Chemists Crack the Recipe for Ring-Shaped Antibiotic Peptides with a Two-Job Molecular Handle

A stubborn synthetic problem in antibiotic chemistry has just become considerably easier. Tryptophan-crosslinked macrocyclic peptides, or TCMPs, are a family of natural products whose rigid, ring-shaped architectures are forged by an unusual chemical bridge running from the amino acid tryptophan to other residues in the peptide chain. Molecules in this class, produced by bacteria as diverse as streptomycetes and streptococci, have attracted intense attention because they display striking biological activities, including the ability to kill drug-resistant pathogens. Yet despite a decade of biosynthetic exploration, chemists have lacked a simple, broadly applicable way to build these molecules in the laboratory. A team at Princeton University led by Mohammad Seyedsayamdost, working with computational chemist Sharon Hammes-Schiffer, now reports in Nature Synthesis a general rhodium-catalysed strategy that closes the critical tryptophan-containing ring in a single, late-stage operation, opening the door to analogues of molecules such as the potent antibiotic tryglysin.

The central challenge in making TCMPs is the macrocyclization itself. Ring-closing reactions are notoriously unforgiving: as the two reactive ends of a growing chain are brought together, entropy works against the chemist, and competing reactions such as dimerization or simple decomposition often win out. The problem is compounded in TCMPs because the bond that must be formed is a carbon–nitrogen linkage onto the indole ring of tryptophan, a relatively unreactive aromatic system buried within a complex, functionally dense peptide. Previous total syntheses of related crosslinked peptides, such as celogentin C, complestatin and streptide, each required bespoke, multi-step campaigns tailored to a single target. What the Princeton group sought instead was a modular disconnection: assemble a linear peptide by standard solid-phase chemistry, then snap it into its macrocyclic form with one catalytic reaction that works across many substrates.

The answer they arrived at is a dual-functional directing group, a small pyridyl-based handle that performs two jobs in the same catalytic cycle. First, it coordinates to a rhodium(III) catalyst and steers the metal to a specific carbon–hydrogen bond on an aniline unit within the peptide, a process known as ortho-metalation. This C–H activation step, which avoids the need for pre-functionalized starting materials, generates a rhodium–carbon bond at exactly the position where the new ring bond must form. Second, the same directing group then acts as an internal nucleophile, delivering a nitrogen atom to forge the carbon–nitrogen bond that completes the macrocycle. In effect, the handle is both the rudder and the engine of the reaction, positioning the catalyst and then participating directly in bond formation before being removed.

Getting both functions to work in one molecule was not straightforward. The team began with simpler pyrimidyl-substituted aniline substrates and developed conditions for intermolecular coupling before moving to the far more demanding intramolecular case. Mechanistic studies of the rhodium-catalysed macrocyclization, combined with detailed density functional theory calculations of the catalytic cycle carried out by Shiloh Andersson in the Hammes-Schiffer group, guided the iterative redesign of the directing group. The computations revealed how electronic properties of the pyridyl ring influenced each elementary step, from metalation through to carbon–nitrogen bond formation, allowing the chemists to tune the group until the reaction worked in complex molecular settings. The optimized handle is a 5-(N,N-dimethylamino)pyrid-2-yl group, whose dimethylamino substituent proved decisive for reactivity.

Practicality was a guiding design principle throughout. The directing group can be installed using well-established cross-coupling chemistry, survives standard peptide coupling sequences used to assemble the linear precursor, and is removed cleanly once the desired macrocycle has formed. This traceless quality matters enormously for a general method, because any residual auxiliary would complicate downstream biology. The reaction itself is an intramolecular transannular macrocyclization, meaning the nitrogen nucleophile reaches across the peptide to attack the metalated tryptophan-derived carbon, closing the ring from within. By conducting the cyclization late in the synthesis, the approach sidesteps many of the solubility and stability problems that plague attempts to build large rings early from small fragments.

