Eriko Takano

Modular polyketide synthases are a class of biosynthetic machinery responsible for the production of a chemically diverse range of natural products, including essential medicines. The modular architecture of this system presents as inherently rearrangeable; however, manipulation of these pathways to produce novel polyketides often results in a reduced output or the pathway stalling entirely. One cause of such breakdown is the ability of ketosynthases (KS), an essential component of the biosynthetic modules, to proofread the substrate produced by preceding steps and reject unexpected chemistries. Here we explore the KS proofreading phenomenon using Graph Neural Networks (GNN) trained on KS AlphaFold structures, as a means to predict when proofreading occurs and how to mitigate it. Our GNN models can distinguish six out of ten β-carbon reduction state pairwise combinations (81–92% AUC). The α-methylation state on the polyketide substrate can also be partially predicted from the KS structure (79% AUC). To explore the utility of the GNNs as a co-pilot in mitigating proofreading, we ran a proof-of-concept experiment, where we show that the GNN models can be used to predict beneficial mutations to a KS receiving a non-native substrate. We also present a pilot study on acyltransferase domains, which identifies novel regions potentially associated with substrate specificity.

Streptomyces are key contributors to soil microbiome function, known for their biosynthetic diversity. While advances in -omics technologies have improved our understanding of microbiome composition and metabolic potential, the mechanisms underpinning interspecies interactions remain poorly resolved. Here, we investigate the molecular basis of interactions among four sympatric Streptomyces soil microbiome isolates, focusing on phenotypic, metabolomic and transcriptomic responses. Co-culture experiments revealed that one isolate, strain A, exhibited pronounced phenotypic changes when grown alongside each of the other three strains. Untargeted metabolomics and RNA-seq analyses showed that strain A undergoes distinct metabolic and transcriptional shifts depending on its partner, with the strongest response elicited by strain C. Despite all four strains possessing a conserved desferrioxamine biosynthetic gene cluster, only strain C constitutively produced desferrioxamine B (DFO-B), a hydroxamate siderophore, indicating a role for iron bioavailability in the interaction. Supplementation with DFO-B or iron mimicked the growth stimulation of strain A observed in co-culture with strain C, and CRISPR base editing of desD in strain C abolished both DFO production and the phenotypic induction of strain A. However, transcriptomic profiles of strain A varied significantly depending on the partner strain, with distinct sets of biosynthetic gene clusters and metabolic pathways activated in response to strains B and C, suggesting additional cues beyond DFO-B. In contrast, strain D did not elicit growth stimulation in its partners, and itself showed downregulation of amino acid and carbon metabolism when co-cultured with strain C. These findings indicate that Streptomyces interactions are not only mediated by siderophore piracy but also involve complex, strain-specific molecular responses. Our findings demonstrate that Streptomyces interactions are highly strain-specific and only partly mediated by siderophore piracy, with DFO-B acting as a potent interspecies cue. The divergent molecular responses to different partners suggest nuanced mechanisms of microbial sensing and competition. These insights advance our understanding of microbial crosstalk and highlight the ecological and evolutionary complexity of siderophore-mediated interactions. By integrating transcriptomics, metabolomics, and biochemical assays, we present a robust framework for dissecting microbial interactions, with implications for microbiome engineering and synthetic community design.

Alessandra Eustaquio, Eriko Takano and Tobias Gulder introduce the Natural Product Reports themed issue on microbial host and metabolic engineering for natural product discovery and production.

Summary: SyntenyQC is a data pre-processing tool for the construction of synteny plots. It supports genomic data collection, annotation and dereplication to facilitate (and in some cases fundamentally enable) the construction of informative synteny plots. Availability and implementation: SyntenyQC is a command line app developed using Python version 3.10 and tested using pytest. SyntenyQC is available on PyPI (https://pypi.org/project/SyntenyQC) under the MIT License, along with a detailed user tutorial. Package tests can be viewed at https://github.com/Tim-Kirkwood/SyntenyQC.

