Skip to main content
NCBI home page
JACS Au logoLink to JACS Au
. 2026 Sep 2;6(9):5455–5464. doi: 10.1021/jacsau.6c01252

Biosynthesis of Complex Phenazines Involves Dimerization and Amination Catalyzed by a Versatile Nuclear Transport Factor 2‑like Superfamily Protein

Jun-Li Dong †, Ying Dan †, Xin Xu †, Bei-Bei He †, Yi Huang †, Jing-Jing Xu †, Qing-Bo Liang †, Yi-Qing Li ‡, Li Han §, Xue-Shi Huang §, Jing He †,*
PMCID: PMC13625586  PMID: 42819288

Abstract

Phenazine natural products represent a structurally diverse class of aromatic alkaloids with broad-spectrum biological activities. Among these compounds, phenazine dimers have attracted considerable attention owing to their intricate molecular architectures and potent antibacterial and anticancer properties. Although the biosynthesis of the core phenazine scaffold is well characterized, the enzymatic postmodifications, especially those governing dimerization processes that yield structurally varied diphenazines, remain poorly understood. In this study, we elucidate the complete biosynthesis of diverse dimeric and aminated phenazines. Following core assembly, the flavoprotein DapS initiates postmodification by catalyzing the C6-selective decarboxylative hydroxylation of phenazine-1,6-dicarboxylic acid (PDC). We further uncover Dap5, a nuclear transport factor 2 (NTF2)-like protein, as a multifunctional enzyme that catalyzes hydroxylation, dimerization, and amination of phenazines. Strikingly, Dap5 operates independently of flavin or metal cofactors, instead activating oxygen by exploiting the redox cycling of its phenazine substrate within the enzyme–substrate complex. Structural and docking analyses based on a modeled protein structure suggest that these reactions likely share a common predicted active-site cavity but employ distinct residue networks for substrate binding and catalysis. This study solves a long-standing puzzle in the biosynthesis of dimeric phenazines and demonstrates a unique substrate-assisted radical strategy, offering new insights into the evolution of multifunctional enzymes.

Keywords: phenazine, biosynthesis, multifunctional, dimerization, amination

graphic file with name au6c01252_0008.webp

graphic file with name au6c01252_0006.webp

Introduction

Phenazines are a significant class of aromatic alkaloids characterized by a tricyclic dibenzopyrazine core, renowned for their potent and wide-ranging bioactivities. , To data, over 200 natural phenazines have been identified, mainly from Pseudomonas and Streptomyces, , with limited examples from other microbes (Figure a). , Among these, only a small subset contains two phenazine units (Figure a), existing either as single-bond-linked dimers (e.g., izumiphenazine B and phenazostatin J) or as fused systems that form new ring scaffolds (e.g., izumiphenazine A, phenazinolin D, diastaphenazine, and baraphenazine E). Their intricate structures and promising bioactivities have garnered increasing interest. While core scaffold biosynthesis is well established (Figure S1), , the enzymatic mechanisms underlying phenazine dimerization remain elusive. Moreover, the chemical synthesis of phenazine dimers with precise regio- and stereochemical control remains challenging. Consequently, elucidating their biosynthesis presents a dual opportunity: to uncover novel dimerization mechanisms and to enable biocatalytic access to diverse diphenazines.

1.

1

Open in a new tab

(a) Chemical structures of representative natural phenazines. (b) Genetic organization of dap BGC from Streptomyces diastaticus and dap-like BGC from Streptomyces sp. 06282-1I. (c) Proposed biosynthetic pathway for diastaphenazine (1), izumiphenazine C (2), izumiphenazine A (4), and phenazinolin D (5). Abbreviations: D-erythrose 4-phosphate (E4P); phosphoenolpyruvate (PEP); 3-deoxy-D-arabinoheptulosonate-7-phosphate (DAHP); chorismic acid (CHA); para aminobenzoic acid (PABA); phenazine-1-carboxylic acid (PCA); phenazine-1,6-dicarboxylic acid (PDC).

In 2014, Li et al. isolated the cytotoxic and antimicrobial dimeric phenazine diastaphenazine (1) and the C–N linked phenazine izumiphenazine C (2) from Streptomyces diastaticus. Our previous work identified the biosynthetic gene cluster (BGC) responsible for their production (Figure b), with genetic analysis implicating dapS and dap5 participate in scaffold tailoring (Figure c). Recently, Zhuang et al. characterized a phenazine BGC from Streptomyces sp. 06282-1I that produces seven dimeric phenazines and harbors conserved homologues of dapS and dap5. This finding strongly suggests an evolutionarily conserved function for DapS and Dap5 in phenazine dimer biosynthesis.

In this study, we demonstrated the essential roles of DapS and Dap5 in the biosynthesis of multiple phenazine compounds, including 1 and 2, through integrated in vivo feeding studies and in vitro enzymatic assays. Remarkably, the NTF2-like superfamily protein Dap5 demonstrates multifunctional catalysis, mediating both phenazine dimerization and amination via a radical-mediated mechanism that requires NAD­(P)H as an external electron donor but proceeds without detectable involvement of metal ions or flavins. This NAD­(P)­H-dependent radical chemistry is unprecedented within this superfamily, which is typified by cofactor-independent acid–base catalysis. − The functional characterization of Dap5 thus not only expands the catalytic scope of the NTF2-like fold but also prompts further investigation into the structural and evolutionary adaptations that enable radical chemistry within this conserved scaffold.

Substrate specificity studies revealed the essential role of carboxyl groups in Dap5 catalysis, facilitating generation of several novel phenazine derivatives. Protein structure analysis and substrate molecular docking studies using a modeled structure suggested different predicted active sites for hydroxylation, amination, and dimerization activities. This work resolves the long-standing mechanistic enigma of enzymatic phenazine dimerization and establishes a framework for discovering, rationally designing, and engineering multifunctional enzymatic systems for producing phenazine derivatives.

