Coupled Natural Fusion Enzymes in a Novel Biocatalytic Cascade Convert Fatty Acids to Amines

Abstract

Tambjamine YP1 is a pyrrole-containing natural product. Analysis of the enzymes encoded in the Pseudoalteromonas tunicata “tam” biosynthetic gene cluster (BGC) identified a unique di-domain biocatalyst (PtTamH). Sequence and bioinformatic analysis predicts that PtTamH comprises an N-terminal, pyridoxal 5′-phosphate (PLP)-dependent transaminase (TA) domain fused to a NADH-dependent C-terminal thioester reductase (TR) domain. Spectroscopic and chemical analysis revealed that the TA domain binds PLP, utilizes l-Glu as an amine donor, accepts a range of fatty aldehydes (C₇–C₁₄ with a preference for C₁₂), and produces the corresponding amines. The previously characterized PtTamA from the “tam” BGC is an ATP-dependent, di-domain enzyme comprising a class I adenylation domain fused to an acyl carrier protein (ACP). Since recombinant PtTamA catalyzes the activation and thioesterification of C₁₂ acid to the holo-ACP domain, we hypothesized that C₁₂ ACP is the natural substrate for PtTamH. PtTamA and PtTamH were successfully coupled together in a biocatalytic cascade that converts fatty acids (FAs) to amines in one pot. Moreover, a structural model of PtTamH provides insights into how the TA and TR domains are organized. This work not only characterizes the formation of the tambjamine YP1 tail but also suggests that PtTamA and PtTamH could be useful biocatalysts for FA to amine functional group conversion.

Introduction

Natural products (NPs) continue to inspire synthetic chemists to develop routes toward a variety of interesting molecules with important, clinically useful functions.¹ Comprehensive genome sequence analysis has revealed that the encoded genes responsible for NP biosynthesis reside in biosynthetic gene clusters (BGCs).^2,3 These BGCs harbor novel and unusual biocatalysts that could potentially be applied in the synthesis of a range of targets.^4,5 If the substrate range of the native biocatalyst is too narrow for a desired function, engineering techniques such as directed evolution can be employed to expand its synthetic utility.^6,7

Prodiginines are a class of secondary metabolites found in various organisms including the prodigiosin-producing Serratia sp. (pig cluster) and Hahella chejuensis (hap cluster), as well as Streptomyces coelicolor (red cluster), which predominantly produces undecylprodiginine, and Streptomyces griseoviridis (rph cluster), which produces prodigiosin R1 (Figure 1).⁸⁻¹¹ They are structurally related to tambjamines (A-K, BE-18591, YP1) through a 4-methoxy-2,2′-bipyrrole-5-carbaldehyde (MBC) core, which is conserved throughout the entire prodiginine and tambjamine NP families (Figure 1). Extensive analysis identified a conserved set of homologous genes responsible for MBC biosynthesis within their respective BGCs. To form prodigiosin, the MBC core is condensed with another intermediate, 2-methyl-3-n-amyl-pyrrole (MAP); the first reported extraction of this NP was from the bacterium Serratia marcescens. Alternatively, a MAP derivative can be used.¹² However, in the case of tambjamines, the MBC intermediate is condensed with an amine to form an enamine moiety in place of the third pyrrole ring. The biosynthetic production of these secondary intermediates can differ, along with the enzymes utilized for their production.

Figure 1.

Structures of prodiginine and tambjamine NPs produced by the biosynthetic pathways of a number of organisms, highlighting the conserved MBC core.

Along with the prodiginine and tambjamine families, NPs’ with pyrroles built into their scaffolds are ubiquitous in nature. The planar, electron-rich ring is able to form hydrogen bonds, chelate metal ions, and participate in π-stacking interactions.¹³ These five-membered N-heterocyclic-containing products often exhibit antimicrobial, antiviral or anticancer bioactivity, which has led to their use as therapeutic agents.¹⁴⁻¹⁶

A yellow-pigmented alkaloid, identified as tambjamine YP1 (YP1), was extracted from Pseudoalteromonas tunicata.¹⁷ Bioinformatic analysis and sequence alignment with homologous red and pig proteins helped to identify a distinct BGC designated as the “tam” cluster, which is responsible for YP1 biosynthesis (Figure S1).¹⁸ Of the 19 genes predicted, only a few have been expressed and their encoded enzymes characterized. The biosynthesis of YP1 requires two converging pathways; the first pathway produces the MBC core, whereas the second pathway produces the fatty amine tail. Since MBC biosynthesis is highly conserved, the genes required to make this bipyrrole core can be confidently predicted and annotated. However, the production and attachment of the unsaturated fatty amine tail are less well-understood.

