⚗️ biochemistry

Substrate-dependent crosslinking by the cytochrome P450 from aminopyruvatide biosynthesis

This study demonstrates that the cytochrome P450 enzyme ApyO exhibits remarkable substrate plasticity, catalyzing diverse crosslinking reactions—including C-C, N-C, and C-O bonds—between aromatic residues in aminopyruvatide precursors to generate macrocyclic peptides with potent protease inhibitory activity.

Original authors: Padhi, C., Nguyen, D. T., Zhu, L., Cha, L., Wald, J. W., Mitchell, D. A., van der Donk, W.

Published 2026-09-01

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Padhi, C., Nguyen, D. T., Zhu, L., Cha, L., Wald, J. W., Mitchell, D. A., van der Donk, W.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Nature has long relied on a single, elegant trick to build complex molecular machines: it takes a simple string of amino acids, the building blocks of proteins, and twists them into tight, durable loops. These loops, known as macrocycles, are often more stable and potent than their straight-chain cousins, making them valuable tools for fighting disease. One of nature's most versatile tools for creating these loops is a family of enzymes called cytochrome P450s. Found in organisms ranging from bacteria to humans, these enzymes act as molecular sculptors, capable of snapping together different parts of a protein chain to form new chemical bonds. While scientists have known for some time that these enzymes can link two aromatic rings—structures found in amino acids like tyrosine and tryptophan—to close a loop, the exact rules governing how they choose which atoms to connect have remained somewhat mysterious. Understanding these rules is crucial because it could allow researchers to reprogram these enzymes to build entirely new types of medicines, potentially targeting diseases that current drugs cannot reach.

In a recent study, researchers set out to map the boundaries of this enzymatic sculpting, focusing on a specific P450 enzyme called ApyO, which is involved in making a natural product known as aminopyruvatide. This enzyme normally takes a precursor peptide containing a specific three-letter sequence, tyrosine-leucine-tyrosine, and fuses the two tyrosine units together with a carbon-carbon bond to form a ring. The team wanted to see if the enzyme would be flexible enough to accept changes in the surrounding sequence and, if so, whether those changes would alter the type of bond formed. To test this, they used a method called cell-free translation, which allows them to synthesize short protein fragments in a test tube without needing living cells. This approach let them rapidly create dozens of variations of the peptide, swapping out specific amino acids to see how the enzyme reacted.

The researchers first confirmed that the enzyme could work on a very short piece of the protein, just ten amino acids long, provided that a specific arginine residue near the start of the sequence was present. This finding suggested that the enzyme recognizes a compact region of the peptide rather than needing the entire protein structure to function. When they swapped the central leucine in the sequence with other amino acids, the enzyme continued to work, but the results were surprisingly complex. In some cases, the enzyme produced two different versions of the same molecule, which looked identical in mass but behaved differently in the lab. By using detailed imaging techniques to look at the atoms within these molecules, the team discovered that the enzyme had not just made the usual carbon-carbon bond. Instead, for certain variants, it had formed a carbon-oxygen bond, linking the two tyrosine rings through an oxygen atom. In another variation where a tryptophan replaced one of the tyrosines, the enzyme formed a nitrogen-carbon bond, connecting the rings through a nitrogen atom.

This discovery revealed a hidden plasticity in the enzyme's behavior. Rather than being a rigid machine that only makes one specific connection, ApyO appears to be sensitive to the shape and chemistry of the space between the two rings it is trying to join. When the researchers changed the amino acid sitting between the two tyrosines, the enzyme adjusted its grip, leading to different chemical outcomes. The study also showed that mass spectrometry, a common tool for weighing molecules, was not enough to distinguish between these different structures; only by looking at the specific arrangement of atoms could the researchers tell the difference between the carbon-carbon, carbon-oxygen, and nitrogen-carbon linkages. This highlights a critical point for future research: assuming that a change in mass always means the same type of chemical bond has formed can lead to incorrect conclusions about what nature is actually building.

