Skip to main content
NCBI home page
RSC Chemical Biology logoLink to RSC Chemical Biology
. 2026 Oct 6. Online ahead of print. doi: 10.1039/d6cb00169f

Exploring RiPP-derived N-acyltransferases for genetically encoded lipopeptide engineering

Nina M Bösch a, Manon Coste a, Johannes A Rudmann Eckert a, Gabriel Cervera Arriaga a, Annina R Meyer a, Eléonore Moore a, Daniel Richter a, Jörn Piel a, Anna L Vagstad a,✉
PMCID: PMC13647313  PMID: 42851787

Abstract

Lipopeptides possess diverse bioactivities with widespread uses as antimicrobial drugs and biosurfactants but their biosynthetic engineering remains challenging. Classical lipopeptides (such as daptomycin or surfactin) are biosynthesized by modular, non-ribosomal megasynthetases. An alternative streamlined route to lipopeptides are the ribosomally synthesized and post-translationally modified peptides (RiPPs) derived from gene-encoded precursor proteins. We recently characterized a novel class of post-translational modification enzymes in RiPPs that catalyze fatty acylation of ornithine and lysine side chains. The resulting lipopeptides were termed selidamides. To assess the use of selidamide N-acyltransferases as biocatalysts, we probed the peptide substrate specificity of the model lauroyl-transferase KspN from kamptornamide biosynthesis. Nearest-neighbor, lysine-scan, and truncation mutants demonstrated promiscuity for the peptide substrate. As a proof-of-concept toward designer lipopeptides using RiPP biosynthetic principles, we produced sequences mimicking bioactive non-ribosomal lipopeptides and tested their conversion by maturases from the kamptornamide biosynthetic pathway. Ornithines, fatty acyl groups, and lanthionine macrocycles were installed in the mimics as a promising start toward the directed biosynthesis of custom lipopeptides.

The selidamide family of ribosomally synthesized peptide natural products are characterized by side-chain fatty acylations. Their biosynthetic enzymes were exploited to produce designer peptides resembling bioactive, non-ribosomal lipopeptides.graphic file with name d6cb00169f-ga.webp

Introduction

Non-proteinogenic moieties can improve the pharmacokinetic properties of peptide drugs.1 In particular, the addition of lipids to peptides has been shown to increase blood plasma stability, improve target specificity, and modulate toxicity.1 Fatty acylated peptides are widely used in pharmaceutical applications, either as naturally occurring lipopeptides or by the chemical introduction of acyl chains.2 A striking example is the anti-diabetes drug detemir, an insulin variant bearing a myristoyl (C14:0) group on the ε-amine of chain B-Lys29, which prolongs the duration of action approximately 5-fold.3 The identity and location of the fatty acyl group can drastically change the potency and toxicity of lipopeptides.4,5 Thus, development of biocatalysts for the site-specific introduction of lipid groups into peptides to complement synthetic tools is a desirable resource for drug development.

Most microbial lipopeptides characterized to date are non-ribosomal peptide (NRP) natural products produced in an assembly line fashion by large, multi-modular protein complexes called non-ribosomal peptide synthetases (NRPSs).6 Modules can introduce non-canonical amino acids to expand the chemical scope beyond the canonical 20 proteinogenic amino acids of ribosomally synthesized polypeptides.7 Intermediates are tethered to the enzyme, and the mature peptide is most commonly released by hydrolysis or macrocyclization. Lipids are generally introduced by the initiation module as starter units but may also be incorporated as rare late-stage tailoring modifications to generate lipopeptide NRPs.8,9 As of March 2026, 38 families of naturally occurring NRP lipopeptides comprising 318 members are annotated in the Norine database.10 Although a wide variety of pharmaceutically relevant activities have been demonstrated for NRP lipopeptides, their clinical use is currently limited to the antibacterial agents daptomycin and polymyxin B against Gram-positive and Gram-negative species, respectively, and the antifungal echinocandins.11 The production and purification of lipopeptides from their natural hosts can be challenging as the fatty acyl AMP-ligase and starter condensation domain enzymes responsible for attaching the fatty acids tend to be promiscuous, leading to a mixture of difficult-to-separate congeners.12,13 Recent breakthroughs in NRPS engineering involve the recombination of multidomain exchange units from different pathways, allowing the custom generation of short lipopeptides as differentially acylated mixtures.14 The size of NRPs appears to be evolutionary constrained and rarely exceeds 15 residues.

A distinct route to lipopeptide natural products is offered by ribosomally synthesized and post-translationally modified peptide (RiPP) pathways.15,16 RiPPs are biosynthesized from a gene-encoded precursor peptide, consisting in most cases of N-terminal leader and C-terminal core parts. Post-translational modifications (PTMs) are introduced by maturase enzymes into the core region. The maturases are generally encoded near the precursor protein gene within a co-regulated biosynthetic gene cluster (BGC), which also tend to contain any genes responsible for pathway regulation, leader removal, and export of the mature peptide natural product derived from the precursor core. RiPP pathways have garnered attention for peptide engineering applications owing to their compact BGCs and easily mutable gene-encoded precursor proteins.17 Characterized RiPPs exhibit a wider range of peptide lengths relative to NRPs, with core peptides numbering from only a few amino acids (e.g., pyrroloquinoline quinone cofactors18 and atropopeptides19) to 49 or even more residues (e.g., polytheonamides20 and some spliceotides21).

We previously described a family of RiPP lipopeptides named selidamides, with the class-defining feature being a side-chain fatty-acyl modification of a lysine or arginine-derived ornithine residue (abbreviated as Orn or O). The acylation is catalyzed by a GCN5-related N-acetyltransferase (GNAT) superfamily enzyme (Fig. 1A).22 Three selidamide BGCs were characterized by heterologous expression in Escherichia coli: ksp from Kamptonema sp. PCC 6506, pha from Pseudophaeobacter arcticus DSM 23566, and npu from Nostoc punctiforme PCC 73102, to reveal distinct fatty acyl-modifications of the side-chain amine of the most N-terminal lysine or ornithine residue within the respective precursor protein cores (Fig. 1B–D). These structures also contain one or more thioether macrocycles introduced by a class II lanthionine synthetase (Fig. S1). In RiPP nomenclature, the core residues are numbered up from one corresponding to the residues in the mature peptide product. In the case of ornithine as the acyl acceptor, arginine residues are first hydrolyzed by a peptide arginase from the pathway to reveal the side-chain amine (Fig. S2).23 For the ksp and pha clusters, the final products were characterized: kamptornamide and phaeornamide bearing dodecanoyl (lauroyl, C12:0) and (S)-3-hydroxydecanoyl [(S)-3-OH-C10:0] fatty acylations of ornithine, respectively. For the npu cluster, only partial characterization of the two putative final products was possible due to low yields: nostolysamide A and B are modified at the NpuA core Lys1 by hexadecenoyl (C16:1) and hydroxyhexadecenoyl (OH-C16:1) fatty acids, respectively. In an independent study, the van der Donk group recently determined the connectivity of the four (methyl)lanthionines facilitated by use of a SUMO (Small Ubiquitin-like Modifier) tag fusion of NpuA to improve soluble production of the unacylated lanthipeptide (termed nostolysamide C), which exhibits antifungal activity.24 They additionally demonstrated in vitro activity for the N-acyltransferase NpuN with various chain lengths of acyl-coenzyme A substrates. Transfer of unsaturated C2:0, C12:0, C14:0 and C16:0 fatty acids and monounsaturated C18:1Δ9 (oleoyl) to Lys1 of nostolysamide C were detected, suggesting that constraints on in vivo substrate specificity can be overcome with controlled in vitro reaction conditions. In E. coli, the distinct N-acylations catalyzed by the pathway GNATs were selective toward a single chain length of endogenous fatty acids, which contrasts with the generally low selectivity of lipoinitiation modules in NRPSs.12,13 Other RiPP lipopeptides modified by GNATs include the lipolanthines,25 goadvionins,26 lipoavitides,27 and aphotorhaptins,28 which are modified at their N-termini with fatty acyl groups following proteolytic removal of the leader. Terpene–RiPP hybrids alkylated by linear and cyclic isoprenoids are also known (e.g., cyanobactins,29 ComX,30 and steromaze,31 among others) as a distinct type of lipopeptides modified by prenyltransferases.