The initial aim of this study was to characterize the formation of the fatty amine precursor to YP1. It was originally proposed that this second pathway in YP1 biosynthesis was initiated by AfaA, a fatty acid CoA ligase (FACL) not found within the “tam” cluster.¹⁹ Our recent investigations showed that another enzyme, PtTamA [a di-domain enzyme comprising a class I adenylation (ANL) domain fused to an acyl carrier protein (ACP)], found within the “tam” cluster catalyzed this step. This enzyme uses adenosine triphosphate (ATP) and a C₁₂ fatty acid (FA) to catalyze the formation of a C₁₂ adenylate, which is then captured by the 4′-phosphopantetheine (4′-PP)-modified ACP domain to form an acyl-ACP-bound PtTamA.^20,21 This PtTamA analysis led us to question how the bound acyl chain is released for further downstream tailoring. Two other enzymes in the BGC (PtTamH and PtTamT) are predicted to be involved in the formation of the amine tail after PtTamA, although their exact function is yet unknown. Here, comprehensive sequence, spectroscopic, and chemical analysis revealed that PtTamH is the unusual fusion enzyme that carries out the acyl chain off-loading function. We also show that recombinant PtTamA and PtTamH act together to convert fatty acids to the corresponding amines.

Results and Discussion

Sequence Analysis

An initial bioinformatic study on PtTamH was performed to identify and characterize its catalytic domains. The deposited amino acid sequence of PtTamH (941 aa, 104 kDa, NCBI reference sequence: WP_009837236.1, Uniprot: A4C5V8) is annotated as a di-domain enzyme, with an N-terminal, pyridoxal 5′-phosphate (PLP)-dependent transaminase (TA) type III domain and a C-terminal, amino acid dehydrogenase domain (Figure S2). Approximately 50 amino acids with no annotated function connect these two domains. An exhaustive phylogenetic analysis performed using the ConSurf²²⁻²⁶ server identified PtTamH homologues from other Pseudoalteromonas species including P. citrea and Pseudoalteromonas sp.A25. More distantly related homologues with similar domain functions/organizations are present in Chitinimonas, Paucibacter, and Streptomyces species (Figure S3). It is currently unclear whether these distant homologues are involved in tambjamine biosynthesis or the closely related prodiginine biosynthetic pathway.²⁷

While the N-terminal domain was confirmed to be a PLP-dependent ω-transaminase²⁸ (ω-TA, residues M₁-K₅₀₀) by BLASTp, further sequence analysis suggests that the C-terminal domain (residues K₅₄₇–S₉₄₁) is more accurately described as a thioester reductase (TR). The proposed TR domain of PtTamH shares sequence homology with acyl-ACP reductases (AARs) involved in alkane biosynthesis. Such homologues originate from cyanobacterial species including Synechococcus elongates(29) (28.3% identity), Synechocystis sp. PCC6803 (27.6% identity), and Nostoc punctiforme(30) (27.2% identity). Following multiple sequence alignments, the canonical nucleotide-binding motif GxxGxxG (G₇₀₇xxG₇₁₀xxG₇₁₃) and a putative catalytic cysteine (C₈₈₇) were identified in PtTamH, as well as other Pseudoalteromonas homologues (Figure S4). As reported in the literature³¹ (and further evidenced by the crystal structure solved by Gao et al., PDB: 6JZY), the active residue C₂₉₄ in S. elongatus AAR (SeAAR) is responsible for acyl chain transfer from a 4'-PP carrier such as holo-ACP or coenzyme A (CoA). The resulting thioester is optimally positioned for hydride attack by nicotinamide adenine dinucleotide phosphate hydrogen (NADPH), releasing an aldehyde. An identical mechanism was proposed by Warui et al. for N. punctiforme AAR.³⁰ Given the apparent sequence similarity around this active site cysteine, it is plausible that the PtTamH TR domain employs a comparable mechanism to liberate 4'-PP-bound intermediates as aldehyde products. Thus, from this sequence analysis, PtTamH was putatively redefined a bifunctional ω-TA-TR fusion biocatalyst.

