Bacterial Cytochrome P450 Catalyzed Macrocyclization of Ribosomal Peptides

Abstract

Macrocyclization is a vital process in the biosynthesis of ribosomally synthesized and post-translationally modified peptides (RiPPs), significantly enhancing their structural diversity and biological activity. Universally found in living organisms, cytochrome P450 enzymes (P450s) are versatile catalysts that facilitate a wide array of chemical transformations and have recently been discovered to contribute to the expansion and complexity of the chemical spectrum of RiPPs. Particularly, P450-catalyzed biaryl-bridged RiPPs, characterized by highly modified structures, represent an intriguing but underexplored class of natural products, as demonstrated by the recent discovery of tryptorubin A, biarylitide and cittilin. These P450 enzymes demonstrate their versatility by facilitating peptide macrocyclization through the formation of carbon-carbon (C-C), carbon-nitrogen (C-N) and ether bonds between the side chains of tyrosine (Tyr), tryptophan (Trp) and histidine (His). This Review briefly highlights the latest progress in P450-catalyzed macrocyclization within RiPP biosynthesis, resulting in the generation of structurally complex RiPPs. These findings have expedited the discovery and detailed analysis of new P450s engaged in RiPP biosynthetic pathways.

Figure 1

(A) Representative RiPPs initially catalyzed by P450s, with the P450s-installed cross-links highlighted in bold red. (B) Different linkages between aromatic residues introduced by P450s in RiPP biosynthesis. The various linkages, represented by red lines, numbered in blue, and featuring red oxygen atoms, represent the bonds introduced by P450 enzymes. Numbers 1–4 represent the P450s catalyzing C–C or C–N bonds between two tryptophan residues. Notably, numbers 3 and 4 also illustrate the capability of the P450s to catalyze an extra C–N bond between α-N and C2 of one of the tryptophan residues. The correlation numbers indicated in Figures 3 and 5 correspond to those in Figure 1B, symbolizing the same linkage.

Figure 2

Generic RiPP biosynthesis and precise mining workflows for P450-catalyzed RiPP BGCs. (A) Generic RiPP biosynthetic gene cluster. (B) Detailed workflow developed by the Kim group to identify P450-catalyzed RiPP BGCs accurately. They utilized PSI-BLAST for the detection of homologous P450s, followed by the application of RODEO to locate potential precursor peptides encoded near these P450s. The EFI-EST tool was then employed to discover shared sequence patterns among the putative precursors. To investigate unique patterns in the C-terminal regions (core regions) of potential precursors with a minimum of two aromatic residues, they used MAFFT (Multiple Alignment using Fast Fourier Transform) analysis and a Python script. (C) Two distinct workflows were developed by the Ge group for genome mining of P450-catalyzed RiPP BGCs. (C,i) They used three P450 enzymes to conduct a BLASTP analysis against the NCBI database, identified 13,896 P450 sequences, predicted potential precursor peptides near these P450s using the RiPPER tool, and selected sequences with two or more conserved aromatic amino acids at the C-terminus. (C,ii) The UniRef90 database was the source of P450 sequences, and then they used a script to predict short peptides to predict the precursor peptide based on the ribosome binding site in combination with short peptides with two or more specific residues. (D) Our group analyzed the actinobacteria genomes using the SPECO workflow, yielding the small-peptide-P450 pairs. A coconservation analysis was conducted for accuracy, and non-RiPP cases were filtered out using structure-based prediction (AlphaFold-Multimer) of precursor-enzyme interaction.

Figure 3

Organization of the 21 P450-catalyzed RiPP BGCs is color-coded, with genes encoding precursor peptides represented in light red and P450-encoding genes represented in light blue. The schematic of the gene cluster is drawn using the ChiPlot online Web site. (i) One P450 enzyme catalyzes a single cyclization on the precursor peptide; (ii) one P450 enzyme catalyzes multiple cross-links on the precursor peptide; (iii) multiple P450 enzymes catalyze multiple cross-links on a single precursor peptide. (The red cross-links in precursor peptide sequences are established by their respective P450 enzymes, while the blue amino acids denote linking residues. The bracketed numbers correspond to the various linkages depicted in Figure 1B.)

Figure 4

Representative P450-catalyzed RiPPs recently discovered by Kim, Ge, and our groups. (The cross-links or hydroxylation indicated in red bold in the compound structures are installed by the corresponding P450s.)

Figure 5

Analysis of the substrate tolerance of the P450s. (A) Cross-reactivities of five P450s (SroB, PruB, SlpB, YokB and AzaB) were tested using hybrid precursor peptides. (B) Substrate tolerance analysis for SroB, KstB, ScnB and MciB on the mutant variants of their corresponding precursor peptides. (Pink indicates mutated amino acids, white signifies deleted amino acids, and yellow highlights added amino acids. The red lines represent cross-links or hydroxylation modifications determined by NMR data, while the gray lines represent cross-links or hydroxylation modifications determined by MS or HPLC data.)