Natural products are crucial for discovering new bioactive molecules and understanding enzymatic chemistry. However, current genomics-driven approaches face challenges like redundancy and the need for laborious genetic manipulation. Halogenated natural products are particularly interesting due to their impact on small molecule properties and bioactivity, with approximately 25% of drugs containing at least one halogen atom. These halogens are introduced by halogenases, and thousands of halogenated natural products are known. Bioinformatic analysis of halogenase sequences and mass spectrometry (MS) have guided discovery efforts. However, MS-based methods struggle to identify natural products derived from cryptic halogenation, where an intermediate is transiently halogenated to activate it for further chemistry, resulting in unhalogenated final products with diverse structures. The alkyl halide-derivatizing enzymes responsible are diverse and poorly characterized, hindering bioinformatic pathway identification. This study aims to accelerate the discovery of natural products and pathways involving halogenation by leveraging our understanding of biosynthesis. Specifically, it investigates if removing halide anions (essential substrates for halogenases) from microbial growth media (‘halide depletion’) affects the production of halogenated natural products and metabolites derived from cryptic halogenation.

Several studies hinted at the influence of halide ion levels in bacterial growth media on the production of natural products derived from enzymatic halogenation. The cylindrocyclophanes, paracyclophane natural products formed via cryptic halogenation, were chosen as a model system. Their biosynthesis involves the dimerization of a halogenated monoalkylresorcinol precursor, catalyzed by the enzyme CylK. Cylindrocyclophane production occurs in both chloride- and bromide-containing media, suggesting flexibility in halide use. This system provided an ideal opportunity to validate the halide depletion workflow. The *Nostoc punctiforme* ATCC 29133 genome provided a target for exploring additional natural products. The organism contains a BGC encoding CyIC and CylK homologues, distinct from the cylindrocyclophane BGC, which was of particular interest because its orphan status suggested difficulties in detecting its product(s).

The halide depletion workflow involved growing microbial cultures in halide-replete and -deficient media with similar osmolarity, ionic strength, and nutrient concentrations. Metabolites were extracted using appropriate solvents (ethanol/water and *n*-heptane/ethyl acetate for *C. licheniforme*, chloroform/methanol for *N. punctiforme*) and analyzed by LC-MS and LC-MS/MS. Comparative metabolomics and molecular networking were used to identify differentially abundant features and structural relationships. For *N. punctiforme*, two different LC-MS methods were used—one for general analysis and a second optimized for lipidomics. To confirm the origin of identified metabolites, markerless deletion mutants of the *pks3* gene cluster (*ngl*) in *N. punctiforme* were created. Metabolite isolation and characterization involved normal-phase silica chromatography and reversed-phase HPLC with mass-guided fractionation. NMR spectroscopy was used for structural elucidation. Acid hydrolysis and permethylation were employed to identify the sugar moiety, and stable isotope feeding experiments with deuterated fatty acids were used to determine the chlorine substituent's position. Heterologous expression and purification of a truncated form of the enzyme NgIO (missing the C-terminal RTX domain) allowed for in vitro biochemical characterization. RNA sequencing of *N. punctiforme* cultures grown with and without chloride explored global gene expression changes resulting from halide depletion. Motility assays were conducted on the wild type and *ngl* mutants.

Halide depletion effectively identified new cylindrocyclophane derivatives in *C. licheniforme*, demonstrating the workflow's validity. The most significantly depleted feature was a cylindrocyclophane biosynthetic intermediate, with a relative fold change of 1525. Features corresponding to cylindrocyclophane F and the hydroxymonoalkylresorcinol monomer were also depleted. Molecular networking revealed 22 features clustering with cylindrocyclophane F, most showing decreased abundance upon halide depletion. Applying the workflow to *N. punctiforme* revealed several chlorinated metabolites significantly depleted in halide-deficient media. These metabolites, identified by LC-MS, were further characterized and identified as chlorinated glycolipids named nostochlorosides. The sugar headgroup was identified as gulose. Stable isotope feeding experiments located the chlorine substituent at the 12-position. Deletion mutants of the *ngl* gene cluster confirmed the cluster's role in nostochloroside biosynthesis. In vitro assays using a purified truncated version of the enzyme NgIO demonstrated an unprecedented enzymatic etherification reaction, resulting in the oligomerization of nostochloroside A. The oligomerization involved an ether linkage between the O6 of one nostochloroside A subunit and the C12 of another. RNA sequencing revealed that halide depletion altered the expression of genes involved in shinorine biosynthesis and various stress responses. Importantly, the *ngl* gene cluster expression was significantly increased in halide-depleted cultures, despite drastically reduced nostochloroside production.

The halide depletion workflow offers a simple, function-agnostic approach to natural product discovery. It avoids laborious genetic manipulations and bioactivity screens. The discovery of nostochlorosides, with their unique features (rare sugar gulose, odd-chain-length lipid tail, and unprecedented etherification reaction), highlights this approach's power. The nostochlorosides' biological roles are unclear, but genetic and phenotypic analyses suggest they are not essential for nitrogen fixation. The absence of an acyltransferase in the *ngl* cluster suggests that another enzyme elsewhere in the genome is responsible for generating the acyl derivatives of nostochloroside A. The observed polymerization of nostochloroside A by NgIO is distinct from previously reported CylK homologue activities, broadening the known reactivity of this enzyme family. The study emphasizes the importance of considering substrate availability in natural product biosynthesis and suggests that manipulating other biosynthetic building blocks could be a powerful strategy for future discovery.

This study demonstrated that manipulating the availability of a key biosynthetic precursor, halide anions, is a powerful strategy for discovering natural products. The halide depletion workflow successfully identified new derivatives of known natural products and a new family of compounds, the nostochlorosides, featuring unusual structural features. Future research could focus on characterizing the full *ngl* biosynthetic pathway, exploring the nostochlorosides' biological roles, and expanding this approach to other biosynthetic building blocks.

The halide depletion method may not be suitable for all organisms, particularly marine organisms requiring halides for growth. The physiological effects of halide depletion are not fully understood, and non-specific metabolic effects in other microbial phyla remain a possibility. The low yields of nostochlorosides hampered complete characterization of the oligomers. The study focused on halogenated compounds, limiting its applicability to other natural product classes.