Characterization of Recombinant PtTamH

Recombinant PtTamH was cloned into the pEHISTEV plasmid and expressed from Escherichia coli with a TEV cleavable HisTag (Figures S5–S7). The addition of 500 mM sorbitol to the growth media promotes correct protein folding, improving protein solubility.³² The soluble PtTamH was isolated using a combination of cobalt-/nickel-immobilized metal affinity chromatography (IMAC) followed by size exclusion chromatography (SEC), which enabled the isolation of ∼1.5 mg/L. After initial isolation to characterize PtTamH, and due to high protein purity after IMAC, SEC was omitted from the purification process, which led to a slightly higher protein yield of ∼3 mg/L. The SEC retention volume confirmed the homodimeric nature of PtTamH (>200 kDa, see Figures S8 and S9); this observation was expected as PLP-dependent enzymes often form dimers or tetramers.³³

PtTamH Is a Transaminase

The purified PtTamH exhibited a strong yellow color and displayed a characteristic PLP spectrum when studied by UV–vis spectroscopy (Figure S10). As exemplified by the homologue CrmG, class III ω-TAs display a preference for either l-Glu or l-Ala as the amino donor; therefore, both amino acids were tested as substrates. The covalent binding of l-Glu to the key catalytic lysine (putative K₃₄₀) was confirmed by UV–vis analysis, where the emergence of a peak at 330 nm indicates the formation of the pyridoxamine 5′-phosphate (PMP) intermediate (Figure S10). To confirm the activity of the predicted N-terminal ω-TA domain, PtTamH was incubated with l-Glu or l-Ala and C₁₂ aldehyde for 24 h, and the formation of the C₁₂ amine product was monitored by liquid chromatography (LC) electrospray ionization-mass spectroscopy (ESI-MS). We observed a peak matching the retention time (16.2 min) of the C₁₂ amine standard in both reactions. The extracted ion chromatograms (EICs) revealed an ion with m/z = 186.2222 Da, which matches with the predicted mass of C₁₂ primary amine ([M + H]⁺, C₁₂H₂₈N) (Figure 2).

Figure 2.

EICs for the C₁₂ amine of the transamination reactions of PtTamH (5 μM) in the presence of either l-Glu (5 mM) or l-Ala (5 mM) and C₁₂ aldehyde (1 mM) for 24 h at 37 °C, leading to a peak with a retention time that corresponds to the amine standard. Each reaction was completed in triplicate.

Since the NP YP1 contains a C₁₂ tail, it was hypothesized that PtTamH prefers C₁₂ aldehyde as its primary substrate. We therefore probed the chain-length specificity of the PtTamH ω-TA domain using LC ESI-MS. When PtTamH was screened against a palette of aliphatic fatty aldehydes (C₆–C₁₄), the corresponding amine products (except C₆) were detected, and a clear preference for C₁₂ aldehyde was also observed (Figures 3 and S11). These data show that PtTamH displays a broad acyl chain selectivity between C₇–C₁₄ fatty aldehydes.

Figure 3.

Monitoring of the EICs of C₇–C₁₄ amine products after incubation of the C₇–C₁₄ aldehyde with the PtTamH TA domain in the presence of l-Glu. Each reaction was completed in triplicate.

PtTamH TR Domain

After demonstrating the activity of the PtTamH N-terminal ω-TA domain, we next focused on the predicted C-terminal TR domain. Inspired by the study of coelimycin biosynthesis by Awodi et al.—which involves an analogous TR-aminotransferase cooperativity³⁴—PtTamH was incubated with C₁₂ CoA and either NADH or NADPH in the presence of l-Glu. Furthermore, the reaction mixture was supplemented with 200 mM KCl and 10 mM MgCl₂ as both magnesium and potassium ions have been found to drastically improve the activity of homologous AARs.²⁹⁻³¹ However, no C₁₂ amine product was observed in either the NADH or NADPH reactions. This suggested that the PtTamH TR domain is either inactive or does not accept acyl-CoA substrates (Figure S12).

Coupled TamA-TamH Cascade

The inability of the PtTamH TR domain to accept a free acyl-CoA thioester substrate was not entirely unexpected. As previously discussed, the tambjamine YP1 BGC encodes PtTamA, a di-domain ACP-ANL natural fusion. Our earlier investigations showed that the PtTamA ANL domain can activate carboxylic acids of various chain lengths (C₆–C₁₄) and attach them to the 4′-PP arm of the fused ACP domain.^20,21 The presence of this enzyme suggests that PtTamH may be acyl-ACP dependent, transferring the acyl chain directly from the PtTamA acyl-ACP domain to the PtTamH TR domain.

