Discovery, biochemical characterization, and bioengineering of cyanobactin prenyltransferases.

Prenylation is a post-translational modification (PTM) widely found in primary and secondary metabolism. This modification can enhance the lipophilicity of molecules, enabling them to interact with lipid membranes more effectively. The prenylation of peptides is often carried out by cyanobactin prenyltransferases (PTases) from cyanobacteria. These enzymes are of interest due to their ability to add prenyl groups to unmodified peptides, thus making them more effective therapeutics through the subsequent acquisition of increased membrane permeability and bioavailability. Herein we review the current knowledge of cyanobactin PTases, focusing on their discovery, biochemistry, and bioengineering, and highlight the potential application of them as peptide alkylation biocatalysts to generate peptide therapeutics.

Selective ligand recognition and activation of somatostatin receptors SSTR1 and SSTR3.

Somatostatin receptors (SSTRs) exert critical biological functions such as negatively regulating hormone release and cell proliferation, making them popular targets for developing therapeutics to treat endocrine disorders, especially neuroendocrine tumors. Although several panagonists mimicking the endogenous ligand somatostatin are available, the development of more effective and safer somatostatinergic therapies is limited due to a lack of molecular understanding of the ligand recognition and regulation of divergent SSTR subtypes. Here, we report four cryoelectron microscopy structures of Gi-coupled SSTR1 and SSTR3 activated by distinct agonists, including the FDA-approved panagonist pasireotide as well as their selective small molecule agonists L-797591 and L-796778. Our structures reveal a conserved recognition pattern of pasireotide in SSTRs attributed to the binding with a conserved extended binding pocket, distinct from SST14, octreotide, and lanreotide. Together with mutagenesis analyses, our structures further reveal the dynamic feature of ligand binding pockets in SSTR1 and SSTR3 to accommodate divergent agonists, the key determinants of ligand selectivity lying across the orthosteric pocket of different SSTR subtypes, as well as the molecular mechanism underlying diversity and conservation of receptor activation. Our work provides a framework for rational design of subtype-selective SSTR ligands and may facilitate drug development efforts targeting SSTRs with improved therapeutic efficacy and reduced side effects.

Structural basis for activation of somatostatin receptor 5 by cyclic neuropeptide agonists.

Somatostatin receptor 5 (SSTR5) is an important G protein-coupled receptor and drug target for neuroendocrine tumors and pituitary disorders. This study presents two high-resolution cryogenicelectron microscope structures of the SSTR5-Gi complexes bound to the cyclic neuropeptide agonists, cortistatin-17 (CST17) and octreotide, with resolutions of 2.7 Å and 2.9 Å, respectively. The structures reveal that binding of these peptides causes rearrangement of a "hydrophobic lock", consisting of residues from transmembrane helices TM3 and TM6. This rearrangement triggers outward movement of TM6, enabling Gαi protein engagement and receptor activation. In addition to hydrophobic interactions, CST17 forms conserved polar contacts similar to somatostatin-14 binding to SSTR2, while further structural and functional analysis shows that extracellular loops differently recognize CST17 and octreotide. These insights elucidate agonist selectivity and activation mechanisms of SSTR5, providing valuable guidance for structure-based drug development targeting this therapeutically relevant receptor.

CyclicPepedia: a knowledge base of natural and synthetic cyclic peptides.

Cyclic peptides offer a range of notable advantages, including potent antibacterial properties, high binding affinity and specificity to target molecules, and minimal toxicity, making them highly promising candidates for drug development. However, a comprehensive database that consolidates both synthetically derived and naturally occurring cyclic peptides is conspicuously absent. To address this void, we introduce CyclicPepedia (https://www.biosino.org/iMAC/cyclicpepedia/), a pioneering database that encompasses 8744 known cyclic peptides. This repository, structured as a composite knowledge network, offers a wealth of information encompassing various aspects of cyclic peptides, such as cyclic peptides' sources, categorizations, structural characteristics, pharmacokinetic profiles, physicochemical properties, patented drug applications, and a collection of crucial publications. Supported by a user-friendly knowledge retrieval system and calculation tools specifically designed for cyclic peptides, CyclicPepedia will be able to facilitate advancements in cyclic peptide drug development.