Beyond the structural surprises, the team explored the potential medical value of these new molecules. The natural aminopyruvatide produced by the original enzyme pathway contains a specific chemical group at its end that is known to inhibit proteases, a class of enzymes that break down proteins and are often involved in disease processes like cancer metastasis and viral infection. The researchers synthesized a version of the natural product and found it to be an incredibly potent inhibitor, stopping the activity of certain proteases at concentrations as low as 0.3 nanomolar. When they tested the new variants created by the modified enzymes, most showed little to no activity. However, one specific variant, which contained the carbon-oxygen bond, did show modest ability to inhibit an enzyme called angiotensin-converting enzyme, which is a target for blood pressure medication. While this activity was far weaker than the natural product, it demonstrated that the new chemical scaffolds could still interact with biological targets.

The study concludes that the ApyO enzyme is far more adaptable than previously thought, capable of forming different types of chemical bridges depending on the subtle details of its substrate. This flexibility suggests that scientists might be able to engineer these enzymes to create a diverse library of macrocyclic compounds, each with unique shapes and chemical properties. While the new variants tested in this study did not immediately surpass the natural product in potency, the ability to generate constitutional isomers—molecules with the same atoms but different connections—opens a new avenue for exploring chemical space. The work underscores that the rules of enzymatic crosslinking are not fixed; they are a dynamic interplay between the enzyme and the specific geometry of the peptide it is modifying, offering a promising path for the future design of peptide-based therapeutics.

Technical Summary: Substrate-Dependent Crosslinking by the Cytochrome P450 from Aminopyruvatide Biosynthesis

Problem Statement
Cytochrome P450 enzymes are versatile catalysts in the biosynthesis of ribosomally synthesized and post-translationally modified peptides (RiPPs), capable of forming macrocyclic scaffolds via C−C, C−N, and C−O bonds between aromatic side chains. While previous studies have explored altering the residues directly involved in crosslinking to generate new chemical structures, the extent to which P450 enzymes tolerate modifications to residues adjacent to the crosslinking sites remains less understood. Specifically, the substrate scope of ApyO, a P450 enzyme from Burkholderia thailandensis E264 responsible for crosslinking two tyrosine residues (Tyr6 and Tyr8) in the YLY motif of the precursor peptide ApyA, had not been systematically probed regarding the conservation of flanking residues or the potential for alternative linkage chemistries (e.g., C−O vs. C−C) when the intervening residue is altered.

Methodology
The study employed a combination of bioinformatic analysis, in silico modeling, and experimental biochemistry to investigate ApyO substrate tolerance:

Bioinformatics: A PHI-BLAST search identified 53 non-redundant aminopyruvatide-like precursor peptides containing YxY, WxY, and YLW motifs. RODEO analysis confirmed the presence of P450 enzymes in the corresponding biosynthetic gene clusters (BGCs).
Structural Modeling: AlphaFold3 was used to model the ApyO enzyme in complex with the C-terminal 10-mer peptide of ApyA (GRMGEGYLYD). This model identified key interactions, including a hydrophilic pocket for the conserved Arg1/Lys1 and hydrophobic pockets for Leu7 and Met2.
Cell-Free Expression (CFE) and Mutagenesis: Using a CFE system (PURExpress), the authors generated truncated ApyA variants (10-mer and 8-mer) and site-directed mutants (e.g., Y6W, Y8W, L7Y, L7W, L7F, L7I, L7V, R1 variants). These were incubated with purified ApyO and redox partners to assess crosslinking efficiency.
Structural Characterization: Products were analyzed via LC-MS, HR-MS/MS, and extensive NMR spectroscopy (1D/2D TOCSY, NOESY, HSQC, HMBC) to determine the connectivity of the crosslinks.
Bioactivity Assays: The natural product (desmethylaminopyruvatide A) and synthetic variants were tested against a panel of proteases (Cathepsin B, Cathepsin L, Papain, ACE, Mpro, Trypsin, Chymotrypsin) to evaluate inhibitory activity.