Results and Discussion

DapS and Dap5 Are Involved in the Biosynthesis of Various Complex Phenazine Compounds

Considering that Streptomyces sp. 06282-1I can produce several dimeric phenazines, we reexamined the metabolite profile of Streptomyces albus (S. albus) carrying the diastaphenazine biosynthetic gene cluster (dap) to probe its capacity for producing phenazines. Beyond the expected compounds 1, 2, and phenazine-1-carboxylic acid (PCA, 3), several additional product peaks were observed (Figure ). Large-scale fermentation of the recombinant strain enabled the isolation of two additional compounds, 4 and 5. Comprehensive ESI-HRMS and NMR analyses (Figures S28–S35; Tables S4 and S5), supported by comparison with published data, established 4 as izumiphenazine A , and 5 as phenazinolin D. , The co-occurrence of these structurally distinct phenazines within a single organism demonstrates a shared biosynthetic origin. While Streptomyces sp. PU-10A was known to produce diverse dimeric phenazines, our work now directly links this chemical diversity to the dap BGC, establishing it as a central player in assembling these complex scaffolds. This genetic assignment establishes the dap cluster as a key model system for deciphering the late-stage steps that drive phenazine diversification.

2.

2

Open in a new tab

HPLC analysis of metabolite profiles. (i) Diastaphenazine (1) standard; (ii) izumiphenazine C (2) standard; (iii) phenazine-1-carboxylic acid (PCA, 3) standard; (iv) Streptomyces albus expressing the dap BGC; (v) ΔdapS mutant; (vi) Δdap5 mutant; (vii) Streptomyces albus expressing the empty vector.

In our previous study, we established that dapS and dap5 are essential for the biosynthesis of 1 and 2. Here, we show that they are also required for the formation of 4 and 5 (Figures and S3). Metabolic profiling of the deletion strains revealed distinct intermediate accumulation patterns: the ΔdapS mutant accumulated 6, whereas the Δdap5 mutant accumulated 7 and 8 at a ratio of approximately 10:1. Scale-up fermentation enabled the isolation of 6–8, which were unequivocally identified as phenazine-1,6-dicarboxylic acid (PDC, 6), 6-hydroxyphenazine-1-carboxylic acid (7), and 1,6-dihydroxyphenazine (8) by ESI-HRMS and NMR spectroscopy (Figures S2 and S36, S43; Tables S6 and S7).

Bioinformatic analysis (Figures S4 and S5) classified DapS as a flavoprotein clustering with known phenazine hydroxylases PhzS, LaphzS, and XpzG. In contrast, Dap5 belongs to the NTF2-like superfamily but forms a distinct evolutionary branch with low similarity to functionally characterized members (Figures S6 and S7). This superfamily exhibits remarkable functional diversity, including intramolecular aldol condensation (SnoaL and AknH), regioselective C2–O bond formation (XimE), and [6 + 4] cycloaddition (StmD). Based on these analyses, we thus propose that DapS catalyzes the C1/C6 decarboxylative hydroxylation of PDC (6) to generate intermediates 7 and 8, while Dap5 mediates downstream tailoring steps to yield the final products 1, 2, 4 and 5 (Figure c). Detailed mechanistic understanding will require further biochemical and structural characterization.

In Vivo and In Vitro Reconstruction of the Biosynthetic Pathway for 1, 2, 4, and 5

To functionally verify the proposed biosynthetic pathway, we first heterologously expressed dapS in S. albus. Upon supplementation with 6, the strain S. albus::DapS produced 7 and 8 at a ratio of approximately 16:1, while the control exhibited no conversion (Figure S8). These results suggest that compound 7 is likely the intermediate for subsequent dimerization, whereas compound 8 is probably a shunt product. Coexpression of dapS and dap5 in S. albus enabled production of 1, 4 and 5 from 6, and of 2 when both 6 and 4-methylaminobenzoic acid (9) were added, as determined by ESI-HRMS (Figure S8).

For in vitro studies, dapS and dap5 were recombinantly expressed in Escherichia coli BL21 (DE3). While soluble Dap5 was readily obtained, DapS consistently formed insoluble inclusion bodies (Figures S9 and S10). To circumvent this issue, XpzG, a soluble homologue of DapS, was expressed instead (Figure S9B). In vitro enzymatic assays showed that XpzG catalyzes the NADPH-dependent sequential decarboxylative hydroxylation of 6 to 7, followed by further conversion to 8 (Figure a). Consistent with in vivo results, the combined XpzG–Dap5 system efficiently converted 6 to 1, 4 and 5, and 6 plus 9 to 2 (Figure a). Collectively, these results establish that dapS and dap5 are indispensable for the biosynthesis of 1, 2, 4 and 5, with 6 serving as the central biosynthetic precursor.

3.

3

Open in a new tab

In vitro biochemical characterization of XpzG and Dap5. (a) In vitro reconstitution of the biosynthetic pathway for compounds 1, 2, 4 and 5. (b–c) HPLC traces of Dap5-catalyzed reactions. (d) Stable isotope labeling experiment with 18O2. (e) Schematic reactions catalyzed by XpzG and Dap5.

Characterization of Dap5′s Catalytic Activity

To elucidate the biosynthetic pathway of dimeric phenazines and determine Dap5’s catalytic role in their formation, we first incubated Dap5 with the proposed substrate 7 in the presence of NADPH. The reaction mixture yielded multiple products, including the target dimers 4 and 5 (Figure b). Notably, however, compound 1 was not detected. A third product, 10, was isolated and characterized as phenacein, a C3-hydroxylated derivative of 7 (Figures S44–S51 and Table S8). Although 4 and 5 possess the backbone of 10, incubation of Dap5 with 10 alone did not yield these dimers (Figure S11A). Time-course experiments revealed the following sequence of product accumulation (Figure S11B): at 5 min, 10 was clearly detectable with only trace amounts of 4 and 5; at 30 min, 10 increased substantially along with 4 and 5; at 1 h, 10 continued to accumulate while 4 and 5 showed only marginal further increase, and 1 became detectable; by 22 h, approximately 90% of substrate 7 had been consumed and 1 reached its maximum yield. The delayed appearance of 4 and 5 relative to 10, together with the observation that 10 alone cannot serve as a substrate for dimerization, indicates that 4 and 5 are instead heterodimers formed between 7 and 10. These results establish that Dap5 catalyzes the NAD­(P)­H-dependent hydroxylation of 7 to 10 and mediates the heterodimerization of 7 with 10 to form 4 and 5.