Fig. 1. Previously characterized selidamide RiPPs. (A) Generic acyl transfer reaction catalyzed by the N-acyltransferases. *Endogenous acyl-coenzyme A thioesters are hypothesized to be the relevant acyl donors in E. coli. BGCs and structures of selidamide natural products (B) kamptornamide, (C) phaeornamide, and (D) nostolysamides A and B. Precursor protein core sequences are numbered up from 1 following the Gly–Gly leader cleavage motif. See ref. 24 for the core lanthipeptide structure of the nostolysamides. #The displayed acyl structures are based on established fatty acid biosynthesis and degradation pathways in E. coli (i.e., Δ9-cis double bond of palmitoleic acid and 3-hydroxyl moieties).

Fig. 1

Open in a new tab

In RiPPs, the separation of peptide substrate recognition elements (i.e., the leader) from the site of catalysis (i.e., the core) generally imparts promiscuity of PTM enzymes toward alternative core peptide sequences.32 Thus, we hypothesized that the N-acyltransferases could be developed as biocatalytic tools to site-specifically acylate diverse peptides fused to a compatible precursor leader and, in combination with other modifying enzymes, to mimic NRP lipopeptide structures. PTMs are already reported from RiPPs that install structural features known from bioactive NRPs,33 such as arginine-to-ornithine hydrolysis,23l- to d-amino acid epimerizations,34,35N-methylations,36 and (macro)cyclizations.37 This approach has had some early successes. For example, the NRP lipopeptide brevicidine was mimicked using an engineered RiPP system in which the native macrolactone was approximated by a methyllanthionine and the lipidated N-terminus was substituted with a series of hydrophobic amino acids to produce designer peptides with antibacterial activity against Xanthomonas campestris.38 Furthermore, late-stage side-chain lipidations using chemical derivatization, for example, via the Cysteine Lipidation on a Peptide or Amino acid (CLipPA) technology, has produced bioactive peptides with improved antibacterial functions over the non-lipidated variants.39

In this study, three selidamide pathways were probed as model systems to expand the biocatalytic toolbox for RiPP-based production of lipopeptides. For the most promising candidate—the kamptornamide C12-acyltransferase KspN—mutant series and designer peptide sequences mimicking bioactive NRPs were incorporated into the KspA precursor protein to show a generally broad peptide substrate sequence tolerance for this enzyme. Future work will focus on characterizing additional selidamide GNAT family members with distinct natural fatty acyl and core-peptide substrate preferences for biocatalytic applications, as well as scaled up productions for bioactivity testing.

Experimental

Materials

Chemicals were used without further purification. Antibiotics came from Applichem: Spectinomycin (spt), kanamycin (kan), and chloramphenicol (chl) antibiotics were used at 25 µg mL−1. Ampicillin (amp) was used at 100 µg mL−1. The media components Bacto Tryptone and Bacto Yeast Extract came from BD Biosciences and Agar–Agar for solid medium came from Serva. Isopropyl-β-d-1-thiogalactopyranoside (IPTG) came from Biosynth AG. Other chemicals were purchased from Sigma Aldrich, unless otherwise stated. Sources of experiment-specific materials and reagents are given in the relevant sections below.

General information on cloning of expression vectors

The plasmids, primers, and synthetic gene sequences used in this study are listed in Tables S1–S3, respectively. E. coli DH5α grown in Lysogeny Broth medium (LB: 10 g L−1 tryptone, 5 g L−1 yeast extract, 10 g L−1 NaCl) was used for cloning and plasmid propagation. For polymerase chain reaction (PCR) amplifications, Phusion High-Fidelity DNA Polymerase (Thermo Scientific) was used with addition of 3% dimethyl-sulfoxide (DMSO) to the reactions. Restriction enzymes and T4 DNA ligase were purchased from New England Biolabs (NEB) and FastAP Thermosensitive Alkaline Phosphatase from Thermo Scientific. NucleoSpin™ Plasmid kit and NucleoSpin™ Gel and PCR Clean-up kit (Macherey-Nagel) were used for plasmid extraction and PCR reaction clean-ups, respectively. Kits and reagents were used as recommended by the manufacturer.

For site-directed mutagenesis of precursor genes, full-vector PCR amplification or Gibson assembly methods were used, as detailed in the SI. For mutagenesis by full-plasmid amplifications, primer sets were generated using the QuikChange Primer Design tool (https://www.agilent.com/store/primerDesignProgram.jsp). Gibson assembly master mix was self-made.40

All plasmids were verified by Sanger or Oxford Nanopore sequencing (Microsynth AG). Strains were cryopreserved in LB supplemented with 25% glycerol at −80 °C. The naming convention for hybrid precursor protein constructs includes ‘L’ for the leader and ‘C’ for the core parts as either subscripts or in parentheses. Plasmids used as templates for kspA, phaA, and npuA in pACYCDuet-1 were previously published.22 pACYCDuet-based clones were selected on LB + chl.

Protein expression and Ni2+-affinity column chromatography purification

Proteins were (co)produced from DUET expression vectors (Novagen). The appropriate expression plasmids for the precursor protein and maturase enzymes were (co)transformed into chemically competent E. coli BL21(DE3) according to the experimental design. Expression cultures were in Terrific Broth medium (TB: 12 g L−1 tryptone, 24 g L−1 yeast extract, 4 mL L−1 glycerol, 17 mM KH2PO4, 72 mM K2HPO4) inoculated with saturated pre-cultures at a 1 : 100 dilution and grown at 37 °C and 250 rpm shaking until reaching OD600 1.5–2.0. Cultures were cooled at 4 °C for 30 min, induced with 1 mM IPTG, and shaken at 180 rpm at 24 °C for 3 days. Cells were harvested by centrifugation (30 min at 5k × g), frozen, and stored at −20 °C until use.

KspA-R1K truncation mutants and His6-bdSUMO-tagged KspN for the in vitro reactions were similarly produced, but cultures were only expressed for ∼16–20 h overnight prior to harvest.

Cell pellet resuspension, lysis by sonication, clarification by centrifugation, and His6-tagged protein purification by standard Ni2+-affinity column chromatography using a benchtop step gradient over Ni-NTA resin (Protino, Macherey-Nagel) were as previously described.41 Purification buffers contained 50 mM Na/PO4 pH 8.0, 300 mM NaCl, 10% glycerol, and varying amounts of imidazole. Proteins were bound and washed with 20 mM imidazole buffer (Ni-B-20) and eluted with 250 mM imidazole buffer (Ni-B-250). Fractions were analyzed by Laemmli SDS-PAGE with 12% acrylamide resolving gel. PageRuler Prestained Protein Ladder (Thermo Scientific) was used as a protein size reference. Protein concentrations were determined by ROTI®Nanoquant assay (Carl ROTH) using Bovine Serum Albumin Fraction V (Sigma-Aldrich) as the standard.

Small-scale cultures for triplicate co-expression analyses were conducted in 96-deep-well plates (Sigma-Aldrich) in 1.5 mL TB supplemented with antibiotics for relevant plasmid selection. Cultures were pelleted, resuspended in 0.5 mL Denaturing Phosphate Buffer (DPB: 100 mM Tris HCl pH 8.0, 300 mM NaCl, 20 mM imidazole, 8 M urea, 10% glycerol in ddH2O) supplemented with 1× FastBreak Cell Lysis Reagent (Promega). Purification by Ni2+-affinity column chromatography was conducted similarly to large-scale expressions. Eluate was concentrated by centrifugation (11k × g) on 0.5 mL Amicon Ultra Centrifugal Filters (3 kDa MWCO, Merck Millipore) for mass spectrometry analyses, as needed.

NRP mimics from the final designs (Fig. 5), were additionally produced at 150 mL scale for structural characterization by HPLC-HRMS2. Two additional cultures at 50 mL scale were completed to give triplicate data for the NRP mimics kahalalide and brevicidine, where small-scale expressions gave low yields.