Key Results

Substrate Scope and Minimal Requirements: ApyO was confirmed to accept the C-terminal 10-mer of ApyA as a viable substrate, with a determined $k_{cat}$ of $1.63 \pm 0.04 \text{ min}^{-1}$ and $K_m$ of $80 \pm 7 \mu\text{M}$ . The AlphaFold model and mutagenesis data revealed that a conserved Arg/Lys at position 1 (N-terminal to the YLY motif) is strictly required for activity; substitution of Arg1 with most amino acids abolished catalysis, though the R1K variant retained low activity (~8% conversion). Conversely, the Leu7 residue (between the two tyrosines) showed significant tolerance for substitution.
Alternative Linkage Chemistries:
- N−C Linkage: Substitution of Tyr6 with Tryptophan (Y6W) resulted in the formation of an N−C bond between the indole nitrogen (N $\epsilon$ 1) of Trp and the C $\epsilon$ 2 of Tyr8, confirmed by NMR.
- Constitutional Isomers (C−C vs. C−O): Substitution of the central Leu7 with Tyr (L7Y) or Trp (L7W) yielded two isobaric products. Structural elucidation revealed that one isomer retained the native C $\epsilon$ 2−C $\epsilon$ 2 C−C linkage, while the second isomer formed a non-native C $\epsilon$ 2−O C−O linkage between the two tyrosine rings.
Bioactivity:
- Natural Product: The desmethylaminopyruvatide A (containing the native C−C crosslink and a C-terminal $\alpha$ -keto- $\beta$ -amino acid moiety) exhibited sub-nanomolar inhibition ( $IC_{50}$ 0.3–0.6 nM) against Cys proteases (Cathepsin L, Cathepsin B, Papain).
- Variants: The L7Y isomer-2 (containing the C−O linkage) showed modest inhibition of ACE ( $IC_{50}$ 1.1 $\mu$ M) and Papain ( $IC_{50}$ 11.7 $\mu$ M). Other variants showed no significant inhibition. The study concluded that the $\alpha$ -keto- $\beta$ -amino acid warhead is critical for the high potency observed in the natural product, as the macrocyclic scaffold alone (without the warhead) displayed only modest activity.
Limitations in Downstream Processing: Attempts to process the L7Y variant with the full suite of tailoring enzymes (ApyH, ApyI, ApyD, ApyS) in E. coli and Burkholderia resulted in inefficient conversion and complex mixtures, preventing the isolation of a C−O linked product with the full $\alpha$ -keto- $\beta$ -amino acid modification for bioactivity testing.

Significance and Claims
The paper demonstrates that the P450 enzyme ApyO possesses significant plasticity in its macrocyclization chemistry. By altering residues adjacent to the crosslinking sites (specifically Leu7), the enzyme can be directed to form constitutional isomers with distinct bond types (C−O vs. C−C) without changing the crosslinking residues themselves. This finding challenges the assumption that crosslinking outcomes are solely determined by the reacting residues and highlights the importance of the local peptide environment.

The authors emphasize that mass spectrometry alone is insufficient for characterizing these products, as isobaric constitutional isomers (C−C vs. C−O) cannot be distinguished without full structural elucidation (e.g., NMR). Furthermore, the study establishes a short 10-mer peptide as a minimal substrate for ApyO, facilitating future engineering efforts. While the C−O linked variants showed only modest protease inhibition compared to the natural product, the study suggests that the combination of the macrocyclic scaffold and the $\alpha$ -keto acid moiety offers a promising route for diversifying protease inhibitors, drawing parallels to known inhibitors like K-13 and OF4949. The work underscores the potential of P450s in RiPP biosynthesis to access diverse chemical spaces through subtle sequence variations.

More like this