HighFold: accurately predicting structures of cyclic peptides and complexes with head-to-tail and disulfide bridge constraints.

In recent years, cyclic peptides have emerged as a promising therapeutic modality due to their diverse biological activities. Understanding the structures of these cyclic peptides and their complexes is crucial for unlocking invaluable insights about protein target-cyclic peptide interaction, which can facilitate the development of novel-related drugs. However, conducting experimental observations is time-consuming and expensive. Computer-aided drug design methods are not practical enough in real-world applications. To tackles this challenge, we introduce HighFold, an AlphaFold-derived model in this study. By integrating specific details about the head-to-tail circle and disulfide bridge structures, the HighFold model can accurately predict the structures of cyclic peptides and their complexes. Our model demonstrates superior predictive performance compared to other existing approaches, representing a significant advancement in structure-activity research. The HighFold model is openly accessible at https://github.com/hongliangduan/HighFold.

CycPeptMP: enhancing membrane permeability prediction of cyclic peptides with multi-level molecular features and data augmentation.

Cyclic peptides are versatile therapeutic agents that boast high binding affinity, minimal toxicity, and the potential to engage challenging protein targets. However, the pharmaceutical utility of cyclic peptides is limited by their low membrane permeability-an essential indicator of oral bioavailability and intracellular targeting. Current machine learning-based models of cyclic peptide permeability show variable performance owing to the limitations of experimental data. Furthermore, these methods use features derived from the whole molecule that have traditionally been used to predict small molecules and ignore the unique structural properties of cyclic peptides. This study presents CycPeptMP: an accurate and efficient method to predict cyclic peptide membrane permeability. We designed features for cyclic peptides at the atom-, monomer-, and peptide-levels and seamlessly integrated these into a fusion model using deep learning technology. Additionally, we applied various data augmentation techniques to enhance model training efficiency using the latest data. The fusion model exhibited excellent prediction performance for the logarithm of permeability, with a mean absolute error of $0.355$ and correlation coefficient of $0.883$. Ablation studies demonstrated that all feature levels contributed and were relatively essential to predicting membrane permeability, confirming the effectiveness of augmentation to improve prediction accuracy. A comparison with a molecular dynamics-based method showed that CycPeptMP accurately predicted peptide permeability, which is otherwise difficult to predict using simulations.

De novo development of small cyclic peptides that are orally bioavailable.

Cyclic peptides can bind challenging disease targets with high affinity and specificity, offering enormous opportunities for addressing unmet medical needs. However, as with biological drugs, most cyclic peptides cannot be applied orally because they are rapidly digested and/or display low absorption in the gastrointestinal tract, hampering their development as therapeutics. In this study, we developed a combinatorial synthesis and screening approach based on sequential cyclization and one-pot peptide acylation and screening, with the possibility of simultaneously interrogating activity and permeability. In a proof of concept, we synthesized a library of 8,448 cyclic peptides and screened them against the disease target thrombin. Our workflow allowed multiple iterative cycles of library synthesis and yielded cyclic peptides with nanomolar affinities, high stabilities and an oral bioavailability (%F) as high as 18% in rats. This method for generating orally available peptides is general and provides a promising push toward unlocking the full potential of peptides as therapeutics.

Evolution-guided discovery of antibiotics that inhibit peptidoglycan remodelling.

Addressing the ongoing antibiotic crisis requires the discovery of compounds with novel mechanisms of action that are capable of treating drug-resistant infections1. Many antibiotics are sourced from specialized metabolites produced by bacteria, particularly those of the Actinomycetes family2. Although actinomycete extracts have traditionally been screened using activity-based platforms, this approach has become unfavourable owing to the frequent rediscovery of known compounds. Genome sequencing of actinomycetes reveals an untapped reservoir of biosynthetic gene clusters, but prioritization is required to predict which gene clusters may yield promising new chemical matter2. Here we make use of the phylogeny of biosynthetic genes along with the lack of known resistance determinants to predict divergent members of the glycopeptide family of antibiotics that are likely to possess new biological activities. Using these predictions, we uncovered two members of a new functional class of glycopeptide antibiotics-the known glycopeptide antibiotic complestatin and a newly discovered compound we call corbomycin-that have a novel mode of action. We show that by binding to peptidoglycan, complestatin and corbomycin block the action of autolysins-essential peptidoglycan hydrolases that are required for remodelling of the cell wall during growth. Corbomycin and complestatin have low levels of resistance development and are effective in reducing bacterial burden in a mouse model of skin MRSA infection.