Fig. 5. Production of NRP-like peptides modified by Ksp RiPP biosynthetic enzymes. (A) Heterologous expression and analytical workflow. Structures of targeted NRPs (left) and the NRP-like product profiles (right) from NRP-mimic precursor protein co-produced with KspMRN for (B) enramycin A, (C) kahalalide F, and (D) brevicidine; KspMN for (E) lokisin and (F) kurstakin; KspRN for (G) amphibactin D and (H) aquachelin D. Modifications are indicated according to the lower legend. The isobaric dehydrated Ser/Thr (i.e., dehydroalanine/dehydrobutyrine) and (methyl)lanthionines are tentatively assigned based on HPLC-HRMS22 evidence for the most abundant product: a lack of MS2 peptide fragments in cyclized regions and no derivatization of cysteine by IAA suggested intact lanthionines. Error bars indicate one standard deviation from biological triplicates. Peptides were detected as the GluC fragments and additionally contain KspA residues −5 to −1 (AIAGG) at the N-terminus of the displayed core sequences.

Fig. 5

Open in a new tab

Proteolytic digest of purified precursor peptides

Small-scale proteolytic digests were performed in a reaction volume of 20 µL at a final concentration of 1 µg µL−1 peptide. Glutamyl endopeptidase (GluC, from NEB) digest was performed at 1 : 50 w : w ratio in the provided buffer supplemented with 25 µM tris(2-carboxyethyl)phosphine (TCEP). Digests were incubated at 37 °C for 18 h. Digested samples were diluted 10-fold in 0.1% (aq) formic acid, clarified by centrifugation (21 k×g, 15 min), and 5 µL injections were analyzed by HPLC-HRMS.

Iodoacetamide (IAA) treatment of free thiol groups

IAA derivatization of purified precursor proteins was performed as described previously.41 Aliquots (500 µL) of the precursor proteins were concentrated to ∼100 µL, diluted to 500 µL with reaction buffer (8 M urea, 50 mM Tris–HCl pH 8.0, 5 mM dithiothreitol [DTT]), and reduced at 37 °C while shaking for 1 h. Freshly prepared 0.1 M IAA (80 µL) was added, and the reaction mixture was incubated at room temperature in the dark for 30 min. Excess IAA was quenched by addition of 100 mM DTT (30 µL) and the reaction was further incubated at room temperature in the dark for 15 min. The derivatized peptides were subjected to GluC protease digest as described above.

In vitro KspA truncation reactions

KspA-R1K truncation variants were tested as substrates for His6-bdSUMO-KspN in vitro. Proteins were purified by Ni2+-affinity column chromatography and buffer exchanged to Reaction Buffer (50 mM HEPES pH 8, 150 mM NaCl, 10% glycerol). Reactions contained 50 µM precursor peptide, 200 µM lauroyl-CoA (Sigma Aldrich), and 5 µM KspN. Triplicate reactions were incubated in 96-well plates at room temperature for 5 h, 1 µL of each reaction was quenched with 100 µL 0.1% formic acid and analyzed by intact-protein HPLC-HRMS.

Peptide analysis by HPLC-HRMS

High-performance liquid chromatography high-resolution mass spectrometry (HPLC-HRMS) was performed on a Dionex Ultimate 3000 UHPLC coupled to a Q Exactive Hybrid Quadrupole-Orbitrap Mass Spectrometer (Thermo Scientific). GluC-digested peptides were separated with a flowrate of 0.7 mL min−1 over a Phenomenex Kinetex C18 column (2.6 µm, 100 Å, 150 × 4.6 mm) heated to 50 °C and detected by heated electrospray ionization (HESI) in positive ion mode. Peptide specific gradient methods are described in the SI. Water and acetonitrile, both supplemented with 0.1% formic acid, were used as solvents A and B, respectively.

Intact-protein HPLC-HRMS was similarly performed with a flowrate of 1 mL min−1 over a Phenomenex C4 Aeris WIDEPORE column (2.6 µm, 150 × 4.6 mm). HPLC-HRMS settings were as described previously unless otherwise stated.21 Parallel Reaction Monitoring using manually annotated mass-to-charge ratio inclusion lists were used for tandem MS experiments to localize modifications.

Relative amounts of the core peptide substrate and product for the KspA point mutants and all rounds of NRP mimics were approximated by peak area of the extracted ion chromatograms (EICs) for the calculated ions with 10 ppm error. Area under the curve (AUC) was determined in Thermo Xcaliber QualBrowser (version 4.1.31.9) or Freestyle (version 1.8) for the most abundant charge state(s).

Relative amounts of the core peptide substrate and product for the KspA-R1K truncation mutants and NRP final designs were analyzed by a custom script documented and available at https://gitlab.ethz.ch/piel_code_public/eckert_johannes/ms_dataanalysis. A further description of the script and output data are available in the SI.

MS2 data was analyzed by a locally installed copy of Metaboseek 0.9.9 downloaded from https://github.com/mjhelf/Metaboseek.42

Results and discussion

Cross-reactivity of N-acyltransferases toward non-native precursors

A series of co-expression experiments were devised to test the application of N-acyltransferases as biocatalysts to generate custom side-chain fatty-acylated peptides. To assess lipopeptide production, our general workflow was as follows: E. coli BL21(DE3) was used as the heterologous host for co-production of His6-tagged precursor proteins with untagged modification enzymes using T7-promoter-based DUET expression vectors. Precursors were purified by Ni2+-affinity column chromatography. Proteins were optionally treated with iodoacetamide (IAA) to modify free cysteine thiols with carboxyamidomethyl groups to prevent disulfide bond formations. Precursor fragments containing the core peptides were released by protease cleavage and analyzed by HPLC-HRMS. The corresponding mass-to-charge (m/z) features associated with the core peptides were identified and analyzed by targeted MS2 to localize the modification sites.

In initial experiments, we assessed the cross-reactivity of previously characterized GNATs towards non-native precursors. The acyltransferases of the ksp, pha, and npu BGCs (designated with postfix N, e.g., KspN) were co-produced with the precursor peptides of the non-native pathways (designated with postfix A, e.g., KspA). The ksp and pha encoded arginases (designated with postfix R, e.g., KspR) were additionally co-produced with the KspA and PhaA precursors to install the ornithine acyl-acceptor sites, whereas NpuA natively has a lysine acceptor site. A lysine installed in the KspA (Arg1Lys) or PhaA (Arg2Lys) cores were previously shown to be highly modified by the respective KspN and PhaN acyltransferases, suggesting these acyltransferases would be capable of modifying a lysine in the NpuA core.22 KspA and PhaA are NHLP-type precursors (also known as ‘proteusins’20) based on leader homology to nitrile hydratase α-subunits, whereas the NpuA precursor has a Nif11-type leader with homology to the nitrogen-fixation protein 11 of unknown function.43 Both NHLP and Nif11 precursors harbor a conserved bacteriocin type Gly–Gly cleavage motif at the leader–core junction. The three precursors were compared by multiple sequence alignment (Fig. S3), revealing that the KspA leader has 48.9% and 27.0% sequence similarity to the PhaA and NpuA leaders, respectively.44

For all mismatched precursor–GNAT pairs, only unacylated precursor protein core peptides were detected (Fig. S4–S6). To test if the inactivity of the GNAT toward the non-native precursors was due to leader or core sequence incompatibilities, we next generated a series of chimeric leader–core precursors. NpuA was generally produced in low soluble yields from E. coli expressions, a problem already noted in our earlier study;22 therefore, this leader was excluded from the core swap experiments. Based on the consensus that the leader peptide is the recognition site for maturases,45 we hypothesized that GNATs would modify non-native cores fused to the native leader. For these hybrid constructs ‘L’ and ‘C’ subscripts were used to indicate the source protein of the leader and core, respectively. For example, the KspA leader peptide fused to the PhaA core peptide is designated ‘KspAL-PhaAC’. Hybrid precursor peptides with KspA and PhaA leaders fused to KspA, PhaA, and NpuA cores were co-produced with the respective KspN and PhaN acyltransferases. The arginases were selectively used to install ornithine acyl acceptors in the KspA and PhaA cores.