The substrate scope reported in the study is broad. The authors demonstrate that the optimized rhodium-catalysed macrocyclization tolerates a range of peptide lengths, amino acid sequences and substitution patterns, delivering a diverse collection of tryptophan-crosslinked macrocycles in the rigid, strained topologies that make these natural products so challenging. X-ray crystallographic analysis of representative products, deposited with the Cambridge Crystallographic Data Center under accession numbers 2487508 and 2487632, confirmed the structures and the geometry of the newly formed crosslinks. The ability to access strained cyclophane-like arrangements, in which aromatic rings are forced into close proximity, is particularly notable, since such architectures were previously reachable only through laborious stepwise routes.

To showcase the method’s power, the team targeted the pharmacophores of biologically significant TCMPs. Among the molecules synthesized were the core ring systems of micitide 982 and of tryglysin, an antimicrobial peptide produced by the oral pathogen Streptococcus mutans. Tryglysin belongs to the RaS-RiPP family, a class of ribosomally synthesized and post-translationally modified peptides whose radical S-adenosylmethionine enzymes install remarkable crosslinks, including the unprecedented lysine-to-tryptophan linkage first characterized by the Seyedsayamdost group in 2015 in the related peptide streptide. By reproducing these pharmacophores chemically, the new route gives researchers a way to make analogues that nature does not produce, a prerequisite for structure–activity studies and drug optimization.

The broader context makes the advance timely. Natural products have historically been the richest source of antibiotics, and ribosomally synthesized peptides in particular are increasingly mined from bacterial genomes as leads against resistant Gram-positive and Gram-negative pathogens. Yet antibiotic development stands at a crossroads, with resistance eroding the clinical value of existing drugs faster than new scaffolds are entering the pipeline. Chemical synthesis complements biosynthetic discovery by allowing medicinal chemists to interrogate and refine natural scaffolds. Methods that make complex peptide architectures accessible on useful scales, and that permit late-stage diversification of a common macrocyclic intermediate, directly address the bottleneck between genome-mined sequence data and testable drug candidates.

The study also highlights how computation and experiment now work hand in hand in catalyst development. Rather than screening directing groups blindly, the Princeton team used mechanistic analysis and quantum chemical calculations to understand why early candidates failed and what electronic features the ideal group would need. This mechanistically guided design philosophy, applied to a rhodium(III)-catalysed C–H functionalization platform the group had previously used to make non-canonical tryptophan variants, converted a niche transformation into a general synthesis. The work has already been protected by a provisional patent application covering the transannular macrocyclization, suggesting the authors see near-term applications in probe development and therapeutic discovery.

For the field of peptide chemistry, the implications extend beyond the specific molecules made here. C–H activation has long promised to simplify synthesis by letting chemists form bonds at otherwise inert positions, but applications in highly functionalized peptide settings have remained rare. A removable, dual-function directing group that promotes both metalation and nucleophilic bond formation could inspire analogous strategies for other crosslink types found in natural products, from carbon–carbon links in darobactin to the ether bridges of glycopeptide antibiotics. As genome mining continues to reveal new TCMPs hidden in streptococci and the human microbiome, chemists now hold a general tool to make them, modify them and find out what they can do.

Subject of Research: Rhodium-catalysed synthesis of tryptophan-crosslinked macrocyclic peptides using a dual-functional directing group

Article Title: A general synthesis of tryptophan-crosslinked macrocyclic peptides enabled by a dual-functional directing group

Article References: A general synthesis of tryptophan-crosslinked macrocyclic peptides enabled by a dual-functional directing group. (n.d.). https://doi.org/10.1038/s44160-026-01184-4

Image Credits: AI Generated

DOI: 10.1038/s44160-026-01184-4

Keywords: macrocyclic peptides, tryptophan crosslink, C–H activation, rhodium catalysis, directing group, natural product synthesis, tryglysin, RaS-RiPPs, antibiotics, macrocyclization, density functional theory, Streptococcus mutans