Evidence for Widespread Class II Microcins in Enterobacterales Genomes.

Microcins are a class of antimicrobial peptides produced by certain Gram-negative bacterial species to kill or inhibit the growth of competing bacteria. Only 10 unique, experimentally validated class II microcins have been identified, and the majority of these come from Escherichia coli. Although the current representation of microcins is sparse, they exhibit a diverse array of molecular functionalities, uptake mechanisms, and target specificities. This broad diversity from such a small representation suggests that microcins may have untapped potential for bioprospecting peptide antibiotics from genomic data sets. We used a systematic bioinformatics approach to search for verified and novel class II microcins in E. coli and other species within its family, Enterobacteriaceae. Nearly one-quarter of the E. coli genome assemblies contained one or more microcins, where the prevalence of hits to specific microcins varied by isolate phylogroup. E. coli isolates from human extraintestinal and poultry meat sources were enriched for microcins, while those from freshwater were depleted. Putative microcins were found in various abundances across all five distinct phylogenetic lineages of Enterobacteriaceae, with a particularly high prevalence in the "Klebsiella" clade. Representative genome assemblies from species across the Enterobacterales order, as well as a few outgroup species, also contained putative microcin sequences. This study suggests that microcins have a complicated evolutionary history, spanning far beyond our limited knowledge of the currently validated microcins. Efforts to functionally characterize these newly identified microcins have great potential to open a new field of peptide antibiotics and microbiome modulators and elucidate the ways in which bacteria compete with each other. IMPORTANCE Class II microcins are small bacteriocins produced by strains of Gram-negative bacteria in the Enterobacteriaceae. They are generally understood to play a role in interbacterial competition, although direct evidence of this is limited, and they could prove informative in developing new peptide antibiotics. However, few examples of verified class II microcins exist, and novel microcins are difficult to identify due to their sequence diversity, making it complicated to study them as a group. Here, we overcome this limitation by developing a bioinformatics pipeline to detect microcins in silico. Using this pipeline, we demonstrate that both verified and novel class II microcins are widespread within and outside the Enterobacteriaceae, which has not been systematically shown previously. The observed prevalence of class II microcins suggests that they are ecologically important, and the elucidation of novel microcins provides a resource that can be used to expand our knowledge of the structure and function of microcins as antibacterials.

Killer Knots: Molecular Evolution of Inhibitor Cystine Knot Toxins in Wandering Spiders (Araneae: Ctenidae).

Venom expressed by the nearly 50,000 species of spiders on Earth largely remains an untapped reservoir of a diverse array of biomolecules with potential for pharmacological and agricultural applications. A large fraction of the noxious components of spider venoms are a functionally diverse family of structurally related polypeptides with an inhibitor cystine knot (ICK) motif. The cysteine-rich nature of these toxins makes structural elucidation difficult, and most studies have focused on venom components from the small handful of medically relevant spider species such as the highly aggressive Brazilian wandering spider Phoneutria nigriventer. To alleviate difficulties associated with the study of ICK toxins in spiders, we devised a comprehensive approach to explore the evolutionary patterns that have shaped ICK functional diversification using venom gland transcriptomes and proteomes from phylogenetically distinct lineages of wandering spiders and their close relatives. We identified 626 unique ICK toxins belonging to seven topological elaborations. Phylogenetic tests of episodic diversification revealed distinct regions between cysteine residues that demonstrated differential evidence of positive or negative selection, which may have structural implications towards the specificity and efficacy of these toxins. Increased taxon sampling and whole genome sequencing will provide invaluable insights to further understand the evolutionary processes that have given rise to this diverse class of toxins.

Adapting antibacterial display to identify serum-active macrocyclic peptide antibiotics.

The lack of available treatments for many antimicrobial-resistant infections highlights the critical need for antibiotic discovery innovation. Peptides are an underappreciated antibiotic scaffold because they often suffer from proteolytic instability and toxicity toward human cells, making in vivo use challenging. To investigate sequence factors related to serum activity, we adapt an antibacterial display technology to screen a library of peptide macrocycles for antibacterial potential directly in human serum. We identify dozens of new macrocyclic peptide antibiotic sequences and find that serum activity within our library is influenced by peptide length, cationic charge, and the number of disulfide bonds present. Interestingly, an optimized version of our most active lead peptide permeates the outer membrane of Gram-negative bacteria without strong inner-membrane disruption and kills bacteria slowly while causing cell elongation. This contrasts with traditional cationic antimicrobial peptides, which kill rapidly via lysis of both bacterial membranes. Notably, this optimized variant is not toxic to mammalian cells and retains its function in vivo, suggesting therapeutic promise. Our results support the use of more physiologically relevant conditions when screening peptides for antimicrobial activity which retain in vivo functionality.