Gratifyingly, the detected mass difference between the unmodified and modified core peptide fragments confirmed that the transferred fatty acid was selected specifically by the acyltransferase–leader pair irrelevant of the core peptide sequence, consistent with the previously characterized transfer of (S)-3-OH-C10:0 by PhaN and C12:0 by KspN (Fig. S7–S15). KspN showed high efficiency in modifying the hybrid precursors KspAL-PhaAC and KspAL-NpuAC, as only minimal amounts of the unmodified core peptides were detected for either variant (Fig. 2 and Figures S8–S9). For PhaN, only moderate modification was detected for both precursors PhaAL-KspAC (Fig. S13) and PhaAL-NpuAC (Fig. S14). These results are consistent with our previous observation that PhaN prefers a lanthionine-cyclized native core peptide over the unmodified core peptide.22 As the KspN acyltransferase appeared to be the most efficient of the enzymes tested, it was selected as the model system for further peptide lipidation experiments.

Fig. 2. The KspN acyltransferase modifies noncognate core peptides fused to the KspA leader. Data for the wild-type KspA positive control (left), PhaA core (middle) and NpuA core (right) are shown. (A) Sequence of the modified core peptide products of KspA-hybrid co-productions. The core peptide sequence is shown in black text with arginine-to-ornithine (O) modifications in green (loss of CH2N2; ΔM = −42.0218 Da per modification), side-chain acylation in pink (gain of C12H22O; ΔM = +182.1671 Da), iodoacetamide alkylation underlined (gain of C2H3NO; ΔM = +57.0215 per modified cysteine), and the remaining leader residues from the GluC protease digest in gray. The site of acylation was determined by HPLC-HRMS2, see Fig. S10–S11. (B) Select HPLC-HRMS extracted ion chromatograms are shown for the unmodified (black) and fully modified (pink) core peptide fragments. (C) Mass spectra of the unmodified (top) and fully modified (bottom) core peptides.

Fig. 2

Open in a new tab

KspA mutants probe the substrate promiscuity of KspN

For each substrate, a single fatty acid was transferred by KspN onto the most N-terminal ornithine or lysine core residue despite the availability of additional acceptor residues for the KspA and NpuA core peptides. This led us to hypothesize that the core is processed in an N- to C-terminal direction with the first encountered acceptor residue as the preferred site of modification. To test if acceptor residues at more C-terminal locations could be lipidated, a lysine scan was conducted along the KspA core sequence. Use of lysine rather than an ornithine acceptor residue simplified the co-expressions to two genes (kspA and kspN) and one potential modification site, allowing the kspR arginase-encoding gene to be excluded. The co-expression experiments were performed as biological triplicates in a 96-well-plate format. Apparent percent conversion was calculated by comparison of the area under the curve (AUC) for the unmodified and C12-acyl-modified GluC fragment peptides assessed by HPLC-HRMS. Lipid modifications are known to reduce the signal intensity of peptide ions detected by electrospray ionization (ESI)-MS;46 thus, the AUC method likely underestimates the amount of acylated product and is a conservative estimate of KspN efficiency.

The lysine scan showed varying modification efficiencies regarding the core position of the lysine with a general trend that more N-terminal residues were preferred acylation sites (Fig. 3A and Fig. S16–S36, Table S4). The Asn12Lys mutant appears to be an exception, with relatively more acylation relative to the neighboring sites, suggesting other factors are at play. Remarkably, acylation of a C-terminal lysine was observed, indicating that the acceptor residue does not need to be at an internal position and can contain a free carboxylate. Only the natively modified core position at Lys1 was highly modified by KspN; all other sites had <50% average apparent conversion.

Fig. 3. Acylation of KspA point mutants by KspN. (A) Lysine scan to examine the position dependence of the acceptor residue. R(1–3)K is the R1K–R2K–R3K triple mutant. (B) G-1X and (C) R2X nearest-neighbor saturation mutant libraries for the KspA-R1K precursor where ‘X’ is the other 19 amino acids. Mutants indicated on the x-axis are organized by side-chain type from left to right: linear–hydrophobic, aromatic, polar–uncharged, positive charge, negative charge, and special cases. Error bars indicate one standard deviation from biological triplicates. The R1K (part A), G-1 (part B), and R2 (part C) data colored in blue are from the identical KspA-R1K mutant substrate with native Gly-1 and Arg2 residues, which served as the positive control for each mutant series. Notes: *The G-1K mutant is acylated on Lys-1 rather than Lys1.22.

Fig. 3

Open in a new tab

Next, the influence of residues neighboring the modification site was tested. Using the KspA-R1K mutant precursor peptide, Gly-1 and Arg2 were individually exchanged for all 19 other canonical amino acids by saturation mutagenesis to produce two libraries termed ‘G-1X’ and ‘R2X’.47 In almost all co-expression experiments, modification of Lys1 by KspN was observed, albeit with variable efficiencies. Especially for the −1 position, only partial modification was detected in most cases (Fig. 3B and Fig. S37–S55, Table S5). The lowest conversions were observed for aromatic and bulky residues (especially for isoleucine, phenylalanine, tyrosine, and tryptophan), suggesting that steric hindrance may contribute to reduced KspN activity for these KspA side-chain variants. Similar low conversion was detected for arginine and conformationally restricted proline. The low conversion of arginine could possibly be explained by competitive binding of the more N-terminal, positively charged Arg-1 over Lys1, since KspN acts in an N- to C-terminal direction. Residues that were tolerated with the highest conversion tended to be small (e.g., alanine, serine, and cysteine). The role of side-chain size was also apparent in the reduced modification of glutamine and glutamate containing an extra methylene relative to the shorter asparagine and aspartate substrates, respectively. With this data, however, we cannot rule out that mutations in the highly conserved Gly–Gly motif at the leader–core junction disrupted substrate recognition in a more general way, for example, through imposing conformational constraints that prevented access to Lys1 in the KspN active site.

In contrast, all Arg2 substitutions were tolerated well with high conversions (Fig. 3C and Fig. S56–S73, Table S6). The lowest average apparent conversion was observed for the Glu2 variant (83.3%). All other residues were modified with very high efficiency (>96%), indicating no substantial interference of the residue at the n + 1 position following the modification site.

With the final mutant series, we probed the ability of KspN to accept truncated core sequences ranging from full-length (20 amino acids) to a single lysine residue. The KspA-R1K truncation mutants were individually produced and tested as in vitro substrates of KspN (purified as a His6-bdSUMO48 solubility tag fusion) together with lauroyl-CoA as the acyl donor substrate in a single timepoint assay. The product distribution was measured by intact protein HPLC-HRMS for triplicate reactions (Fig. 4 and Fig. S74–S81). The data show acylation of all substrates. Precursors truncated after residues 10 and 15 had higher apparent modification relative to the full-length KspA-R1K core, whereas cores of 2–5 residues were acylated at similar efficiencies. Although the acylation rate decreased for the single-lysine core, conversion of this construct confirms that KspN can modify lysine residues with a free carboxyl terminus, as also observed for the KspA-N20K mutant (Fig. 3A).

Fig. 4. KspN modifies truncated KspA-R1K variants. In vitro reactions were performed in triplicate with lauroyl-CoA as the acyl donor and R1K truncation variants as the acyl acceptor. The acylated product, as a percent of the total precursor protein, was detected by intact protein HLPC-LCMS at the 5 h timepoint. Error bars indicate one standard deviation.

Fig. 4

Open in a new tab

Taken together, the core-swap and point-mutant experiments establish that (i) non-native core peptides can be modified by acyltransferases when fused to the cognate leader peptide; (ii) the most N-terminal lysine or ornithine acceptor site is modified by KspN and N-terminal core positions are favored, but other positions are accepted with lower efficiencies; (iii) neighboring residues influence the acyltransferase activity of KspN at the n − 1 position but have only a minimal effect at the n + 1 site; and (iv) KspN modifies core peptides of variable lengths. These results underpin the use of custom core-replacement sequences to rationally produce designer lipopeptides.