The thiopeptide GE2270A is a clinically relevant, ribosomally synthesised, and post-translationally modified peptide naturally produced by Planobispora rosea. Due to the genetically intractable nature of P. rosea, heterologous expression is considered a possible route to yield improvement. In this study, we focused on improving GE2270A production through heterologous expression of the biosynthetic gene cluster (BGC) in the model organism Streptomyces coelicolor M1146. A statistically significant yield improvement was obtained in the S. coelicolor system through the data-driven rational engineering of the BGC, including the introduction of additional copies of key biosynthetic and regulatory genes. However, despite our best efforts, the highest production level observed in the strains generated in this study is 12× lower than published titres achieved in the natural producer and 50× lower than published titres obtained using Nonomuraea ATCC 39727 as expression host. These results suggest that, while using the most genetically amenable strain as host can be the right choice when exploring different BGC designs, the choice of the most suitable host has a major effect on the achievable yield and should be carefully considered. The analysis of the multiomics data obtained in this study suggests an important role of PbtX in GE2270A biosynthesis and provides insights into the differences in production metabolic profiles between the different strains. One Sentence Summary: Data-driven rational engineering of Streptomyces coelicolor for heterologous production of the thiopeptide antibiotic GE2270A resulted in increased production but encountered unexpected challenges compared to production in the natural producer or the alternative host Nonomuraea ATCC 39727.

Advances in genome engineering and associated technologies have reinvigorated natural products research. Here we highlight the latest developments in the field across the discover-design-build-test-learn cycle of bioengineering, from recent progress in computational tools for AI-supported genome mining, enzyme and pathway engineering, and compound identification to novel host systems and new techniques for improving production levels, and place these trends in the context of responsible research and innovation, emphasizing the importance of anticipatory analysis at the early stages of process development.

The synergy between silk proteins and nanomaterials can lead to novel materials with improved mechanical and electrical properties. Designed peptides have been previously utilised in the functionalisation of two-dimensional material surfaces in a self-assembly manner, including graphite, to develop highly sensitive electrical biosensors. These studies have predominantly focused on functionalising the surfaces with peptides of less than several kDa in size. In this work, we assessed the capability of a ∼35 kDa synthetic spider silk protein to serve as a candidate biomolecular scaffold on a range of two-dimensional materials: graphite, molybdenum disulphide and boron nitride. The structural properties of the synthetic spider silk protein at the 2D material surface were characterised at the nanoscale for the first time using a multi-analysis approach incorporating atomic force microscopy and fluorescence microscopy, in addition to polarised Raman and tip-enhanced Raman spectroscopy. The synthetic spider silk protein was revealed to self-assemble into stable nanowire structures of monolayer thickness. Our findings have demonstrated the feasibility of functionalising 2D materials with spider silk-based proteins and will unlock new possibilities in the development of next-generation high-performance biosensing devices.

Cannabinoids are a therapeutically valuable class of secondary metabolites with a vast number of substituents. The native cannabinoid biosynthetic pathway of Cannabis sativa generates cannabigerolic acid, the common substrate to multiple cannabinoid synthases. The bioactive decarboxylated analogue of this compound, cannabigerol, represents an alternate gateway into the cannabinoid space either as a substrate to non-canonical cannabinoid synthase homologues or to synthetic chemical reactions. Herein, we describe the identification and repurposing of the aromatic prenyl transferase, AtaPT, which when coupled with native enzymes of C. sativa can form an E. coli production system for cannabigerolic acid in cell lysates, and cannabigerol in whole cells. Engineering of AtaPT, guided by structural analysis, was performed to enhance its kinetics towards cannabigerolic acid production for subsequent use in a proof-of-concept lysate system. For the first time, we show a synthetic biology platform for cannabigerol biosynthesis in E. coli cells by employing AtaPT under an optimised microbial system. Our results have therefore set the foundation for sustainable production of well researched and rarer cannabinoids in an E. coli chassis.