The dimeric structure of 1 suggested a biosynthetic route involving two distinct monomers. In vitro, Dap5 exclusively catalyzed the heterodimerization of 6 and 7 to form 1, showing no activity with alternative substrates (Figures b and S11). For the biosynthesis of 2, Dap5 specifically mediated C–N coupling between 7 and 9, yielding 2 as a minor product (Figures c and S12). Collectively, these findings establish Dap5 as a versatile phenazine-tailoring enzyme that catalyzes both C–C heterodimerization and intermolecular C–N bond formation.

To investigate Dap5’s substrate specificity, we tested various phenazines (3, 8, 11, 12) and nonphenazine analogues (13–15) (Figures and S13). Dap5 hydroxylated 3 to 16, which was characterized as 4-hydroxyphenazine-1-carboxylic acid (4-OH-PCA) by ESI-HRMS and NMR analysis (Figures S52–S58 and Table S9). No hydroxylation was observed with 8, 11, or 12, all of which lack a C1-carboxylate, suggesting that this group is essential for activity. Furthermore, no products were detected with 13–15, underscoring Dap5’s stringent substrate specificity for phenazine carboxylic acids.

4.

4

Open in a new tab

Chemical structures of compounds 3, 8, 11–16, and 18–20.

To further explore Dap5’s catalytic capacity, we coincubated 3 with 9 or 7. Reaction with 9 afforded three new products (17–19) (Figure S13C). While 17 was unstable and could not be isolated, 18 and 19 were isolated and characterized as novel aminated phenazines by ESI-HRMS and NMR analysis (Figures S58–S71; Tables S10 and S11). The NMR data for 18 and 19 have been deposited in the Natural Products Magnetic Resonance Database (NP-MRD, www.np-mrd.org) under accession codes NP0351361 and NP0351362. Incubation with 7 produced a new peak (20), which gave an [M + H]+ ion at m/z 453.1199 (calcd for C25H16N4O5) (Figures S13D and S72). Based on HRMS/MS data (Figure S73) alone, a tentative structure is proposed for 20. However, due to its instability, full NMR data could not be obtained; thus, the structure remains to be confirmed by future isolation and comprehensive spectroscopic analysis. The remarkable ability of Dap5 to generate such structurally diverse derivatives raises the fundamental question of how a single enzyme can accommodate a broad range of transformations.

Dap5 Catalyzes the Biosynthesis of Complex Phenazines via a Substrate-Assisted Radical Mechanism

Given the observed requirement for NAD­(P)H (Figure S14A), we hypothesized that molecular oxygen serves as the terminal oxidant. Isotope labeling experiments (Figures d and S15) confirmed that 4, 5, and 10 incorporated oxygen from 18O2 but not H2 18O. However, 1 consistently failed to show any 18O incorporation from either 18O2 or H2 18O, even when the labeling experiments were repeated multiple times under identical conditions. The origin of the oxygen atom in 1 therefore remains unclear and warrants further investigation.

Although Dap5-catalyzed reactions, like those of canonical monooxygenases, require NAD­(P)H for O2 activation, they proceed without detectable involvement of metal ions or flavins (Figure S14), pointing to a distinct catalytic mechanism. We attribute this to the intrinsic redox activity of the phenazine substrate, which like flavin undergoes reversible redox cycling, a capability that enables it to replace flavin in electron transfer in Shewanella oneidensis. Within the Dap5 active site, this cycling is tightly controlled to generate reactive intermediates for O2 activation.

Mechanistically, substrate 7 is reduced by NAD­(P)­H, and the resulting reduced species then reacts with O2 to form superoxide anion (Figure S16). Protonation of superoxide affords the electrophilic hydroperoxyl radical, which attacks the C3 position of 7 to yield a C3-peroxy adduct. The O–O bond of this adduct may undergo homolytic cleavage to produce a C3-oxy radical (7b) and a hydroxyl radical, or heterolytic cleavage to give a C3-oxy anion (7a) and water. Subsequent electron rearrangement restores aromaticity, ultimately furnishing the stable hydroxylated product 10. The anionic intermediate 7a is proposed to serve as a key reactive species in the formation of dimeric compounds. Specifically, Michael addition between 7 and 7a generates intermediate 5a, which undergoes decarboxylation to yield 5b. 5b is then proposed to yield 5c via epoxidation. The phenolic hydroxyl group of 5c attacks the epoxide at one carbon (Pathway a) to afford the cyclized product 4, whereas attack at the other carbon (Pathway b) yields 5. For the formation of 1, 7a undergoes Michael addition with 6 to generate intermediate 1d. Intramolecular Michael addition and decarboxylation of 1d then afford 1e. A second decarboxylation of 1e, inferred from the product structure, gives 1f, whose ketone carbonyl group is then reduced to give 1. In the formation of 2, the reduced 7 is oxidized by a single electron transfer within the enzyme active site, generating a radical cation. This unstable intermediate undergoes homolytic cleavage of the C1–COOH bond, leading to decarboxylation and formation of a C1 radical, which is further oxidized to a C1 cation. Concurrently, the nitrogen atom of substrate 9 is proposed to be activated (e.g., by an active-site base), enabling nucleophilic attack on the C1 cation via its lone pair, ultimately furnishing 2. Consistent with the proposed radical mechanism, the radical scavenger TEMPO suppressed all Dap5-catalyzed reactions (Figure S17).