Generation of NRP lipopeptide mimics

As a proof-of-concept, we tested if the Ksp biosynthetic enzymes could modify engineered core peptides to emulate non-ribosomal lipopeptides. In a preliminary round of experiments, we generated a series of KspA hybrid precursor peptides with the KspA leader fused to core peptides that mimic the underlying peptide sequence of NRPs with diverse bioactivities: daptomycin,4 echinocandin B,49 enramycin A (also known as enduracidin),50 kahalalide F,51 lokisin,52,53 and surfactin A.54 These NRPs are naturally acylated at their N-termini with medium-to-long chain-length fatty acids and contain macrolactam or macrolactone heterocycles. A lysine residue was placed at core position 1 as the acyl acceptor to mimic the N-terminal fatty-acyl groups. As a precaution, a glycine spacer was inserted in the daptomycin and enramycin mimicking cores to ensure that the neighboring bulky (tryptophan) or negatively charged (aspartate) residue would not interfere with acylation. In hindsight, this was not necessary since subsequent results for the R2X mutant series showed that all amino acid substitutions were highly tolerated at the n + 1 position (Fig. 3C). Naturally occurring ornithines, which are common monomers in NRPs, were substituted with arginine at the appropriate sites for conversion by the KspR arginase. KspR was previously shown to broadly tolerate alternative substrates.23

To mimic the natural macrocycles in the selected NRP structures, we designed a second series of substrates that could be macrocyclized by the pathway lanthionine synthetase KspM at non-native positions. KspM naturally installs a nine-residue methyllanthionine between KspA Thr11 and Cys19. The substrate scope of KspM had not previously been tested beyond the native KspA core, but other type II lanthionine synthetases exhibit high regio and sequence promiscuity.55,56 Furthermore, the natural macrolactone ring of the antibacterial NRP brevicidine has previously been mimicked by the thioether of a methyllanthionine installed by the two-component type I lanthionine synthetase NisBC.38 We introduced a C-terminal cysteine extension and an internal serine/threonine as the KspM substrate residues to allow macrocycles to be installed in the designer cores. For surfactin, a glycine was inserted after the introduced threonine to mimic the C1–C3 of the 3-hydroxy-fatty acyl starter unit, which is included in the native macrolactone.

Similar to our earlier workflow, the engineered precursor peptides were co-produced with selected sets of Ksp maturases and purified by Ni2+-affinity column chromatography (Fig. 5A). Product profiles of GluC-released core peptide fragments were analyzed by HPLC-HRMS2. Data for the co-production of the NRP-like precursor variants with Ksp(R)N (Table S7 and Fig. S82–S87), KspM (Table S8 and Fig. S88–S93), and KspM(R)N (Table S9 and Fig. S94–S100) are shown. KspR was optionally included for NRPs containing ornithine residues (i.e., daptomycin, enramycin, and kahalalide). Alkylation of uncyclized cysteines and MS2 fragmentation data instructed our assignment of linear and (methyl)lanthionine cyclized products (Fig. S101–S109).

For all co-expressions that included KspN, we detected the C12 fatty-acylated peptides. For the KspM(R)N co-production data, four of the tested substrates (daptomycin, echinocandin, enramycin, and kahalalide) were >95% modified (Fig. S94). Only the lokisin and surfactin mimics had relatively high amounts of unacylated peptide remaining, accounting for ∼60% and 35% of the detected core peptides, respectively.

Modification of the NRP-like peptides by the lanthionine synthetase KspM and the arginase KspR was more variable. For the KspM co-expressions, only up to two dehydrations were observed in the NRP-like cores, even when more than two serines and threonines were present (Table S8). The surfactin-like core was not modified at all by KspM and only small amounts of the single-dehydration product (∼9%) was observed for daptomycin. The other cores ranged from 20–99% of mixed dehydrated/cyclized products. Installation of ornithine by KspR was generally efficient and ornithine containing products were the predominant species for the relevant daptomycin, enramycin, and kahalalide mimics (Tables S7 and S9), consistent with the general promiscuity observed by KspR toward non-native substrates in our previous study.23 The limited modification efficiency by KspM might be caused by the inherent primary and secondary structures of the non-canonical core peptides prohibiting efficient modification, which has been shown to affect cyclizations by the model type II lanthionine synthetase ProcM from cyanobactin biosynthesis.57 Whereas the lysine acylation site is located at its native core position 1, the arginine, serine, threonine, and cysteine residues are located at unnatural positions and sequence contexts. It is especially remarkable that KspM installs (methyl)lanthionines using a cysteine with a free carboxylate at the C-terminus instead of an internal cysteine, like in the native KspA core. In all cases where lanthionine formation by KspM was detected (enramycin, kahalalide, and lokisin), MS2 data suggested that the smallest ring was formed between the serine/threonine residue closest to the C-terminal cysteine, leading to smaller cyclic structures relative to the desired NRP targets. This observation stands in contrast to the native kamptornamide system where Thr11 (rather than the more C-terminal Thr13) is dehydrated and coupled with the penultimate Cys19 to give a 9-residue macrocycle. IAA derivatization of the remaining free cysteines differentiated the lanthionine-cyclized and linear-dehydrated products of KspM. The echinocandin, enramycin, kahalalide, and lokisin mimics predominantly gave a monocyclized species, as the linear dehydrated species were only minor products of the IAA-derivatized treatment.

Based on the lessons learned from these preliminary results, we generated a third series of NRP-mimics (Fig. 5B–H and Table S9, Fig. S110–S124). The redesigned enramycin, kahalalide, and lokisin constructs featured threonine and cysteine residues installed with spacing to more closely resemble the natural sizes of the NRP macrocycles, where cysteine replaced the last residue in the core. Brevicidine58 and kurstakin59,60 were added as targets because their four-residue macrocycles lack additional serine or threonine residues that could give rise to competing KspM-installed lanthionine regioisomers. In addition, two linear lipopeptide targets containing several modified ornithine residues were selected, the siderophores amphibactin and aquachelin.61

These constructs were heterologously expressed with the relevant PTM enzyme genes from the kamptornamide biosynthetic pathways: kspMRN for enramycin, kahalalide, and brevicidine, kspMN for lokisin and kurstakin lacking ornithines, or kspRN for the linear amphibactin and aquachelin mimics. Products were characterized from 150 mL cultures to ensure sufficient material for HPLC-HRMS2 and IAA cysteine-derivatization assays (Fig. S117–S124). Small-scale (1.5 mL) 96-well-plate expressions followed in triplicate to quantify the product profiles (Fig. 5B, C and E–H). For the kahalalide and brevicidine mimics, small-scale expressions led to significantly less modification, and the product profiles were averaged from duplicate 50 mL cultures and the 150 mL expression (Fig. 5C and D).

All redesigned NRP targets were highly modified by the Ksp PTM enzymes (Fig. 5B–H). For the three mimics co-produced with KspMRN, enramycin, kahalalide, and brevicidine predominantly contained a lanthionine macrocycle, ornithines, and a lysine C12-acylation, accounting for 83%, 86%, and 61% of their respective product profiles. The only other appreciable products were unacylated KspMR-modified species for enramycin (∼13%) and kahalalide (∼12%) and the ornithine-containing KspR-modified species for brevicidine (∼36%). All other modification states were minor (<1%) or not detected. Likewise, for the lokisin and kurstakin mimics co-produced with KspMN and the linear amphibactin and aquachelin mimics co-produced with KspRN, fully modified species dominated the isolated product profiles.

To differentiate between isobaric dehydroamino acids and lanthionine-containing species, products were reacted with IAA to derivatize any free cysteine thiols. This analysis suggested that the C-terminal cysteine residues were primarily occupied in lanthionine-bridged macrocycles installed by KspM. The proposed major regioisomers of the lanthionines were based on MS2 data and are indicated in Fig. 5B–F but were not conclusively defined by further analytical methods. For kahalalide and lokisin mimics with three serines/threonines each, there were ambiguities in the location of the mass losses associated with the dehydrations. Kahalalide was predominantly doubly dehydrated (Fig. S118 and S123) and lokisin predominantly singly dehydrated (Fig. S120 and S124). In these two mimics, the MS2 data gave mixed signals, which is attributed to co-eluting isobaric product mixtures with different combinations of dehydrated sites. The proposed major product based on relative b- and y-ion intensities is shown in Fig. 5C and E, with alternative sites of the dehydroamino acids indicated by asterisks.

Although the lipopeptides products from this study were generally produced with high titers (>30 mg L−1 purified precursor protein for all conditions), attempts to release the modified cores by digestion with the promiscuous bacteriocin endopeptidase LahT62 were plagued by low cleavage efficiency and poor recovery. We attribute this to the bulky and hydrophobic acyl modification adjacent to the Gly–Gly protease recognition site. We are currently evaluating alternative chemical and proteolytic cleavage strategies for core peptide release to enable bioactivity assays of the biocatalytically generated lipopeptides.