MuCoCP: a priori chemical knowledge-based multimodal contrastive learning pre-trained neural network for the prediction of cyclic peptide membrane penetration ability.

MOTIVATION: There has been a burgeoning interest in cyclic peptide therapeutics due to their various outstanding advantages and strong potential for drug formation. However, it is undoubtedly costly and inefficient to use traditional wet lab methods to clarify their biological activities. Using artificial intelligence instead is a more energy-efficient and faster approach. MuCoCP aims to build a complete pre-trained model for extracting potential features of cyclic peptides, which can be fine-tuned to accurately predict cyclic peptide bioactivity on various downstream tasks. To maximize its effectiveness, we use a novel data augmentation method based on a priori chemical knowledge and multiple unsupervised training objective functions to greatly improve the information-grabbing ability of the model. RESULTS: To assay the efficacy of the model, we conducted validation on the membrane-permeability of cyclic peptides which achieved an accuracy of 0.87 and R-squared of 0.503 on CycPeptMPDB using semi-supervised training and obtained an accuracy of 0.84 and R-squared of 0.384 using a model with frozen parameters on an external dataset. This result has achieved state-of-the-art, which substantiates the stability and generalization capability of MuCoCP. It means that MuCoCP can fully explore the high-dimensional information of cyclic peptides and make accurate predictions on downstream bioactivity tasks, which will serve as a guide for the future de novo design of cyclic peptide drugs and promote the development of cyclic peptide drugs. AVAILABILITY AND IMPLEMENTATION: All code used in our proposed method can be found at https://github.com/lennonyu11234/MuCoCP.

HELM-GPT: de novo macrocyclic peptide design using generative pre-trained transformer.

MOTIVATION: Macrocyclic peptides hold great promise as therapeutics targeting intracellular proteins. This stems from their remarkable ability to bind flat protein surfaces with high affinity and specificity while potentially traversing the cell membrane. Research has already explored their use in developing inhibitors for intracellular proteins, such as KRAS, a well-known driver in various cancers. However, computational approaches for de novo macrocyclic peptide design remain largely unexplored. RESULTS: Here, we introduce HELM-GPT, a novel method that combines the strength of the hierarchical editing language for macromolecules (HELM) representation and generative pre-trained transformer (GPT) for de novo macrocyclic peptide design. Through reinforcement learning (RL), our experiments demonstrate that HELM-GPT has the ability to generate valid macrocyclic peptides and optimize their properties. Furthermore, we introduce a contrastive preference loss during the RL process, further enhanced the optimization performance. Finally, to co-optimize peptide permeability and KRAS binding affinity, we propose a step-by-step optimization strategy, demonstrating its effectiveness in generating molecules fulfilling both criteria. In conclusion, the HELM-GPT method can be used to identify novel macrocyclic peptides to target intracellular proteins. AVAILABILITY AND IMPLEMENTATION: The code and data of HELM-GPT are freely available on GitHub (https://github.com/charlesxu90/helm-gpt).

Atomic structure of Lanreotide nanotubes revealed by cryo-EM.

Functional and versatile nano- and microassemblies formed by biological molecules are found at all levels of life, from cell organelles to full organisms. Understanding the chemical and physicochemical determinants guiding the formation of these assemblies is crucial not only to understand the biological processes they carry out but also to mimic nature. Among the synthetic peptides forming well-defined nanostructures, the octapeptide Lanreotide has been considered one of the best characterized, in terms of both the atomic structure and its self-assembly process. In the present work, we determined the atomic structure of Lanreotide nanotubes at 2.5-Å resolution by cryoelectron microscopy (cryo-EM). Surprisingly, the asymmetric unit in the nanotube contains eight copies of the peptide, forming two tetramers. There are thus eight different environments for the peptide, and eight different conformations in the nanotube. The structure built from the cryo-EM map is strikingly different from the molecular model, largely based on X-ray fiber diffraction, proposed 20 y ago. Comparison of the nanotube with a crystal structure at 0.83-Å resolution of a Lanreotide derivative highlights the polymorphism for this peptide family. This work shows once again that higher-order assemblies formed by even well-characterized small peptides are very difficult to predict.