To gain mechanistic insight into the Dap5-catalyzed transformation, we performed Electron Paramagnetic Resonance (EPR) spin-trapping experiments using DMPO (5,5-dimethyl-1-pyrroline N-oxide) as the radical scavenger (Figure S18). Phenazines are structurally and electronically analogous to flavins, capable of existing in three oxidation states: oxidized (ox), reduced (red), and semiquinone radical (SQ•). In the nonenzymatic control containing 7 and NADPH alone, a clear DMPO adduct EPR signal was observed, indicating the formation of SQ•, which subsequently generates reactive oxygen species (ROS) via auto-oxidation. However, no substrate conversion or product formation was detected under these conditions (Figure b ii), demonstrating that this nonenzymatic electron transfer constitutes a “futile cycle” that fails to support C–O/C–N bond formation. Upon addition of Dap5, the EPR signal intensity decreased in an enzyme-concentration-dependent manner, while the formation of hydroxylated and dimerized products became evident (Figure b iii). This attenuation reflects the enzyme’s catalytic function: rather than simply generating the radical, Dap5 channels the inherent phenazine redox cycle into the productive hydroxylation/amination/dimerization pathway. By accelerating the consumption of the reduced/semiquinone intermediates toward products, the enzyme lowers their steady-state concentration, leading to the diminished EPR signal. The inverse correlation between enzyme concentration and signal intensity is thus the hallmark of a genuine catalytic intermediate, distinguishing enzyme-mediated radical chemistry from nonspecific background oxidation. Collectively, these EPR data provide direct spectroscopic evidence for a radical intermediate in the Dap5-catalyzed reaction and support the proposed substrate-assisted radical mechanism, wherein Dap5 orchestrates the coupling of radical formation with subsequent bond-forming events through precise substrate positioning and electrostatic stabilization of the transition state.

Protein Structure Analysis of Dap5

To elucidate the structural basis of Dap5′s multifunctionality, we modeled its structure using AlphaFold2. The overall structure adopts a canonical α+β barrel fold and exhibits significant similarity to proteins within the NTF2-like superfamily (Figures and S19). Molecular docking was subsequently performed to model the binding of NADPH and the relevant substrates for each reaction (Figures and S20). Analysis of the docking models revealed that in all complexes, substrate 7 binds in a large cavity between α3 and β1, with NADPH at the β5 surface. NADPH forms hydrogen bonds with Gly36, Pro88, and Asp108, hydrophobic interactions with Trp76 and Pro88, and a salt bridge with Arg110. Substrate 7, the common substrate for all three reactions, is stabilized by hydrophobic interactions with Phe34, Lys50, Ile53, Ala54, Val105 and Phe107, π–π stacking with Tyr38, and a hydrogen bond with Lys50. The second substrates for amination (9) and dimerization (6) are positioned near 7 in the same cavity and are stabilized by Tyr38, Ile53, Ile57, Val112 and Thr113 via π–π stacking, hydrophobic interactions, and hydrogen bonds.

5.

5

Open in a new tab

Structural analysis of Dap5. (a) The 3D structure of Dap5 predicted by AlphaFold2. (b) Molecular docking of substrate 7 and NADPH in the binding pocket of Dap5. Binding site residues (cyan sticks), substrate 7 (yellow sticks), and NADPH (pink sticks) are shown. Molecular interactions are indicated by dashed lines: hydrogen bonds (magenta), hydrophobic contacts (green), π–π stacking (yellow), and salt bridges (blue).

To validate the predicted interactions, we performed site-directed mutagenesis of residues lining the Dap5 active site (Figure S21). K50A, I53A, P88A, and R110A retained wild-type activity in all reactions (Figure S22). In contrast, F34A, G36A, Y38A, A54G, W76F, F107A and D108A markedly reduced or abolished Dap5 activity. Notably, I57A, V105A, V112A, and T113A showed reaction-specific defects: while dimerization and amination were markedly impaired, they retained ∼80% of wild-type hydroxylation activity. Docking models predict that Ile57, Val112, and Thr113 directly contact the second substrates (9 and 6), whereas Val105 interacts with the common substrate 7. This correlation suggests that the observed defects likely result from impaired second-substrate binding, while binding of 7 and NADPH remains largely intact.

Notably, NADPH binds at the β5 surface, spatially distant from the substrate cavity. This has precedents: in DHFR, distal NADPH-binding mutations affect catalysis 25 Å away; in CYPOR and GrhO5, NADPH binds at surface sites distinct from the active site and triggers conformational changes essential for function. Given that NADPH is strictly required for Dap5 activity, we propose that NADPH binding may similarly induce a conformational change, or that the static docking model may not fully capture the dynamic solution interactions.

This NADPH dependency distinguishes Dap5 from the vast majority of catalytically characterized NTF2-like enzymes. Members of this superfamily, such as ketosteroid isomerase, scytalone dehydratase, and the polyketide cyclase SnoaL, are known to catalyze diverse reactions (isomerization, dehydration, and aldol condensation) relying solely on conserved catalytic residues such as Asp-Tyr, Tyr-His, or Arg-Tyr pairs. These exemplify the cofactor-independent acid–base catalysis that is characteristic of the NTF2-like fold. More recently, the NTF2-like aldolase AtoB was structurally characterized and found to employ an Arg-Tyr acid–base pair for intramolecular aldol condensation without cofactor assistance. Similarly, in the context of natural product biosynthesis, the NTF2-like protein SgcJ, which is essential for the biosynthesis of the enediyne antibiotic C-1027, has also been proposed to operate via a cofactor-independent general acid–base mechanism. In sharp contrast, Dap5 requires NAD­(P)H as an external electron donor to initiate radical chemistry, a mechanistic feature thus far unprecedented among NTF2-like enzymes. This radical-based catalysis, which enables C–C and C–N bond formation, demands a level of chemical reactivity that acid–base catalysis alone cannot achieve. Dap5 therefore represents a unique member of the NTF2-like superfamily that has adapted the conserved α+β barrel scaffold to accommodate NAD­(P)­H-dependent radical transformations, significantly expanding the functional repertoire of this protein family. Future structural studies of Dap5 with its substrates will provide deeper mechanistic insights into its catalytic versatility.