Conclusions

We presented here a promiscuous enzymatic system to rationally produce diverse lipopeptides using enzymes from the kamptornamide RiPP natural product pathway. The bioactivity of natural and synthetic lipopeptides is highly dependent on the chain length and structure of the lipid moiety.39,63 Although our Ksp model system was limited to transferring C12 lauric acid and showed higher conversion rates for N-terminal core residues than for internal or C-terminal residues, we propose that KspN homologs with different acyl-chain and regiochemical preferences may exhibit similar peptide-substrate promiscuities. Such variants could enable the engineering of lipopeptides with alternative lipid structures. We are currently working to characterize diverse selidamide N-acyltransferase family members to develop a biocatalytic arsenal for the generation of custom lipopeptides.

Despite acylation rates and overall yields of the modified precursor proteins being generally high, this work also exposes some limitations in our current technology to produce NRP mimics: the sometimes low efficiencies of combinatorial modifications by different RiPP maturases, the inability of RiPP PTM enzymes to incorporate all non-proteinogenic residues observed in NRP targets, and the need to optimize efficient release and recovery of the modified core peptide from the leader peptide. Our future work will focus on circumventing these limitations and expanding from rationally designed cores to randomized library-based screenings. A 2020 publication by Kuipers et al. suggested that replicating key structural features of the NRP brevicidine (i.e., the hydrophobic N-terminus and the macrocycle size) using core mutagenesis and RiPP biosynthetic enzymes was sufficient to mimic the antibacterial activity of the parent compound.38 Thus, we envision that the combinatorial use of peptide maturases from distinct pathways will be able to generate NRP mimics with high enough similarity to the native structures to replicate their natural bioactivities and could also be used to generate engineered lipopeptides with improved or alternative functions. Together, our work highlights the potential of RiPP engineering to unlock a versatile and modular platform for generating lipopeptide therapeutics inspired by, yet not limited to, native NRP architectures.

Author contributions

NB: conceptualization, methodology, investigation, data analysis, writing original draft preparation; MC: investigation, data analysis, writing – review & editing draft; JARE: conceptualization, investigation, validation, writing – review & editing draft; GAC: data analysis, writing – review & editing draft; ARM: investigation, data analysis; EM: investigation, data analysis; DR: methodology, writing – review & editing draft; JP: conceptualization, funding acquisition; ALV: conceptualization, validation, writing – review & editing draft (lead), funding acquisition.

Conflicts of interest

The authors report no conflicts of interest.

Supplementary Material

CB-OLF-D6CB00169F-s001

Acknowledgments

We would like to thank the following funding agencies for financial support of this work: Novartis FreeNovation FN20-0000000194, ETH Research Grant ETH-21 22-1, European Union's Horizon 2020 grant 101000392 (MARBLES), and Swiss National Science Foundation Project Grant #10006495.

Biography

Anna L. Vagstad .

Anna L. Vagstad

Open in a new tab

Anna Vagstad is Group Leader at the Institute of Microbiology at ETH Zurich. Her research involves the discovery, mechanistic understanding, and application of post-translational modification enzymes from the biosynthesis of modified peptide natural products. She earned her PhD in Chemical Biology at Johns Hopkins University with Prof. Craig Townsend, studying the biosynthetic logic of fungal aromatic polyketide assembly. During postdoctoral studies with Prof. Jörn Piel at ETH Zurich, she focused on peptide modifications catalyzed by radical S-adenosylmethionine enzymes. Her interdisciplinary work integrates biochemistry, molecular biology, and analytical chemistry to uncover new principles of microbial metabolism and enzymatic catalysis.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information included are methods describing the cloning procedures and mass spectrometry data analysis, as well as supporting figures and tables summarizing the mass spectrometry results. See DOI: https://doi.org/10.1039/d6cb00169f.

The script used to scan mzXML files and extract peaks matching predicted masses from user-defined chemical formula compositions is available at https://gitlab.ethz.ch/piel_code_public/eckert_johannes/ms_dataanalysis.