Targeted Inhibition of Lymphovascular Invasion Formation with CREKA Peptide-Modified Silicasomes to Boost Chemotherapy in Bladder Cancer.

Despite its significant clinical efficacy as a first-line treatment for advanced bladder cancer, cisplatin-based chemotherapy provides a limited benefit for patients with lymphovascular invasion (LVI), which is characterized by the presence of tumor emboli within blood vessels and associated with enhanced cisplatin resistance and metastatic potential. Notably, platelets, a critical component of LVI, hinder the delivery of chemotherapeutic agents to tumors and facilitate metastasis. Consequently, platelet function inhibition holds the potential to disrupt LVI formation, as well as augment the antitumor activity of cisplatin. Herein, we developed a tumor microenvironment-targeted nanodrug with lipid-coated mesoporous silica nanoparticles (silicasomes) that synergistically combines cisplatin with an antiplatelet agent, tirofiban, for bladder cancer treatment. The customized nanodrug can concurrently prevent LVI formation and enhance the chemotherapeutic efficacy without significant adverse effects. This study supports the integration of chemotherapy and antiplatelet therapy via a silicasome-based nanosystem as a highly promising strategy for bladder cancer management.

Clustering of the Membrane Protein by Molecular Self-Assembly Downregulates the Signaling Pathway for Cancer Cell Inhibition.

This work reports a cyclic peptide appended self-assembled scaffold that recognizes the membrane protein EGFR and arrests the EGFR signaling through multivalent interactions by assembly-induced aggregation. When incubated with cells, the oligomers of PAD-1 first recognize the overexpressed EGFR on cancer cell membranes for arresting EGFR, which then initiates cellular uptake through endocytosis. The accumulation of PAD-1 and EGFR in the lysosome results in the formation of nanofibers, leading to the lysosomal membrane permeabilization (LMP). These processes disrupt the homeostasis of EGFR and inhibit the downstream signaling transduction of EGFR for cancer cell survival. Moreover, LMP induced the release of protein aggregates that could generate endoplasmic reticulum (ER) stress, resulting in cancer cell death selectively. In vivo studies indicate the efficient antitumor efficiency of PAD-1 in tumor-bearing mice. As a first example, this work provides an alternative strategy for controlling protein behavior for tuning cellular events in living cells.

Time-resolved fluorescence of tryptophan characterizes membrane perturbation by cyclic lipopeptides.

Viscosin is a membrane-permeabilizing, cyclic lipopeptide (CLiP) produced by Pseudomonas species. Here, we have studied four synthetic analogs (L1W, V4W, L5W, and L7W), each with one leucine (Leu; L) or valine residue exchanged for tryptophan (Trp; W) by means of time-resolved fluorescence spectroscopy of Trp. To this end, we recorded the average fluorescence lifetime, rotational correlation time and limiting anisotropy, dipolar relaxation time and limiting extent of relaxation, rate constant of acrylamide quenching, effect of H2O-D2O exchange, and time-resolved half-width of the spectrum in the absence and presence of POPC (1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine) liposomes. Structure, localization, and hydration of the peptides were described by molecular dynamics simulations. The combination of the parameters provides a good description of the molecular environments of the Trp positions and the behavior of viscosin as a whole. Of particular value for characterizing the impact of viscosin on the membrane is the dipolar relaxation of Trp4 in V4W, which is deeply embedded in the hydrophobic core. The limiting relaxation level represents the membrane perturbation-unlike typical membrane probes-at the site of the perturbant. Fractions of Trp4 relax at different rates; the one not in contact with water upon excitation relaxes via recruitment of a water molecule on the 10-ns timescale. This rate is sensitive to the concerted membrane perturbation by more than one lipopeptide, which appears at high lipopeptide concentration and is assumed a prerequisite for the final formation of a membrane-permeabilizing defect. Trp7 relaxes primarily with respect to neighboring Ser residues. Trp5 flips between a membrane-inserted and surface-exposed orientation.

Determinants of maturation of the Staphylococcus aureus autoinducing peptide.