Mining for Dap5 Homologous Proteins

Guided by these findings, we used Dap5 as a query for genome mining and identified 19 BGCs containing dap5 homologues (Figure S23; Table S3). The majority share a conserved organization with the original dap clustersuggesting potential for dimeric phenazine production. Strikingly, however, several exhibit distinct variations, including the absence of dapS in three clusters and the fusion of dapS and dap5 into a single open reading frame in five others. To explore their biosynthetic potential despite the inaccessibility of native strains, we selected two representative Dap5 homologues based on phylogenetic analysis (Figure S24), and purified them for in vitro characterization (Figure S25). ISL-Dap5 efficiently produced compounds 1, 2, 4, 5, and 10, mirroring Dap5 activity, whereas Rer75-Dap5 showed no detectable activity under the same conditions (Figure S26). Sequence alignment (Figure S27) suggested that substitutions at key residues (F34, Y38, A54, I57, F107) may contribute to the functional divergence between Rer75-Dap5 and Dap5, although mutagenesis studies are required to confirm their individual roles.

Conclusions

In summary, we have biochemically characterized Dap5, an NTF2-like superfamily protein, which generates diverse dimeric and aminated phenazines through a substrate-assisted radical mechanism. Remarkably, Dap5 accomplishes this catalytic versatility without the need for flavin or metal cofactors, instead exploiting the intrinsic redox activity of the phenazine substrate. These findings not only resolve a key step in phenazine biosynthesis but also establish a blueprint for the discovery and engineering of related enzymes to access complex phenazines and other bioactive molecules for pharmaceutical applications.

Methods

Heterologous Expression of dapS and dap5 in S. albus

For heterologous expression of dapS in S. albus, the encoding gene was amplified by PCR and ligated into the PstI/XbaI sites of the shuttle vector pWHM4S, yielding pWHM4S-dapS. The resulting construct was then introduced into S. albus via E. coli S17-1-mediated conjugation, generating the expression strain S. albus::DapS. For coexpression, dap5 was inserted into the XbaI site of pWHM4S-dapS to generate pWHM4S-dapS5, which was similarly introduced into S. albus by conjugation to obtain S. albus::DapS5.

In Vivo Feeding Experiments

Spores of the heterologous expression strains (S. albus::DapS or S. albus::DapS5) were inoculated into 5 mL of fresh YEME medium and cultured at 30 °C with shaking at 180 rpm for about 48 h. The resulting seed culture was then transferred to 50 mL of YM medium. After 36–48 h of growth, 0.1 mM substrate was added, and the cultures were incubated for another 3 days under the same conditions. Culture supernatants were collected by centrifugation, and metabolites were adsorbed onto XAD-16 resin (2%, v/v). The resin was then washed and extracted with 50 mL of MeOH. The combined organic extracts were concentrated to dryness under reduced pressure, and the residue was dissolved in 1.0 mL of MeOH for HPLC and ESI-HRMS analysis as described above.

Enzyme Activity Assays

In vitro assays for XpzG and/or Dap5 were carried out in 100 μL reaction volumes composed of 50 mM Tris-HCl buffer (pH 8.0), 0.2 mM substrate, 2.4 mM NADPH, and 10 μM purified enzyme. Following incubation at 37 °C for 12 h, reactions were quenched by the addition of 100 μL of EtOAc. The samples were lyophilized, and the residues were dissolved in 200 μL of MeOH. After centrifugation at 13,780 × g for 20 min, the resulting sample was analyzed by HPLC and ESI-HRMS.

Scaled-Up Reactions and Purification of Products 10, 16, 18, and 19

For preparative-scale reactions, a mixture containing 0.2 mM substrates, 2.4 mM NADPH, and 10 μM Dap5 in 50 mM Tris-HCl buffer (50 mL, pH 8.0) was incubated at 37 °C for 16 h. The reaction was terminated by the adding of EtOAc (50 mL) and extracted with a separatory funnel. The organic phase was concentrated by rotary evaporation, and the resulting residue was purified by semipreparative HPLC.

NMR Analysis

For NMR analysis, purified product was dissolved in 0.6 mL of DMSO-d 6. Spectra, including one-dimensional 1H, 13C) and two-dimensional (1H–13C HMBC, 1H–15N HMBC, HSQC, NOESY, and COSY), were recorded on a Bruker AVANCE III 600 MHz spectrometer. Data processing was performed using MestReNova software. Chemical shifts (δ) and coupling constants (J) are reported in ppm and Hz, respectively.

Phylogenetic Tree Analysis

Homologous sequences of DapS or Dap5 were identified via BLASTP against the NCBI database. Multiple sequence alignment was performed using ClustalW implemented in MEGA 7.0.26 with the following parameters: pairwise gap penalties of 10 (opening) and 0.1 (extension); multiple gap penalties of 10 (opening) and 0.2 (extension); Gonnet protein weight matrix; residue-specific and hydrophilic penalties enabled; gap separation distance of 4; end gap separation disabled; and a delay divergent cutoff of 30%.

Phylogenetic trees were reconstructed using the Maximum Likelihood (ML) method based on the Jones-Taylor-Thornton (JTT) amino acid substitution model, with uniform rates among sites (no Gamma correction). The ML tree was inferred using the Nearest-Neighbor-Interchange (NNI) heuristic search, starting from an initial tree automatically generated by NJ/BioNJ from pairwise distances. Gaps were treated by complete deletion, and branch supports were not assessed via bootstrapping. All remaining parameters were left at their default values.

Protein Modeling and Molecular Docking

The three-dimensional structure of Dap5 was predicted using AlphaFold2 via the ColabFold platform (v1.6.2). Multiple sequence alignments were generated by searching the UniRef100 and metagenomic databases using MMseqs2, as implemented in the ColabFold pipeline. Predictions were performed with three recycles, and template-based homology modeling was disabled to allow template-free prediction. All remaining parameters were kept at their default values. Model quality was assessed by stereochemical analysis using PROCHECK (assessing residue geometry) and Ramachandran plot evaluation (verifying backbone dihedral angles).