References

  1. Lee A. C.-L. Harris J. L. Khanna K. K. Hong J.-H. A comprehensive review on current advances in peptide drug development and design. Int. J. Mol. Sci. 2019;20:2383. doi: 10.3390/ijms20102383. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bech E. M. Pedersen S. L. Jensen K. J. Chemical strategies for half-life extension of biopharmaceuticals: Lipidation and its alternatives. ACS Med. Chem. Lett. 2018;9:577–580. doi: 10.1021/acsmedchemlett.8b00226. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Philips J.-C. Scheen A. Insulin detemir in the treatment of Type 1 and Type 2 diabetes. Vasc. Health Risk Manage. 2006;2:277–283. doi: 10.2147/vhrm.2006.2.3.277. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Robbel L. Marahiel M. A. Daptomycin, a bacterial lipopeptide synthesized by a nonribosomal machinery. J. Biol. Chem. 2010;285:27501–27508. doi: 10.1074/jbc.R110.128181. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Jensen S. K. Thomsen T. T. Oddo A. Franzyk H. Løbner-Olesen A. Hansen P. R. Novel cyclic lipopeptide antibiotics: Effects of acyl chain length and position. Int. J. Mol. Sci. 2020;21:5829. doi: 10.3390/ijms21165829. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Süssmuth R. D. Mainz A. Nonribosomal peptide synthesis-principles and prospects. Angew. Chem., Int. Ed. 2017;56:3770–3821. doi: 10.1002/anie.201609079. [DOI] [PubMed] [Google Scholar]
  7. Stachelhaus T. Mootz H. D. Marahiel M. A. The specificity-conferring code of adenylation domains in nonribosomal peptide synthetases. Chem. Biol. 1999;6:493–505. doi: 10.1016/S1074-5521(99)80082-9. [DOI] [PubMed] [Google Scholar]
  8. Kahne D. Leimkuhler C. Lu W. Walsh C. Glycopeptide and lipoglycopeptide Antibiotics. Chem. Rev. 2005;105:425–448. doi: 10.1021/cr030103a. [DOI] [PubMed] [Google Scholar]
  9. Zhong L. Diao X. Zhang N. Li F. Zhou H. Chen H. Bai X. Ren X. Zhang Y. Wu D. Bian X. Engineering and elucidation of the lipoinitiation process in nonribosomal peptide biosynthesis. Nat. Commun. 2021;12:296. doi: 10.1038/s41467-020-20548-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Caboche S. Pupin M. Leclère V. Fontaine A. Jacques P. Kucherov G. NORINE: a database of nonribosomal peptides. Nucleic Acids Res. 2007;36:D326. doi: 10.1093/nar/gkm792. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Ledger E. V. K. Sabnis A. Edwards A. M. Polymyxin and lipopeptide antibiotics: membrane-targeting drugs of last resort. Microbiology. 2022;168:001136. doi: 10.1099/mic.0.001136. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Strieker M. Marahiel M. A. The structural diversity of acidic lipopeptide antibiotics. ChemBioChem. 2009;10:607–616. doi: 10.1002/cbic.200800546. [DOI] [PubMed] [Google Scholar]
  13. Baltz R. H. Miao V. Wrigley S. K. Natural products to drugs: daptomycin and related lipopeptide antibiotics. Nat. Prod. Rep. 2005;22:717–741. doi: 10.1039/B416648P. [DOI] [PubMed] [Google Scholar]
  14. Bozhüyük K. A. J. Präve L. Kegler C. Schenk L. Kaiser S. Schelhas C. Shi Y.-N. Kuttenlochner W. Schreiber M. Kandler J. Alanjary M. Mohiuddin T. M. Groll M. Hochberg G. K. A. Bode H. B. Evolution-inspired engineering of nonribosomal peptide synthetases. Science. 2024;383:eadg4320. doi: 10.1126/science.adg4320. [DOI] [PubMed] [Google Scholar]
  15. Zheng Y. Cong Y. Schmidt E. W. Nair S. K. Catalysts for the enzymatic lipidation of peptides. Acc. Chem. Res. 2022;55:1313–1323. doi: 10.1021/acs.accounts.2c00108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Montalbán-López M. Scott T. A. Ramesh S. Rahman I. R. van Heel A. J. Viel J. H. Bandarian V. Dittmann E. Genilloud O. Goto Y. Grande Burgos M. J. Hill C. Kim S. Koehnke J. Latham J. A. Link A. J. Martínez B. Nair S. K. Nicolet Y. Rebuffat S. Sahl H.-G. Sareen D. Schmidt E. W. Schmitt L. Severinov K. Süssmuth R. D. Truman A. W. Wang H. Weng J.-K. van Wezel G. P. Zhang Q. Zhong J. Piel J. Mitchell D. A. Kuipers O. P. van der Donk W. A. New developments in RiPP discovery, enzymology and engineering. Nat. Prod. Rep. 2021;38:130–239. doi: 10.1039/D0NP00027B. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Wenski S. L. Thiengmag S. Helfrich E. J. N. Complex peptide natural products: Biosynthetic principles, challenges and opportunities for pathway engineering. Synth. Syst. Biotechnol. 2022;7:631–647. doi: 10.1016/j.synbio.2022.01.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Magnusson O. T. Toyama H. Saeki M. Schwarzenbacher R. Klinman J. P. The structure of a biosynthetic intermediate of pyrroloquinoline quinone (PQQ) and elucidation of the final step of PQQ biosynthesis. J. Am. Chem. Soc. 2004;126:5342–5343. doi: 10.1021/ja0493852. [DOI] [PubMed] [Google Scholar]
  19. Tan B. Breunig P. Arbib L. Kakumu Y. Biermann F. Hardes K. Hefendehl J. Helfrich E. J. N. Phylogeny-guided discovery of a promiscuous P450 macrocyclase for the production of diverse atropopeptides. Chem. Sci. 2025;16:16240–16249. doi: 10.1039/D5SC03525B. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Freeman M. F. Gurgui C. Helf M. J. Morinaka B. I. Uria A. R. Oldham N. J. Sahl H.-G. Matsunaga S. Piel J. Metagenome mining reveals polytheonamides as posttranslationally modified ribosomal peptides. Science. 2012;338:387–390. doi: 10.1126/science.1226121. [DOI] [PubMed] [Google Scholar]
  21. Scott T. A. Verest M. Farnung J. Forneris C. C. Robinson S. L. Ji X. Hubrich F. Chepkirui C. Richter D. U. Huber S. Rust P. Streiff A. B. Zhang Q. Bode J. W. Piel J. Widespread microbial utilization of ribosomal β-amino acid-containing peptides and proteins. Chem. 2022;8:2659–2677. [Google Scholar]
  22. Hubrich F. Bösch N. M. Chepkirui C. Morinaka B. I. Rust M. Gugger M. Robinson S. L. Vagstad A. L. Piel J. Ribosomally derived lipopeptides containing distinct fatty acyl moieties. Proc. Natl. Acad. Sci. U. S. A. 2022;119:e2113120119. doi: 10.1073/pnas.2113120119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Mordhorst S. Morinaka B. I. Vagstad A. L. Piel J. Posttranslationally acting arginases provide a ribosomal route to non-proteinogenic ornithine residues in diverse peptide sequences. Angew. Chem., Int. Ed. 2020;59:21442–21447. doi: 10.1002/anie.202008990. [DOI] [PubMed] [Google Scholar]
  24. Weir E. Anterola I. van der Donk W. A. Structure, biosynthesis, and bioactivity of nostolysamides. Biochemistry. 2026;65:1578–1591. doi: 10.1021/acs.biochem.6c00103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Wiebach V. Mainz A. Siegert M.-A. J. Jungmann N. A. Lesquame G. Tirat S. Dreux-Zigha A. Aszodi J. Le Beller D. Süssmuth R. D. The anti-staphylococcal lipolanthines are ribosomally synthesized lipopeptides. Nat. Chem. Biol. 2018;14:652–654. doi: 10.1038/s41589-018-0068-6. [DOI] [PubMed] [Google Scholar]
  26. Kozakai R. Ono T. Hoshino S. Takahashi H. Katsuyama Y. Sugai Y. Ozaki T. Teramoto K. Teramoto K. Tanaka K. Abe I. Asamizu S. Onaka H. Acyltransferase that catalyses the condensation of polyketide and peptide moieties of goadvionin hybrid lipopeptides. Nat. Chem. 2020;12:869–877. doi: 10.1038/s41557-020-0508-2. [DOI] [PubMed] [Google Scholar]
  27. Ren H. Huang C. Pan Y. Dommaraju S. R. Cui H. Li M. Gadgil M. G. Mitchell D. A. Zhao H. Non-modular fatty acid synthases yield distinct N-terminal acylation in ribosomal peptides. Nat. Chem. 2024;16:1320–1329. doi: 10.1038/s41557-024-01491-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Ma S. Li R. Gao X. Gemmell E. Guo S. Chen H. Huang X. Capper M. Liu Z. Deng Z. Ding W. Köhnke J. Wang X. Zhang Q. An N-acetylated daropeptide modulates nematode development. Proc. Natl. Acad. Sci. U. S. A. 2026;123:e2617449123. doi: 10.1073/pnas.2617449123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Donia M. S. Ravel J. Schmidt E. W. A global assembly line for cyanobactins. Nat. Chem. Biol. 2008;4:341–343. doi: 10.1038/nchembio.84. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Okada M. Sato I. Cho S. J. Iwata H. Nishio T. Dubnau D. Sakagami Y. Structure of the Bacillus subtilis quorum-sensing peptide pheromone ComX. Nat. Chem. Biol. 2005;1:23–24. doi: 10.1038/nchembio709. [DOI] [PubMed] [Google Scholar]
  31. Hubrich F. Kandy S. K. Chepkirui C. Padhi C. Mordhorst S. Moosmann P. Zhu T. Gugger M. Chekan J. R. Piel J. Ribosomal peptides with polycyclic isoprenoid moieties. Chem. 2024;10:3224–3242. doi: 10.1016/j.chempr.2024.07.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Chekan J. R. Ongpipattanakul C. Nair S. K. Steric complementarity directs sequence promiscuous leader binding in RiPP biosynthesis. Proc. Natl. Acad. Sci. U. S. A. 2019;116:24049–24055. doi: 10.1073/pnas.1908364116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Mordhorst S. Ruijne F. Vagstad A. L. Kuipers O. P. Piel J. Emulating nonribosomal peptides with ribosomal biosynthetic strategies. RSC Chem. Biol. 2023;4:7–36. doi: 10.1039/D2CB00169A. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Morinaka B. I. Vagstad A. L. Helf M. J. Gugger M. Kegler C. Freeman M. F. Bode H. B. Piel J. Radical S-adenosyl methionine epimerases: regioselective introduction of diverse d-amino acid patterns into peptide natural products. Angew. Chem., Int. Ed. 2014;53:8503–8507. doi: 10.1002/anie.201400478. [DOI] [PubMed] [Google Scholar]