The accessory gene regulatory (Agr) system is required for virulence factor gene expression and pathogenesis of Staphylococcus aureus. The Agr system is activated in response to the accumulation of a cyclic autoinducing peptide (AIP), which is matured and secreted by the bacterium. The precursor of AIP, AgrD, consists of the AIP flanked by an N-terminal [Formula: see text]-helical Leader and a charged C-terminal tail. AgrD is matured to AIP by the action of two proteases, AgrB and MroQ. AgrB cleaves the C-terminal tail and promotes the formation of a thiolactone ring, whereas MroQ cleaves the N-terminal Leader in a manner that depends on the four-amino acid linker immediately following a conserved IG helix breaker motif. However, the attributes of AgrD that dictate the sequence of events in peptide maturation are not fully defined. Here, we used engineered AgrD peptide intermediates to ascertain the sufficiency of MroQ for N-terminal peptide cleavage, peptide export, and generation of mature AIP. We found that MroQ promotes the removal of the N-terminal Leader peptide from both linear and cyclic peptide intermediates, while peptide cyclization remained essential for signaling. The expression of the Leader peptide in isolation was sufficient for MroQ-dependent cleavage proximal to the four-amino-acid linker. In addition, active site mutations within AgrB destabilized full-length AgrD and thiolactone-containing intermediates and prevented the release of the Leader peptide. Altogether, our data support a tandem peptide maturation event involving both MroQ and AgrB that appears to couple protease activity and export of bioactive AIP.IMPORTANCEThe accessory gene regulatory (Agr) system is important for S. aureus pathogenesis. Activation of the Agr system requires recognition of a cyclic peptide pheromone, which must be fully matured to exert its biological activity. The complete events in cyclic peptide maturation and export from the bacterial cell remain to be fully defined. We and others recently discovered that the membrane peptidase MroQ is required for pheromone maturation. This study builds off the identification of MroQ and considers the attributes of the pheromone pro-peptide that are required for MroQ-mediated processing as well as uncovers features important for peptide stability and export. Overall, the findings in this study have implications for understanding bacterial pheromone maturation and virulence.

Cyclic Peptides from Graspetide Biosynthesis and Native Chemical Ligation.

The ribosomally synthesized and post-translationally modified peptide (RiPP) superfamily of natural products includes many examples of cyclic peptides with diverse macrocyclization chemistries. The graspetides, one family of macrocyclized RiPPs, harbor side chain-side chain ester or amide linkages. We recently reported the structure and biosynthesis of the graspetide pre-fuscimiditide, a 22-amino-acid (aa) peptide with two ester cross-links forming a stem-loop structure. These cross-links are introduced by a single graspetide synthetase, the ATP-grasp enzyme ThfB. Here we show that ThfB can also catalyze the formation of amide or thioester cross-links in prefuscimiditide, with thioester formation being especially efficient. We further show that upon proteolysis to reveal an N-terminal cysteine residue, the thioester-linked peptide rapidly and quantitatively rearranges via native chemical ligation into an isopeptide-bonded head-to-tail cyclic peptide. The solution structure of this rearranged peptide was determined by using 2D NMR spectroscopy experiments. Our methodology offers a straightforward recombinant route to head-to-tail cyclic peptides.

Crafting Unnatural Peptide Macrocycles via Rh(III)-Catalyzed Carboamidation.

Contemporary developments in the field of peptide macrocyclization methodology are imperative for enabling the advance of drug design in medicinal chemistry. This report discloses a Rh(III)-catalyzed macrocyclization via carboamidation, reacting acryloyl-peptide-dioxazolone precursors and arylboronic acids to form complex cyclic peptides with concomitant incorporation of noncanonical α-amino acids. The diverse and modular technology allows for expedient access to a wide variety of cyclic peptides from 4 to 15 amino acids in size and features simultaneous formation of unnatural phenylalanine and tyrosine derivatives with up to >20:1 diastereoselectivity. The reaction showcases an expansive substrate scope with 45 examples and is compatible with the majority of standard protected amino acids used in Fmoc-solid phase peptide synthesis. The methodology is applied to the synthesis of multiple peptidomimetic macrocyclic analogs, including derivatives of cyclosomatostatin and gramicidin S.

Membrane Permeability in a Large Macrocyclic Peptide Driven by a Saddle-Shaped Conformation.