Molecular docking was performed using the AutoDockTools graphical interface within MGLTools 1.5.7. The receptor and ligands were prepared by removing water molecules, adding polar hydrogens, and assigning Gasteiger charges, then converted to PDBQT format. A grid box encompassing the putative active site was defined with center coordinates (x = −0.055, y = 0.8, z = −0.995), dimensions of 96 × 102 × 126 grid points, and a spacing of 0.375 Å. Docking was performed using AutoDock Vina via the AutoDockTools interface. A configuration file (config.txt) was generated to define the docking parameters, and the docking job was executed with default settings. The optimal binding pose was selected based on the lowest binding energy, and protein–ligand interactions were visualized and analyzed in PyMOL (http://www.pymol.org).

Construction of Dap5 Site-Directed Mutants

Site-directed mutagenesis of Dap5 was conducted via overlap extension PCR with the primers listed in Table S2 and pET28a-dap5 as the template. Following digestion with NdeI and EcoRI, the PCR products were cloned into the corresponding sites of pET28a. After transformation into E. coli DH5α, mutant plasmids were verified by DNA sequencing and subsequently introduced into E. coli BL21­(DE3) for protein expression.

Investigation of the Metal Dependence of Dap5

To investigate the dependence on metal ions, EDTA-pretreated Dap5 was incubated at 37 °C for 12 h in 50 mM Tris-HCl buffer (pH 8.0) containing substrates (6 and 7) and 2.4 mM NADPH. In parallel, reactions containing 10 μM untreated Dap5 were supplemented with 0.1 mM of various metal ions (Mg2+, Zn2+, Ni2+, Cu2+, or Fe2+). All reactions were quenched by the addition of ethyl acetate, followed by HPLC analysis as described above.

Isotope Labeling with H2 18O and 18O2

For H2 18O labeling experiments, reactions were carried out in 1.5 mL microcentrifuge tubes with a total volume of 100 μL. The reaction mixture, consisting of 50 mM Tris-HCl (pH 8.0), 0.2 mM substrate, 2.4 mM NADPH, and 10 μM enzyme, was prepared in buffer containing H2 18O. For 18O2 labeling experiments, reactions were performed in 5 mL Schlenk tubes, which were first purged with nitrogen to establish an anaerobic atmosphere, followed by the introduction of 18O2 via injection.

EPR Spin-Trapping Experiments

EPR spectra were recorded at room temperature on a Bruker EMXnano spectrometer equipped with a suitable resonator. Instrument parameters were as follows: microwave frequency, 9.6 GHz; microwave power, 7.9 mW; modulation frequency, 100 kHz; modulation amplitude, 1.0 G; center field, 3440 G; sweep width, 100 G; sweep time, 100 s; number of scans, 4; receiver gain, 1.0 × 104; temperature, 298 K. DMPO was used as the spin-trapping agent at a final concentration of 300 mM. All samples were prepared in 50 mM Tris-HCl buffer (pH 7.5) containing 13.5 mM substrate 7 and 3.6 mM NADPH, with or without purified Dap5 (10–50 μM). The reaction mixtures were incubated at 25 °C for 7 min, then transferred to a quartz capillary tube (inner diameter 2.0 mm) for immediate EPR analysis. For enzyme-concentration-dependent experiments, Dap5 was added at varying concentrations (0, 10, 25, and 50 μM). Control experiments were performed under identical conditions: (a) Fenton reaction (H2O2 + Fe2+) to verify DMPO adduct formation; (b) complete reaction mixture without Dap5; (c) without substrate 7; (d) without NADPH; and (e) with inactivated Dap5. Spectral acquisition and processing were performed using Bruker Xenon software; all spectra were baseline-corrected and normalized to the same receiver gain.

Supplementary Material

au6c01252_si_001.pdf (53MB, pdf)

Acknowledgments

This work is supported by the National Natural Science Foundation of China (32370087 and 32070080) (grants to J.H.), the Major Science and Technology Special Project of Wuhan City (2025021102020384), the Open Funds of the State Key Laboratory of Agricultural Microbiology (AMLKF201703), and the Natural Science Foundation for Distinguished Young Scholars of Hubei Province of China (2018CFA069). We are grateful to the National Key Laboratory of Agricultural Microbiology at Huazhong Agricultural University for access to its HPLC and HR-ESI-TOF-MS facilities.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacsau.6c01252.

  • Reagents and materials; strains and culture conditions; chemical analysis; fermentation and isolation of compounds 4–8; overexpression and purification of proteins; strains and plasmids used in this study; primers used in this study; homologous BGCs harboring dap5-like gene from public databases; NMR data for compounds 4, 5, 7, 8, 10, 16, 18, and 19; sequence alignment; SDS-PAGE gel analyses; the Dap5 protein structure; HPLC analyses; ESI-HR-MS spectra; NMR spectra; and other supporting figures (PDF)

J.L.D. designed the research, performed experiments, and wrote the original draft; J.H. provided financial support; Y.D., X.X., B.B.H., Y.H., J.J.X., and Q.B.L. performed experiments; Y.Q.L., L.H., and X.S.H. analyzed the NMR data. All authors have reviewed and approved the final version of the manuscript.

The authors declare no competing financial interest.

Published as part of JACS Au special issue “Molecular Editing in Natural Products and Drug Molecules”.