Therefore, a cascade reaction was performed using purified PtTamA (Figure S13), prepared with the ACP domain in its 4'-PP-activated holo-form.^20,21PtTamA was incubated with PtTamH in the presence of C₁₂ acid, MgATP, KCl, NADH or NADPH, and l-Glu, and the production of the corresponding C₁₂ amine was monitored by LC ESI-MS (Figure 4). A peak with m/z = 186.2222 Da was detected by LC ESI-MS in the NADH reaction. In contrast, no amine is produced in the NADPH reaction, illustrating that PtTamH is NADH-specific (Figure 4). The amine product was also absent in the control reactions. Furthermore, the gas chromatography (GC)–MS study of the PtTamA-PtTamH cascade in the absence of l-Glu enabled the detection of the C₁₂ aldehyde intermediate (Figure S14). Taken all together, our data not only confirm that the PtTamH TR domain is catalytically active but also confirm that it exhibits an innate specificity for the C₁₂ ACP substrate. The four domains are therefore working in concert to convert the acid to an amine. The C₁₂ acid is activated by the PtTamA ANL domain^35,36 in an ATP-dependent reaction that leads to the formation of the C₁₂ adenylate intermediate. The PtTamA holo-ACP 4'-PP thiol acts as a nucleophile, attacking the C₁₂ adenylate and releasing AMP. Once captured, the C₁₂ ACP is then reduced by the PtTamH TR domain using NADH, forming the C₁₂ aldehyde. The final step involves transamination to give the final C₁₂ amine, catalyzed by the PtTamH TA domain (Figure 5). The experimental data supports the initial bioinformatic annotations of both PtTamA and PtTamH and show that the conversion of C₁₂ FA to C₁₂ fatty amine is catalyzed by this cascade of di-domain fusion biocatalysts (Figure 5).

Figure 4.

EICs for the coupled cascade of PtTamA and PtTamH, transforming the C₁₂ acid to the C₁₂ amine, through incubation with l-Glu, C₁₂ acid, KCl, MgCl₂, ATP, PLP, and NADH or NADPH for 24 h at 37 °C, leading to a peak with a retention time that corresponds to the amine standard in the NADH reaction. This reaction was completed in triplicate.

Figure 5.

Coupled PtTamA-PtTamH cascade for the formation of the long-chain C₁₂ amine product.

Predictive Structural Modeling

Following experimental characterization, we sought to understand the structural logic underpinning the activity of PtTamH. To this end, a head-to-head homodimeric model of PtTamH was predicted using the accurate deep-learning program ColabFold³⁷⁻³⁹ (Figure 6, see also the Supporting Information for further details) and studied by molecular dynamics simulation (MDS). The predicted ω-TA domain shares structural homology with several ω-TA crystal structures including YgjG⁴⁰ (PDB: 4UOX, 30.1% identity), PigE⁴¹ (PDB: 4PPM, 30.5% identity), ArgD (PDB: 1VEF, 31.4% identity), and CrmG⁴² (PDB: 5DDS, 29.0% identity), with the latter displaying the highest sequence coverage of the homologues identified (90%, Figure 6D). Evolutionary conservation analysis and superimposition of the aforementioned ω-TA structures show that the PLP-binding core is highly conserved, including the key PLP-binding residue K₃₄₀ (Figure S15A,B). This critical lysine covalently binds PLP via a Schiff base linkage, thereby enabling the coenzyme to participate in transamination (Figure S15B).^43,44 Furthermore, the ω-TA domains are predicted to comprise the homodimeric interface; identical interfaces exist across all solved ω-TA structures to date, and ω-TA dimerization is essential for defining the active site. Using this model, the C₁₂ external aldimine could be docked in the predicted ω-TA binding tunnel, with the PMP moiety flanked by K₃₄₀ and F₂₀₆ and the alkyl chain extending toward a hydrophobic pocket comprising F₅₁, L₄₇₀, and L₄₇₃ from chain A and M₃₆₅ from chain B (Figure S15B,C).

Figure 6.

Predicted homodimeric model of PtTamH. (A) PtTamH monomer chains A and B highlighted in green and cyan, respectively. (B) pLDDT score of the PtTamH monomer. (C) PtTamH ω-TA and TR domains highlighted in yellow and steel blue, respectively. (D) Closest known structural homologues CrmG (PDB: 5DDS, RMSD: 0.994 Å between 323 pruned atom pairs) and SeAAR (PDB: 6JZY, RMSD: 1.13 Å between 209 pruned atom pairs) superimposed on the predicted PtTamH monomer chain A.