  35. Korneli M. Fuchs S. W. Felder K. Ernst C. Zinsli L. V. Piel J. Promiscuous installation of d-amino acids in gene-encoded peptides. ACS Synth. Biol. 2021;10:236–242. doi: 10.1021/acssynbio.0c00470. [DOI] [PubMed] [Google Scholar]
  36. Lee A. R. Carter R. S. Imani A. S. Dommaraju S. R. Hudson G. A. Mitchell D. A. Freeman M. F. Discovery of borosin catalytic strategies and function through bioinformatic profiling. ACS Chem. Biol. 2024;19:1116–1124. doi: 10.1021/acschembio.4c00066. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Truman A. W. Cyclisation mechanisms in the biosynthesis of ribosomally synthesised and post-translationally modified peptides. Beilstein J. Org. Chem. 2016;12:1250–1268. doi: 10.3762/bjoc.12.120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Zhao X. Li Z. Kuipers O. P. Mimicry of a non-ribosomally produced antimicrobial, brevicidine, by ribosomal synthesis and post-translational modification. Cell Chem. Biol. 2020;27:1262–1271. doi: 10.1016/j.chembiol.2020.07.005. [DOI] [PubMed] [Google Scholar]
  39. Yim V. Kavianinia I. Knottenbelt M. K. Ferguson S. A. Cook G. M. Swift S. Chakraborty A. Allison J. R. Cameron A. J. Harris P. W. R. Brimble M. A. “CLipP”ing on lipids to generate antibacterial lipopeptides. Chem. Sci. 2020;11:5759–5765. doi: 10.1039/D0SC01814G. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Gibson D. G. Young L. Chuang R.-Y. Venter J. C. Hutchison C. A. Smith H. O. Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat. Methods. 2009;6:343–345. doi: 10.1038/nmeth.1318. [DOI] [PubMed] [Google Scholar]
  41. Bösch N. M. Borsa M. Greczmiel U. Morinaka B. I. Gugger M. Oxenius A. Vagstad A. L. Piel J. Landornamides: antiviral ornithine-containing ribosomal peptides discovered through genome mining. Angew. Chem., Int. Ed. 2020;59:11763–11768. doi: 10.1002/anie.201916321. [DOI] [PubMed] [Google Scholar]
  42. Helf M. J. Fox B. W. Artyukhin A. B. Zhang Y. K. Schroeder F. C. Comparative metabolomics with Metaboseek reveals functions of a conserved fat metabolism pathway in C. elegans. Nat. Commun. 2022;13:782. doi: 10.1038/s41467-022-28391-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Haft D. H. Basu M. K. Mitchell D. A. Expansion of ribosomally produced natural products: a nitrile hydratase- and Nif11-related precursor family. BMC Biol. 2010;8:70. doi: 10.1186/1741-7007-8-70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Madeira F. Pearce M. Tivey A. R. N. Basutkar P. Lee J. Edbali O. Madhusoodanan N. Kolesnikov A. Lopez R. Search and sequence analysis tools services from EMBL-EBI in 2022. Nucleic Acids Res. 2022;50:W276–W279. doi: 10.1093/nar/gkac240. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Bregman M. H. Cogan D. P. Shelton K. E. Rice A. J. Dommaraju S. R. Nair S. K. Mitchell D. A. Structure-based discovery and definition of RiPP recognition elements. mSystems. 2025;10:e01252. doi: 10.1128/msystems.01252-25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Koivusalo M. Haimi P. Heikinheimo L. Kostiainen R. Somerharju P. Quantitative determination of phospholipid compositions by ESI-MS: effects of acyl chain length, unsaturation, and lipid concentration on instrument response. J. Lipid Res. 2001;42:663–672. doi: 10.1016/S0022-2275(20)31176-7. [DOI] [PubMed] [Google Scholar]
  47. Acevedo-Rocha C. G. Reetz M. T. Nov Y. Economical analysis of saturation mutagenesis experiments. Sci. Rep. 2015;5:10654. doi: 10.1038/srep10654. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Frey S. Görlich D. A new set of highly efficient, tag-cleaving proteases for purifying recombinant proteins. J. Chromatogr. A. 2014;1337:95–105. doi: 10.1016/j.chroma.2014.02.029. [DOI] [PubMed] [Google Scholar]
  49. Szymański M. Chmielewska S. Czyżewska U. Malinowska M. Tylicki A. Echinocandins – structure, mechanism of action and use in antifungal therapy. J. Enzyme Inhib. Med. Chem. 2022;37:876. doi: 10.1080/14756366.2022.2050224. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Kawakami M. Nagai Y. Fujii T. Mitsuhashi S. Anti-microbial activities of enduracidin (enramycin) in vitro and in vivo. J. Antibiot. 1971;24:583–586. doi: 10.7164/antibiotics.24.583. [DOI] [PubMed] [Google Scholar]
  51. Cruz L. J. Luque-Ortega J. R. Rivas L. Albericio F. Kahalalide F, an antitumor depsipeptide in clinical trials, and its analogues as effective antileishmanial agents. Mol. Pharm. 2009;6:813–824. doi: 10.1021/mp8001039. [DOI] [PubMed] [Google Scholar]
  52. Sørensen D. Nielsen T. H. Sørensen J. Christophersen C. Cyclic lipoundecapeptide lokisin from Pseudomonas sp. strain DSS41. Tetrahedron Lett. 2002;43:4421–4423. doi: 10.1016/S0040-4039(02)00856-0. [DOI] [Google Scholar]
  53. Götze S. Herbst-Irmer R. Klapper M. Görls H. Schneider K. R. A. Barnett R. Burks T. Neu U. Stallforth P. Structure, biosynthesis, and biological activity of the cyclic lipopeptide anikasin. ACS Chem. Biol. 2017;12:2498–2502. doi: 10.1021/acschembio.7b00589. [DOI] [PubMed] [Google Scholar]
  54. Chen X. Lu Y. Shan M. Zhao H. Lu Z. Lu Y. A mini-review: mechanism of antimicrobial action and application of surfactin. World J. Microbiol. Biotechnol. 2022;38:143. doi: 10.1007/s11274-022-03323-3. [DOI] [PubMed] [Google Scholar]
  55. Arias-Orozco P. Inklaar M. Lanooij J. Cebrián R. Kuipers O. P. Functional expression and characterization of the highly promiscuous lanthipeptide synthetase SyncM, enabling the production of lanthipeptides with a broad range of ring topologies. ACS Synth. Biol. 2021;10:2579–2591. doi: 10.1021/acssynbio.1c00224. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Le T. Jeanne Dit Fouque K. Santos-Fernandez M. Navo C. D. Jiménez-Osés G. Sarksian R. Fernandez-Lima F. A. van der Donk W. A. Substrate sequence controls regioselectivity of lanthionine formation by ProcM. J. Am. Chem. Soc. 2021;143:18733–18743. doi: 10.1021/jacs.1c09370. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Mi X. Desormeaux E. K. Le T. T. van der Donk W. A. Shukla D. Sequence controlled secondary structure is important for the site-selectivity of lanthipeptide cyclization. Chem. Sci. 2023;14:6904–6914. doi: 10.1039/D2SC06546K. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Li Y.-X. Zhong Z. Zhang W.-P. Qian P.-Y. Discovery of cationic nonribosomal peptides as Gram-negative antibiotics through global genome mining. Nat. Commun. 2018;9:3273. doi: 10.1038/s41467-018-05781-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Hathout Y. Ho Y. P. Ryzhov V. Demirev P. Fenselau C. Kurstakins: a new class of lipopeptides isolated from Bacillus thuringiensis. J. Nat. Prod. 2000;63:1492–1496. doi: 10.1021/np000169q. [DOI] [PubMed] [Google Scholar]
  60. Campbell R. Mevers E. Structural characterization and multiomics analysis reveal extensive diversity and global distribution of kurstakin lipopeptides. J. Nat. Prod. 2025;88:2960–2967. doi: 10.1021/acs.jnatprod.5c01212. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Martinez J. S. Carter-Franklin J. N. Mann E. L. Martin J. D. Haygood M. G. Butler A. Structure and membrane affinity of a suite of amphiphilic siderophores produced by a marine bacterium. Proc. Natl. Acad. Sci. U. S. A. 2003;100:3754–3759. doi: 10.1073/pnas.0637444100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Bobeica S. C. Dong S.-H. Huo L. Mazo N. McLaughlin M. I. Jiménez-Osés G. Nair S. K. van der Donk W. A. Insights into AMS/PCAT transporters from biochemical and structural characterization of a double glycine motif protease. eLife. 2019;8:e42305. doi: 10.7554/eLife.42305. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Kowalczyk R. Harris P. W. R. Williams G. M. Yang S.-H. Brimble M. A. Peptide lipidation – A synthetic strategy to afford peptide based therapeutics. Adv. Exp. Med. Biol. 2017;1030:185–227. doi: 10.1007/978-3-319-66095-0_9. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

CB-OLF-D6CB00169F-s001
CB-OLF-D6CB00169F-s001.pdf (10.3MB, pdf)

Data Availability Statement

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information included are methods describing the cloning procedures and mass spectrometry data analysis, as well as supporting figures and tables summarizing the mass spectrometry results. See DOI: https://doi.org/10.1039/d6cb00169f.

The script used to scan mzXML files and extract peaks matching predicted masses from user-defined chemical formula compositions is available at https://gitlab.ethz.ch/piel_code_public/eckert_johannes/ms_dataanalysis.

Articles from RSC Chemical Biology are provided here courtesy of Royal Society of Chemistry

RESOURCES