References

  1. Laursen J. B., Nielsen J.. Phenazine natural products: biosynthesis, synthetic analogues, and biological activity. Chem. Rev. 2004;104(3):1663–1686. doi: 10.1021/cr020473j. [DOI] [PubMed] [Google Scholar]
  2. Guttenberger N., Blankenfeldt W., Breinbauer R.. Recent developments in the isolation, biological function, biosynthesis, and synthesis of phenazine natural products. Bioorg. Med. Chem. 2017;25(22):6149–6166. doi: 10.1016/j.bmc.2017.01.002. [DOI] [PubMed] [Google Scholar]
  3. Blankenfeldt W., Parsons J. F.. The structural biology of phenazine biosynthesis. Curr. Opin. Struct. Biol. 2014;29(29C):26–33. doi: 10.1016/j.sbi.2014.08.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. a Thanabalasingam D., Kumar N. S., Jayasinghe L., Fujimoto Y.. Endophytic Fungus Nigrospora oryzae from a medicinal plant Coccinia grandis, a high yielding new source of phenazine-1-carboxamide. Nat. Prod. Commun. 2015;10(10):1659–1660. [PubMed] [Google Scholar]; b Wu C., van Wezel G. P., Hae Choi Y.. Identification of novel endophenaside antibiotics produced by Kitasatospora sp. MBT66. J. Antibiot. (Tokyo) 2015;68(7):445–452. doi: 10.1038/ja.2015.14. [DOI] [PubMed] [Google Scholar]
  5. Shi Y. M., Brachmann A. O., Westphalen M. A., Neubacher N., Tobias N. J., Bode H. B.. Dual phenazine gene clusters enable diversification during biosynthesis. Nat. Chem. Biol. 2019;15(4):331–339. doi: 10.1038/s41589-019-0246-1. [DOI] [PubMed] [Google Scholar]
  6. Wang X., Abbas M., Zhang Y., Elshahawi S. I., Ponomareva L. V., Cui Z., Van Lanen S. G., Sajid I., Voss S. R., Shaaban K. A.. et al. Baraphenazines A-G, divergent fused phenazine-based metabolites from a Himalayan Streptomyces. J. Nat. Prod. 2019;82(6):1686–1693. doi: 10.1021/acs.jnatprod.9b00289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Abdelfattah M. S., Kazufumi T., Ishibashi M.. Izumiphenazines A-C: isolation and structure elucidation of phenazine derivatives from Streptomyces sp. IFM 11204. J. Nat. Prod. 2010;73(12):1999–2002. doi: 10.1021/np100400t. [DOI] [PubMed] [Google Scholar]
  8. Lee H. S., Kang J. S., Cho D. Y., Choi D. K., Shin H. J.. Isolation, Structure Determination, and Semisynthesis of Diphenazine Compounds from a Deep-Sea-Derived Strain of the Fungus Cystobasidium laryngis and Their Biological Activities. J. Nat. Prod. 2022;85(4):857–865. doi: 10.1021/acs.jnatprod.1c00985. [DOI] [PubMed] [Google Scholar]
  9. Ding Z. G., Li M. G., Ren J., Zhao J. Y., Huang R., Wang Q. Z., Cui X. L., Zhu H. J., Wen M. L.. Phenazinolins A-E: novel diphenazines from a tin mine tailings-derived Streptomyces species. Org. Biomol. Chem. 2011;9(8):2771–2776. doi: 10.1039/c1ob05044c. [DOI] [PubMed] [Google Scholar]
  10. Li Y., Han L., Rong H., Li L., Zhao L., Wu L., Xu L., Jiang Y., Huang X.. Diastaphenazine, a new dimeric phenazine from an endophytic Streptomyces diastaticus subsp. ardesiacus. J. Antibiot. (Tokyo) 2015;68(3):210–212. doi: 10.1038/ja.2014.124. [DOI] [PubMed] [Google Scholar]
  11. Huang W., Wan Y., Zhang S., Wang C., Zhang Z., Su H., Xiong P., Hou F.. Recent Advances in Phenazine Natural Products: Chemical Structures and Biological Activities. Molecules. 2024;29(19):4771. doi: 10.3390/molecules29194771. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Dong J., He B., Wang R., Zuo X., Zhan R., Hu L., Li Y., He J.. Characterization of the diastaphenazine/izumiphenazine C biosynthetic gene cluster from plant endophyte Streptomyces diastaticus W2. Microb. Biotechnol. 2022;15(4):1168–1177. doi: 10.1111/1751-7915.13909. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Zhuang Y., Yang F., Menon A., Song J. M., Espinoza R. V., Schultz P. J., Garner A. L., Tripathi A.. An ECD and NMR/DP4+ Computational Pipeline for Structure Revision and Elucidation of Diphenazine-Based Natural Products. J. Nat. Prod. 2023;86(7):1801–1814. doi: 10.1021/acs.jnatprod.3c00306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Kim S. W., Cha S.-S., Cho H.-S., Kim J.-S., Ha N.-C., Cho M.-J., Joo S., Kim K. K., Choi K. Y., Oh B.-H.. High-resolution crystal structures of Δ5–3-ketosteroid isomerase with and without a reaction intermediate analogue. Biochemistry. 1997;36(46):14030–14036. doi: 10.1021/bi971546+. [DOI] [PubMed] [Google Scholar]
  15. Lundqvist T., Rice J., Hodge C. N., Basarab G. S., Pierce J., Lindqvist Y.. Crystal structure of scytalone dehydratasea disease determinant of the rice pathogen, Magnaporthe grisea. Structure. 1994;2(10):937–944. doi: 10.1016/S0969-2126(94)00095-6. [DOI] [PubMed] [Google Scholar]
  16. Kallio P., Sultana A., Niemi J., Mäntsälä P., Schneider G.. Crystal structure of the polyketide cyclase AknH with bound substrate and product analogue: implications for catalytic mechanism and product stereoselectivity. J. Mol. Biol. 2006;357(1):210–220. doi: 10.1016/j.jmb.2005.12.064. [DOI] [PubMed] [Google Scholar]