The predicted PtTamH TR model and the crystal structure of SeAAR could be comfortably superimposed (Figure 6D), enabling further structural study. PtTamH TR is modeled with a classic Rossmann fold for nucleotide binding. An alkyl-binding pocket was also inferred by structural homology and topological analysis (Figure S16A), encompassing several hydrophobic residues (including F₅₅₀, L₅₅₄, I₅₅₅, L₅₆₀, I₅₆₃, L₅₉₁, and L₈₇₄) complementary to the lipophilicity of PtTamH’s natural substrate (Figure S16B). Like SeAAR, the putative catalytic cysteine C₈₈₇ is positioned at the intersection between the NADH- and alkyl-binding subdomains. Evolutionary conservation analysis suggests that the active core of the reductase is highly conserved, including residues involved in nucleotide binding as well as C₈₈₇ itself (Figure S17A). By comparison, the residues predicted to stabilize the acyl chain display greater variability, which may suggest an additional substrate-recognition role for this pocket. With this knowledge, both NAD⁺ and C₁₂ aldehyde ligands were docked in this predicted active site. The top-ranked outputs position the aldehyde carbonyl near C₈₈₇ and the NAD⁺ pyridine (Figure S17B), with the alkyl chain extending into the predicted hydrophobic pocket (Figure S17C,D). These poses are remarkably similar to the ligand orientation in the SeAAR crystal structure (Figure S17E) and suggest a conserved active site architecture within this class of reductases.

The fully predicted PtTamH homodimer maintained its structural integrity over the course of a 10 ns MDS (see Figures S18 and S19 in the Supporting Information for detailed analysis). The simulated 208 kDa complex positions the TR domains laterally from the ω-TA core. Studying the electrostatic properties of PtTamH revealed an electropositive surface on the solvent-exposed substrate channel of the TR domain, which may underpin the recognition/docking of the comparatively acidic PtTamA ACP (Figure S20).⁴⁵ Thus, the predicted domain organization of PtTamH would permit the facile transfer of acyl intermediates between the ACP of PtTamA and the TR domain of PtTamH. Taken alongside the experimental data, the PtTamH TR domain can be confidently classified as an AAR. The fascinating predicted architecture of PtTamH makes it an attractive target for further structural studies.

Conclusions

The application of biocatalysts for the synthesis of a range of high-value molecules (e.g., sitagliptin and islatravir)⁴⁶ is gaining in popularity.⁴⁷⁻⁴⁹ Once an enzyme has been discovered, its substrate specificity can be engineered to widen the substrate scope for a bespoke synthetic application.⁶ A rich source of these enzymes has been NP biosynthetic pathways which have evolved to efficiently transform simple building blocks into complex structures with a myriad of functionalities.⁵⁰ These NPs are often encoded in BGCs where each enzyme plays a specific role in the stepwise transformation along a linear path. Along with these discrete, single-domain biocatalysts, the polyketide and nonribosomal peptide classes of NPs are members of a large family whose biosynthesis is driven by large, multidomain assemblies.^45,51,52 The fused domains within these molecular machines have clearly evolved to efficiently transfer intermediates between active sites. A key player within the complex is the acyl-ACP substrate which relays covalently tethered substrates between domains.

The tambjamine YP1 BGC encodes a pathway to convert long-chain FAs to the essential hydrophobic fatty amine tails of these biologically active NPs. The details of this part of the pathway were unclear until we reported that the PtTamA enzyme is a di-domain fusion of an N-terminal ANL to a C-terminal ACP.^20,21 However, the details involved in the downstream processing of the novel PtTamA acyl-ACP intermediate were unknown. In this study, the putative biocatalyst PtTamH was analyzed using comprehensive bioinformatic analysis. This predicted PtTamH to also be a di-domain fusion with an N-terminal domain that displays high-sequence homology to a class III PLP-dependent TA. This is fused to a C-terminal domain whose sequence and structural homology suggest that it is a member of the AAR family. To assign a function to the individual domains, the recombinant PtTamH was initially shown to bind PLP and catalyze the conversion of C₁₂ aldehyde to the corresponding C₁₂ amine. The preferred amine donor was found to be l-Glu (but it can also accept l-Ala to a lesser extent), and it also displays a broad substrate promiscuity by being able to accept C₇–C₁₄ aldehydes. In future, the synthetic utility of this TA could be further expanded by studying its activity with smart amine donors such as cadaverine, o-xylylenediamine, and N-phenylputrescine (NPP).⁵³⁻⁵⁵