The effort to modulate challenging protein targets has stimulated interest in ligands that are larger and more complex than typical small-molecule drugs. While combinatorial techniques such as mRNA display routinely produce high-affinity macrocyclic peptides against classically undruggable targets, poor membrane permeability has limited their use toward primarily extracellular targets. Understanding the passive membrane permeability of macrocyclic peptides would, in principle, improve our ability to design libraries whose leads can be more readily optimized against intracellular targets. Here, we investigate the permeabilities of over 200 macrocyclic 10-mers using the thioether cyclization motif commonly found in mRNA display macrocycle libraries. We identified the optimal lipophilicity range for achieving permeability in thioether-cyclized 10-mer cyclic peptide-peptoid hybrid scaffolds and showed that permeability could be maintained upon extensive permutation in the backbone. In one case, changing a single amino acid from d-Pro to d-NMe-Ala, representing the loss of a single methylene group in the side chain, resulted in a highly permeable scaffold in which the low-dielectric conformation shifted from the canonical cross-beta geometry of the parent compounds into a novel saddle-shaped fold in which all four backbone NH groups were sequestered from the solvent. This work provides an example by which pre-existing physicochemical knowledge of a scaffold can benefit the design of macrocyclic peptide mRNA display libraries, pointing toward an approach for biasing libraries toward permeability by design. Moreover, the compounds described herein are a further demonstration that geometrically diverse, highly permeable scaffolds exist well beyond conventional drug-like chemical space.

Synthesis of Membrane-Permeable Macrocyclic Peptides via Imidazopyridinium Grafting.

Macrocyclic peptides (MPs) are a class of compounds that have been shown to be particularly well suited for engaging difficult protein targets. However, their utility is limited by their generally poor cell permeability and bioavailability. Here, we report an efficient solid-phase synthesis of novel MPs by trapping a reversible intramolecular imine linkage with a 2-formyl- or 2-keto-pyridine to create an imidazopyridinium (IP+)-linked ring. This chemistry is useful for the creation of macrocycles of different sizes and geometries, including head-to-side and side-to-side chain configurations. Many of the IP+-linked MPs exhibit far better passive membrane permeability than expected for "beyond Rule of 5" molecules, in some cases exceeding that of much lower molecular weight, traditional drug molecules. We demonstrate that this chemistry is suitable for the creation of libraries of IP+-linked MPs and show that these libraries can be mined for protein ligands.

In Vitro Selection of Macrocyclic l-α/d-α/ß/γ-Hybrid Peptides Targeting IFN-γ/IFNGR1 Protein-Protein Interaction.

Nonproteinogenic amino acids, including d-α-, ß-, and γ-amino acids, present in bioactive peptides play pivotal roles in their biochemical activities and proteolytic stabilities. d-α-Amino acids (dαAA) are widely used building blocks that can enhance the proteolytic stability. Cyclic ß2,3-amino acids (cßAA), for instance, can fold peptides into rigid secondary structures, improving the binding affinity and proteolytic stability. Cyclic γ2,4-amino acids (cγAA) are recently highlighted as rigid residues capable of preventing the proteolysis of flanking residues. Simultaneous incorporation of all dαAA, cßAA, and cγAA into a peptide is expected to yield l-α/d-α/ß/γ-hybrid peptides with improved stability and potency. Despite challenges in the ribosomal incorporation of multiple nonproteinogenic amino acids, our engineered tRNAPro1E2 successfully reaches such a difficulty. Here, we report the ribosomal synthesis of macrocyclic l-α/d-α/ß/γ-hybrid peptide libraries and their application to in vitro selection against interferon gamma receptor 1 (IFNGR1). One of the resulting l-α/d-α/ß/γ-hybrid peptides, IB1, exhibited remarkable inhibitory activity against the IFN-γ/IFNGR1 protein-protein interaction (PPI) (IC50 = 12 nM), primarily attributed to the presence of a cßAA in the sequence. Additionally, cγAAs and dαAAs in the resulting peptides contributed to their serum stability. Furthermore, our peptides effectively inhibit IFN-γ/IFNGR1 PPI at the cellular level (best IC50 = 0.75 µM). Altogether, our platform expands the chemical space available for exploring peptides with high activity and stability, thereby enhancing their potential for drug discovery.