Two sequence- and two structure-based ML models have learned different aspects of protein biochemistry.

Deep learning models are seeing increased use as methods to predict mutational effects or allowed mutations in proteins. The models commonly used for these purposes include large language models (LLMs) and 3D Convolutional Neural Networks (CNNs). These two model types have very different architectures and are commonly trained on different representations of proteins. LLMs make use of the transformer architecture and are trained purely on protein sequences whereas 3D CNNs are trained on voxelized representations of local protein structure. While comparable overall prediction accuracies have been reported for both types of models, it is not known to what extent these models make comparable specific predictions and/or generalize protein biochemistry in similar ways. Here, we perform a systematic comparison of two LLMs and two structure-based models (CNNs) and show that the different model types have distinct strengths and weaknesses. The overall prediction accuracies are largely uncorrelated between the sequence- and structure-based models. Overall, the two structure-based models are better at predicting buried aliphatic and hydrophobic residues whereas the two LLMs are better at predicting solvent-exposed polar and charged amino acids. Finally, we find that a combined model that takes the individual model predictions as input can leverage these individual model strengths and results in significantly improved overall prediction accuracy.

Adapting antibacterial display to identify serum active macrocyclic peptide antibiotics.

The lack of available treatments for many antimicrobial resistant infections highlights the critical need for antibiotic discovery innovation. Peptides are an underappreciated antibiotic scaffold because they often suffer from proteolytic instability and toxicity towards human cells, making in vivo use challenging. To investigate sequence factors related to serum activity, we adapt an antibacterial display technology to screen a library of peptide macrocycles for antibacterial potential directly in human serum. We identify dozens of new macrocyclic peptide antibiotic sequences and find that serum activity within our library is influenced by peptide length, cationic charge, and the number of disulfide bonds present. Interestingly, an optimized version of our most active lead peptide permeates the outer membrane of gram-negative bacteria without strong inner membrane disruption and kills bacteria slowly while causing cell elongation. This contrasts with traditional cationic antimicrobial peptides, which kill rapidly via lysis of both bacterial membranes. Notably, this optimized variant is not toxic to mammalian cells and retains its function in vivo , suggesting therapeutic promise. Our results support the use of more physiologically relevant conditions when screening peptides for antimicrobial activity which retain in vivo functionality. Significance: Traditional methods of natural antibiotic discovery are low throughput and cannot keep pace with the development of antimicrobial resistance. Synthetic peptide display technologies offer a high-throughput means of screening drug candidates, but rarely consider functionality beyond simple target binding and do not consider retention of function in vivo . Here, we adapt a function-based, antibacterial display technology to screen a large library of peptide macrocycles directly for bacterial growth inhibition in human serum. This screen identifies an optimized non-toxic macrocyclic peptide antibiotic retaining in vivo function, suggesting this advancement could increase clinical antibiotic discovery efficiency.

Designing and identifying ß-hairpin peptide macrocycles with antibiotic potential.

Peptide macrocycles are a rapidly emerging class of therapeutic, yet the design of their structure and activity remains challenging. This is especially true for those with ß-hairpin structure due to weak folding properties and a propensity for aggregation. Here, we use proteomic analysis and common antimicrobial features to design a large peptide library with macrocyclic ß-hairpin structure. Using an activity-driven high-throughput screen, we identify dozens of peptides killing bacteria through selective membrane disruption and analyze their biochemical features via machine learning. Active peptides contain a unique constrained structure and are highly enriched for cationic charge with arginine in their turn region. Our results provide a synthetic strategy for structured macrocyclic peptide design and discovery while also elucidating characteristics important for ß-hairpin antimicrobial peptide activity.

ExoDx prostate test as a predictor of outcomes of high-grade prostate cancer - an interim analysis.