The outstanding activity to be defined was that of the origin of the aldehyde substrate, and we predicted that the TR domain would catalyze the reduction of an acyl-thioester substrate, but PtTamH was unable to convert C₁₂ CoA to the corresponding C₁₂ aldehyde. However, we established that the PtTamH TR domain was catalytically active by successfully constructing a biocatalytic cascade which converted the C₁₂ acid to the corresponding amine. Moreover, we also detected the formation of an aldehyde intermediate by omission of the amine donor. This suggests the four domains, present as two fusions within PtTamA and PtTamH, worked together and that the PtTamH TR domain requires a specific acyl-ACP-bound thioester substrate. Defining the function of PtTamH, and coupling it with PtTamA, finally resolves the origin of the fatty amine tail of tambjamine YP1.^18,20,21,56

Further work on the PtTamA-PtTamH system could be carried out to fully explore the substrate scope of this novel biocatalytic cascade. This would be enabled by determination of the three-dimensional structures of both enzymes; our sequence and structural analyses has provided initial insights into the chemistry and logic underpinning PtTamH-PtTamA cooperativity. The dimeric PtTamH is >200 kDa and is also an excellent candidate for cryogenic electron microscopy studies.⁵⁷ Further issues to be resolved include the molecular details of how each domain interacts with each other and how substrates and products navigate between catalytic sites. Tambjamine YP1 contains an oxidized acyl chain (with cis geometry between C₃ and C₄), and PtTamT within the P. tunicata BGC has been proposed to perform this oxidation. The active biocatalytic cascade described here will allow functional analysis of this remaining useful enzyme. Furthermore, along with YP1, Picott et al. have recently characterized the macrocyclic tambjamine derivative MYP1 produced by P. citrea. Initial analysis identifies PtTamA and PtTamH homologues [TreaA (Uniprot: U1J4V2) and TreaH (Uniprot: U1KHB2)], in the BGC of this organism, which both show 71% identity to PtTamA and PtTamH, respectively. Our work lays the foundation to reveal the similarities and differences between the two biosynthetic pathways.⁵⁶

Members of the ANL family are versatile biocatalysts that can be used to activate a range of FA substrates and prepare useful amides, esters, and thioesters.⁵⁸ New tools such as the database RetroBioCat should become a useful resource to incorporate such adaptable biocatalysts into synthetic routes for the preparation of a range of target molecules.⁵⁹ The work described here suggests that the functional group conversion displayed by the PtTamA–PtTamH biocatalytic cascade would be a valuable addition to this biocatalyst repository.

Glossary

Abbreviations

4′-PP

4′-phosphopantetheine

AAR

acyl-ACP reductase

ACP

acyl-carrier protein

ANL

adenylation

ATP

adenosine triphosphate

BGC

biosynthetic gene cluster

EIC

extracted ion chromatogram

FA

fatty acid

FACL

fatty acid CoA ligase

IMAC

immobilized metal affinity chromatography

MAP

2-methyl-3-n-amyl-pyrrole

MBC

4-methoxy-2,2′-bipyrrole-5-carbaldehyde

NP

natural product

NRPS

nonribosomal peptide synthase

PLP

pyridoxal 5′-phosphate

PKS

polyketide synthases

TA

transaminase

TR

thioester reductase

SEC

size exclusion chromatography

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acscatal.2c02954.

• Details of the YP1 biosynthetic pathways; sequence alignments; gene cloning; enzyme purification; enzyme assay; mass spectrometry analysis; and predictive modeling, docking, and simulation (PDF)

Author Contributions

S.M.R., P.M.M., and M.A.H. contributed equally. S.M.R. and P.M.M. carried out the biocatalyst experiments, and M.A.H. performed mass spectrometry and bioinformatics analysis. S.M.R., M.A.H., and D.J.C. wrote the manuscript. All authors have given approval to the final version of the manuscript.

We thank the Derek Stewart Charitable Trust and the School of Chemistry for PhD studentship funding (to S.M.R.). We thank the Biotechnology and Biological Sciences Research Council (BBSRC) East of Scotland Bioscience (EastBio) Doctoral Training Partnership (DTP) for funding a PhD studentship (to P.M.M., BB/J01446X/1). The BBSRC and Industrial Biotechnology Innovation Centre (IBIOIC) is also thanked for PhD funding (to M.A.H., BB/S506953/1). MS data were acquired on an instrument funded by the Engineering and Physical Sciences Research Council (EPSRC, EP/K039717/1).

The authors declare no competing financial interest.