.

François Diederich - In Memoriam*.

François Diederich - In Memoriam. In this Guest Editorial, Fraser Hof and Anna K. H. Hirsch help us remember the life and scientific legacy of Prof. François Diederich, a beloved mentor and inspiration to many, as well as an extraordinary scientist who made significant impacts in remarkably diverse areas.

Pan-specific and partially selective dye-labeled peptidic inhibitors of the polycomb paralog proteins.

Epigenetic regulation of gene expression is in part controlled by post-translational modifications on histone proteins. Histone methylation is a key epigenetic mark that controls gene transcription and repression. There are five human polycomb paralog proteins (Cbx2/4/6/7/8) that use their chromodomains to recognize trimethylated lysine 27 on histone 3 (H3K27me3). Recognition of the methyllysine side chain is achieved through multiple cation-pi interactions within an 'aromatic cage' motif. Despite high structural similarity within the chromodomains of this protein family, they each have unique functional roles and are linked to different cancers. Selective inhibition of different CBX proteins is desirable for both fundamental studies and potential therapeutic applications. We report here on a series of peptidic inhibitors that target certain polycomb paralogs. We have identified peptidic scaffolds with sub-micromolar potency, and will report examples that are pan-specific and that are partially selective for individual members within the family. These results highlight important structure-activity relationships that allow for differential binding to be achieved through interactions outside of the methyllysine-binding aromatic cage motif.

Optimization of Ligands Using Focused DNA-Encoded Libraries To Develop a Selective, Cell-Permeable CBX8 Chromodomain Inhibitor.

Polycomb repressive complex 1 (PRC1) is critical for mediating gene expression during development. Five chromobox (CBX) homolog proteins, CBX2, CBX4, CBX6, CBX7, and CBX8, are incorporated into PRC1 complexes, where they mediate targeting to trimethylated lysine 27 of histone H3 (H3K27me3) via the N-terminal chromodomain (ChD). Individual CBX paralogs have been implicated as drug targets in cancer; however, high similarities in sequence and structure among the CBX ChDs provide a major obstacle in developing selective CBX ChD inhibitors. Here we report the selection of small, focused, DNA-encoded libraries (DELs) against multiple homologous ChDs to identify modifications to a parental ligand that confer both selectivity and potency for the ChD of CBX8. This on-DNA, medicinal chemistry approach enabled the development of SW2_110A, a selective, cell-permeable inhibitor of the CBX8 ChD. SW2_110A binds CBX8 ChD with a Kd of 800 nM, with minimal 5-fold selectivity for CBX8 ChD over all other CBX paralogs in vitro. SW2_110A specifically inhibits the association of CBX8 with chromatin in cells and inhibits the proliferation of THP1 leukemia cells driven by the MLL-AF9 translocation. In THP1 cells, SW2_110A treatment results in a significant decrease in the expression of MLL-AF9 target genes, including HOXA9, validating the previously established role for CBX8 in MLL-AF9 transcriptional activation, and defining the ChD as necessary for this function. The success of SW2_110A provides great promise for the development of highly selective and cell-permeable probes for the full CBX family. In addition, the approach taken provides a proof-of-principle demonstration of how DELs can be used iteratively for optimization of both ligand potency and selectivity.

Parallel Synthesis and Screening of Supramolecular Chemosensors That Achieve Fluorescent Turn-on Detection of Drugs in Saliva.

Programming and controlling molecular recognition in aqueous solutions is increasingly common, but creating supramolecular sensors that detect analytes in biologically relevant solutions remains a nontrivial task. We report here a parallel synthesis-driven approach to create a family of self-assembling dimeric sensors that we call DimerDyes and its use for the rapid identification of salt-tolerant sensors for illicit drugs. We developed an efficient method that involves parallel synthesis and screening in crude form without the need to purify each potential sensor. Structurally diverse "hit" DimerDyes were resynthesized and purified and were each shown to assemble into homodimers in water in the programmed way. DimerDyes provided a "turn-on" fluorescence detection of multiple illicit drugs at low micromolar concentrations in water and in saliva. The combination of multiple agents into a sensor array was successfully able to detect and discriminate between closely related drugs and metabolites in multiple important drug families.

In Vitro Assessment of Putative PD-1/PD-L1 Inhibitors: Suggestions of an Alternative Mode of Action.

The programmed cell death protein 1 (PD-1) signaling axis is among the most important therapeutic targets in modern oncology. Aurigene Discovery Technologies Ltd. (Aurigene) has patented a series of peptidomimetic small molecules derived from the PD-1 protein sequence for use in targeting the interaction between PD-1 and its ligand, PD-L1. We evaluated three of Aurigene's most potent compounds in SPR binding assays. Our results showed that these compounds-each of which is known to be potently effective in a splenocyte recovery assay-do not directly inhibit the PD-1/PD-L1 interaction nor do they appear to bind to either of the constituent proteins, indicating that another mechanism is at play. As a result of these studies and upon consideration of structural features within the PD-1/PD-L1 complex, we hypothesize that the Aurigene molecules may interact with a currently unknown protein capable of regulating the PD-1 axis.