  17. Ma K., Liu J., Huang Z., Wu M., Liu D., Ren J., Fan A., Lin W.. Three-dimensional structural alignment based discovery and molecular basis of AtoB, catalyzing linear tetracyclic formation. Chem. Sci. 2024;15(44):18490–18496. doi: 10.1039/D4SC05590J. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Huang T., Chang C.-Y., Lohman J. R., Rudolf J. D., Kim Y., Chang C., Yang D., Ma M., Yan X., Crnovcic I.. Crystal structure of SgcJ, an NTF2-like superfamily protein involved in biosynthesis of the nine-membered enediyne antitumor antibiotic C-1027. J. Antibiot. 2016;69(10):731–740. doi: 10.1038/ja.2016.88. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Römer A.. 1H NMR spectra of substituted phenazines. Magn. Reson. Chem. 1982;19:66–68. doi: 10.1002/mrc.1270190203. [DOI] [Google Scholar]
  20. Akabori H., Nakamura M.. 1, 6-Di-hydroxyphenazine, an antibiotic produced by Streptomyces thioluteus. J. Antibiot. (Tokyo) 1959;12(1):17–20. doi: 10.11554/antibioticsa.12.1_17. [DOI] [PubMed] [Google Scholar]
  21. Greenhagen B. T., Shi K., Robinson H., Gamage S., Bera A. K., Ladner J. E., Parsons J. F.. Crystal structure of the pyocyanin biosynthetic protein PhzS. Biochemistry. 2008;47(19):5281–5289. doi: 10.1021/bi702480t. [DOI] [PubMed] [Google Scholar]
  22. Zhao Y., Qian G., Ye Y., Wright S., Chen H., Shen Y., Liu F., Du L.. Heterocyclic aromatic N-Oxidation in the biosynthesis of phenazine antibiotics from Lysobacter antibioticus. Org. Lett. 2016;18(10):2495–2498. doi: 10.1021/acs.orglett.6b01089. [DOI] [PubMed] [Google Scholar]
  23. Sultana A., Kallio P., Jansson A., Wang J. S., Niemi J., Mäntsälä P., Schneider G.. Structure of the polyketide cyclase SnoaL reveals a novel mechanism for enzymatic aldol condensation. EMBO J. 2004;23(9):1911–1921. doi: 10.1038/sj.emboj.7600201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. He B.-B., Zhou T., Bu X.-L., Weng J.-Y., Xu J., Lin S., Zheng J.-T., Zhao Y.-L., Xu M.-J.. Enzymatic Pyran Formation Involved in Xiamenmycin Biosynthesis. ACS Catal. 2019;9(6):5391–5399. doi: 10.1021/acscatal.9b01034. [DOI] [Google Scholar]
  25. Zhang B., Wang K. B., Wang W., Wang X., Liu F., Zhu J., Shi J., Li L. Y., Han H., Xu K.. et al. Enzyme-catalysed [6 + 4] cycloadditions in the biosynthesis of natural products. Nature. 2019;568(7750):122–126. doi: 10.1038/s41586-019-1021-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. a Bush K., Henry P. R., Souser-Woehleke M., Trejo W. H., Slusarchyk D. S.. Phenacein–an angiotensin-converting enzyme inhibitor produced by a streptomycete. I. Taxonomy, fermentation and biological properties. J. Antibiot. (Tokyo) 1984;37(11):1308–1312. doi: 10.7164/antibiotics.37.1308. [DOI] [PubMed] [Google Scholar]; b Liu W. C., Parker W. L., Brandt S. S., Atwal K. S., Ruby E. P.. Phenacein - an angiotensin-converting enzyme inhibitor produced by a streptomycete. II. Isolation, structure determination and synthesis. J. Antibiot. 1984;37(11):1313–1319. doi: 10.7164/antibiotics.37.1313. [DOI] [PubMed] [Google Scholar]
  27. Yu Y. Y., Zhang Y., Peng L.. Investigating the interaction between Shewanella oneidensis and phenazine 1-carboxylic acid in the microbial electrochemical processes. Sci. Total Environ. 2022;838(Pt 3):156501. doi: 10.1016/j.scitotenv.2022.156501. [DOI] [PubMed] [Google Scholar]
  28. a Castellano A., Catteau J., Lablache-Combier A., Planckaert B., Allan G.. Mechanism of the photochemical addition-substitution reactions of six-membered aza aromatic compounds in hydrogen-containing solvents. Chem. Heterocycl. Compd. 1974;10(7):755–768. doi: 10.1007/BF00471348. [DOI] [Google Scholar]; b Suzuki S., Takeda T., Kuratsu M., Kozaki M., Sato K., Shiomi D., Takui T., Okada K.. Pyrene-dihydrophenazine bis (radical cation) in a singlet ground state. Org. Lett. 2009;11(13):2816–2818. doi: 10.1021/ol9009927. [DOI] [PubMed] [Google Scholar]
  29. Adams J., Johnson K., Matthews R., Benkovic S. J.. Effects of distal point-site mutations on the binding and catalysis of dihydrofolate reductase from Escherichia coli . Biochemistry. 1989;28(16):6611–6618. doi: 10.1021/bi00442a012. [DOI] [PubMed] [Google Scholar]
  30. Xia C., Hamdane D., Shen A. L., Choi V., Kasper C. B., Pearl N. M., Zhang H., Im S.-C., Waskell L., Kim J.-J. P.. Conformational changes of NADPH-cytochrome P450 oxidoreductase are essential for catalysis and cofactor binding. J. Biol. Chem. 2011;286(18):16246–16260. doi: 10.1074/jbc.M111.230532. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Toplak M., Saleem-Batcha R., Piel J., Teufel R.. Catalytic control of spiroketal formation in rubromycin polyketide biosynthesis. Angew. Chem. 2021;133(52):27166–27176. doi: 10.1002/ange.202109384. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kumar S., Stecher G., Tamura K.. MEGA7: Molecular Evolutionary Genetics Analysis Version 7.0 for Bigger Datasets. Mol. Biol. Evol. 2016;33(7):1870–1874. doi: 10.1093/molbev/msw054. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Jumper J., Evans R., Pritzel A., Green T., Figurnov M., Ronneberger O., Tunyasuvunakool K., Bates R., Žídek A., Potapenko A.. Highly accurate protein structure prediction with AlphaFold. Nature. 2021;596(7873):583–589. doi: 10.1038/s41586-021-03819-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Mirdita M., Schütze K., Moriwaki Y., Heo L., Ovchinnikov S., Steinegger M.. ColabFold: making protein folding accessible to all. Nat. Methods. 2022;19(6):679–682. doi: 10.1038/s41592-022-01488-1. [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

au6c01252_si_001.pdf (53MB, pdf)

Articles from JACS Au are provided here courtesy of American Chemical Society

RESOURCES