BACKGROUND: Patient outcomes were assessed based on a pre-biopsy ExoDx Prostate (EPI) score at 2.5 years of the 5-year follow-up of ongoing prostate biopsy Decision Impact Trial of the ExoDx Prostate (IntelliScore). METHODS: Prospective, blinded, randomized, multisite clinical utility study was conducted from June 2017 to May 2018 (NCT03235687). Urine samples were collected from 1049 men (≥50 years old) with a PSA 2-10 ng/mL being considered for a prostate biopsy. Patients were randomized to EPI vs. standard of care (SOC). All had an EPI test, but only EPI arm received results during biopsy decision process. Clinical outcomes, time to biopsy and pathology were assessed among low (<15.6) or high (≥15.6) EPI scores. RESULTS: At 2.5 years, 833 patients had follow-up data. In the EPI arm, biopsy rates remained lower for low-risk EPI scores than high-risk EPI scores (44.6% vs 79.0%, p < 0.001), whereas biopsy rates were identical in SOC arm regardless of EPI score (59.6% vs 58.8%, p = 0.99). Also in the EPI arm, the average time from EPI testing to first biopsy was longer for low-risk EPI scores compared to high-risk EPI scores (216 vs. 69 days; p < 0.001). Similarly, the time to first biopsy was longer with EPI low-risk scores in EPI arm compared to EPI low-risk scores in SOC arm (216 vs 80 days; p < 0.001). At 2.5 years, patients with low-risk EPI scores from both arms had less HGPC than high-risk EPI score patients (7.9% vs 26.8%, p < 0.001) and the EPI arm found 21.8% more HGPC than the SOC arm. CONCLUSIONS: This follow-up analysis captures subsequent biopsy outcomes and demonstrates that men receiving EPI low-risk scores (<15.6) significantly defer the time to first biopsy and remain at a very low pathologic risk by 2.5-years after the initial study. The EPI test risk stratification identified low-risk patients that were not found with the SOC.

Synthetic antibacterial discovery of symbah-1, a macrocyclic ß-hairpin peptide antibiotic.

The rapid development and spread of antibiotic resistance necessitate the development of novel strategies for antibiotic discovery. Symbah-1, a synthetic peptide antibiotic, was identified in a high-throughput antibacterial screen of random peptide sequences. Symbah-1 functions through membrane disruption and contains broad spectrum bactericidal activity against several drug-resistant pathogens. Circular dichroism and high-resolution mass spectrometry indicate symbah-1 has a ß-hairpin structure induced by lipopolysaccharide and is cyclized via an intramolecular disulfide bond. Together these data classify symbah-1 as an uncommon synthetic member of the ß-hairpin antimicrobial peptide class. Symbah-1 displays low hemolysis but loses activity in human serum. Characterization of a symbah-1 peptide library identified two variants with increased serum activity and protease resistance. The method of discovery and subsequent characterization of symbah-1 suggests large synthetic peptide libraries bias toward macrocyclic ß-hairpin structure could be designed and screened to rapidly expand and better understand this rare peptide antibiotic class.

Pick Your Poison: Molecular Evolution of Venom Proteins in Asilidae (Insecta: Diptera).

Robber flies are an understudied family of venomous, predatory Diptera. With the recent characterization of venom from three asilid species, it is possible, for the first time, to study the molecular evolution of venom genes in this unique lineage. To accomplish this, a novel whole-body transcriptome of Eudioctria media was combined with 10 other publicly available asiloid thoracic or salivary gland transcriptomes to identify putative venom gene families and assess evidence of pervasive positive selection. A total of 348 gene families of sufficient size were analyzed, and 33 of these were predicted to contain venom genes. We recovered 151 families containing homologs to previously described venom proteins, and 40 of these were uniquely gained in Asilidae. Our gene family clustering suggests that many asilidin venom gene families are not natural groupings, as delimited by previous authors, but instead form multiple discrete gene families. Additionally, robber fly venoms have relatively few sites under positive selection, consistent with the hypothesis that the venoms of older lineages are dominated by negative selection acting to maintain toxic function.

TOXIFY: a deep learning approach to classify animal venom proteins.

In the era of Next-Generation Sequencing and shotgun proteomics, the sequences of animal toxigenic proteins are being generated at rates exceeding the pace of traditional means for empirical toxicity verification. To facilitate the automation of toxin identification from protein sequences, we trained Recurrent Neural Networks with Gated Recurrent Units on publicly available datasets. The resulting models are available via the novel software package TOXIFY, allowing users to infer the probability of a given protein sequence being a venom protein. TOXIFY is more than 20X faster and uses over an order of magnitude less memory than previously published methods. Additionally, TOXIFY is more accurate, precise, and sensitive at classifying venom proteins.

FUSTr: a tool to find gene families under selection in transcriptomes.

BACKGROUND: The recent proliferation of large amounts of biodiversity transcriptomic data has resulted in an ever-expanding need for scalable and user-friendly tools capable of answering large scale molecular evolution questions. FUSTr identifies gene families involved in the process of adaptation. This is a tool that finds genes in transcriptomic datasets under strong positive selection that automatically detects isoform designation patterns in transcriptome assemblies to maximize phylogenetic independence in downstream analysis. RESULTS: When applied to previously studied spider transcriptomic data as well as simulated data, FUSTr successfully grouped coding sequences into proper gene families as well as correctly identified those under strong positive selection in relatively little time. CONCLUSIONS: FUSTr provides a useful tool for novice bioinformaticians to characterize the molecular evolution of organisms throughout the tree of life using large transcriptomic biodiversity datasets and can utilize multi-processor high-performance computational facilities.