Rational Adaptation of L3MBTL1 Inhibitors to Create Small-Molecule Cbx7 Antagonists.

Chromobox homolog 7 (Cbx7) is an epigenetic modulator that is an important driver of multiple cancers. It is a methyl reader protein that operates by recognizing and binding to methylated lysine residues on specific partners. Herein we report our efforts to create low-molecular-weight inhibitors of Cbx7 by making rational structural adaptations to inhibitors of a different methyl reader protein, L3MBTL1, inhibitors that had previously been reported to be inactive against Cbx7. We evaluated each new inhibitor for Cbx7 inhibition by fluorescence polarization assay, and also confirmed the binding of selected inhibitors to Cbx7 by saturation-transfer difference NMR spectroscopy. This work identified multiple small-molecule inhibitors with modest (IC50 : 257-500 µm) potency.

Using reversible non-covalent and covalent bonds to create assemblies and equilibrating molecular networks that survive 5 molar urea.

The limits of self-assembly and host-guest chemistry in water solutions containing competitive solutes are largely unexplored. We report here a new family of self-assembling systems that are stitched together at two levels by reversible hydrazone bonds and by non-covalent self-assembly in strongly denaturing conditions. Three different hydrazides of various charge and hydrophobicity are combined with an aldehyde-containing calixarene, and each system spontaneously forms AB hydrazones that subsequently self-assemble into four-component (AB)2 structures in water. The assemblies display varying responses to added NaCl and/or urea. The most robust assembly survives completely intact in solution up to 5 M urea. We also combine the aldehyde calixarene with two different hydrazides in the same tube to create complex, competitive dynamic libraries. We report experiments in which the composition of the dynamic equilibrating library is under the control of self-assembly, allowing the systems to choose the components that form the most stable assemblies under a variety of competitive solutions conditions. These dynamic networks of equilibrating molecules maintain remarkably similar equilibrium positions under widely varying concentrations of urea and NaCl.

Analyte-Driven Disassembly and Turn-On Fluorescent Sensing in Competitive Biological Media.

Many indicator displacement assays can detect biological analytes in water, but these often have reduced performance in the presence of an unavoidable component: NaCl. We report here a new self-assembled sensor, DimerDye, that uses a novel photochemical guest-sensing mechanism and that is intrinsically tolerant of cosolutes. We synthetically integrated a dye into a calixarene macrocycle, forming two new merocyanine calixarenes (MCx-1 and MCx-2). Both compounds self-assemble into nonemissive dimers in water. The addition of good guests like trimethyllysine induces a turn-on fluorescence response of MCx-1 due to simultaneous dimer dissociation and formation of an emissive host-guest complex. DimerDyes remain functional in solutions containing the various salts, metal ions, and cofactors that are needed for enzymatic reactions. MCx-1 provides a real-time, turn-on fluorescence signal in response to the lysine methyltransferase reaction of PRDM9.

Robustness of In Vitro Selection Assays of DNA-Encoded Peptidomimetic Ligands to CBX7 and CBX8.

The identification of protein ligands from a DNA-encoded library is commonly conducted by an affinity selection assay. These assays are often not validated for robustness, raising questions about selections that fail to identify ligands and the utility of enrichment values for ranking ligand potencies. Here, we report a method for optimizing and utilizing affinity selection assays to identify potent and selective peptidic ligands to the highly related chromodomains of CBX proteins. To optimize affinity selection parameters, statistical analyses (Z' factors) were used to define the ability of selection assay conditions to identify and differentiate ligands of varying affinity. A DNA-encoded positional scanning library of peptidomimetics was constructed around a trimethyllysine-containing parent peptide, and parallel selections against the chromodomains from CBX8 and CBX7 were conducted over three protein concentrations. Relative potencies of off-DNA hit molecules were determined through a fluorescence polarization assay and were consistent with enrichments observed by DNA sequencing of the affinity selection assays. In addition, novel peptide-based ligands were discovered with increased potency and selectivity to the chromodomain of CBX8. The results indicate low DNA tag bias and show that affinity-based in vitro selection assays are sufficiently robust for both ligand discovery and determination of quantitative structure-activity relationships.

Aza-amino acid scanning of chromobox homolog 7 (CBX7) ligands.

An aza-amino acid scan of peptide inhibitors of the chromobox homolog 7 (CBX7) was performed to study the conformational requirements for affinity to the methyllysine reader protein. Twelve azapeptide analogues were prepared using three different approaches employing respectively N-(Fmoc)aza-amino acid chlorides and submonomer azapeptide synthesis to install systematically aza-residues at the first four residues of the peptide, as well as to provide aza-lysine residues possessing saturated and unsaturated side chains. The aza-peptide ligands were evaluated in a chromobox homolog 7 binding assay, providing useful insight into structural requirements for affinity. Copyright © 2017 European Peptide Society and John Wiley & Sons, Ltd.