Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Dec 1.
Published in final edited form as: Nat Chem. 2017 Jul 31;9(12):1235–1242. doi: 10.1038/nchem.2829

A unifying paradigm for naphthoquinone-based meroterpenoid (bio)synthesis

Zachary D Miles 1,†, Stefan Diethelm 1,†, Henry P Pepper 2,†, David M Huang 2, Jonathan H George 2,*, Bradley S Moore 1,3,*
PMCID: PMC5960991  NIHMSID: NIHMS967051  PMID: 29168495

Abstract

Bacterial meroterpenoids constitute an important class of natural products with diverse biological properties and therapeutic potential. The biosynthetic logic for their production is unknown and defies explanation via classical biochemical paradigms. A large subgroup of naphthoquinone-based meroterpenoids exhibits a substitution pattern of the polyketide-derived aromatic core that seemingly contradicts the established reactivity pattern of polyketide phenol nucleophiles and terpene diphosphate electrophiles. We report the discovery of a hitherto unprecedented enzyme-promoted α-hydroxyketone rearrangement catalysed by vanadium-dependent haloperoxidases to account for these discrepancies in the merochlorin and napyradiomycin class of meroterpenoid antibiotics, and we demonstrate that the α-hydroxyketone rearrangement is potentially a conserved biosynthetic reaction in this molecular class. The biosynthetic α-hydroxyketone rearrangement was applied in a concise total synthesis of naphthomevalin, a prominent member of the napyradiomycin meroterpenes, and sheds further light on the mechanism of this unifying enzymatic transformation.

Meroterpenoids are natural products of mixed biosynthetic origin that exhibit antimicrobial1–6, anticancer2,3,6,7 and antioxidant activities8,9. They are formed in nature by the regioselective addition of an electron-rich, polyketide-derived aromatic nucleophile to an electrophilic terpene diphosphate, a reaction catalysed by the ABBA prenyltransferase (PTase) class of enzymes10,11. Despite the impressive body of knowledge that pertains to terpenoid biosynthesis12, surprisingly little is known about the biosynthesis of these hybrid compounds. Moreover, meroterpenoids are ubiquitous across all domains of life and are a prevalent source of natural product structural diversity in fungi13. Among bacteria, biosynthetic pathways to secondary metabolites that incorporate terpene fragments are often difficult to unravel with established terpene biosynthetic concepts14–16.

Recently, a group of hybrid isoprenoid–polyketide natural products, which include naphthomevalin (1)17, napyradiomycin A1 (2)18,19, neomarinone (3)20, naphterpin (4)8,21 and merochlorins A, B and D (5–7)1,22, have attracted the attention of the biosynthetic community because of their unique molecular architectures (Fig. 1). As demonstrated previously by 13C-labelling studies20,21,23, these naphthoquinone-based meroterpenoids derive from a tetrahydroxynaphthalene (THN) (8) precursor formed through the action of a single polyketide synthase using the substrate malonyl-coenzyme A (ref. 24). Single or multiple prenylation events, in addition to further embellishment by a multitude of predominantly uncharacterized modifying enzymes, adorn the THN precursor and form the complex final products. Curiously, the most-abundant subgroup of THN-derived meroterpenes (Fig. 1b, class II) exhibit a C3-prenylation pattern, which contradicts the expected C2/C4-nucleophilic reactivity of the parent THN molecule, as the C3 position is not nucleophilic and not poised for such an alkylation reaction. As such, they clearly distinguish themselves from the C2- or C4-prenylated meroterpenes, which include neomarinone (3) and merochlorins A (5) and B (6) (Fig. 1a, class I). To explain prenylation at the non-nucleophilic C3 carbon of the THN core, the involvement of ‘promiscuous’ PTases has been postulated25–27. To the best of our knowledge, no literature precedence exists for prenylation at such a non-nucleophilic aromatic carbon by a PTase, and therefore the lack of a satisfying explanation called for further investigation.

Figure 1. Bacterial THN-derived meroterpenes.

Figure 1

Open in a new tab

a,b, THN-derived meroterpenoids can be divided into two classes based on C2 or C4 prenylation (class I (a)) or C3 prenylation (class II (b)). The majority of naphthoquinone-based terpenoids are categorized as class II. Naphthomevalin (1) and napyradiomycin A1 (2) share features of both class I and II, with terpene units at both C2 and C3. All of the compounds could originate from a C4-prenylated molecule, with an α-hydroxyketone rearrangement of the prenyl appendage from C4 to C3 accounting for the creation of class II (bottom).

Herein we report the discovery of a halogenation-induced α-hydroxyketone rearrangement to account for the enigmatic substitution pattern of the prevalent class II meroterpenoid natural products (Fig. 1, bottom). This unifying transformation occurs in the biosynthesis of both the merochlorin and napyradiomycin classes of compounds through the action of the vanadium-dependent chloroperoxidase (VCPO) Mcl24 (merochlorin) and a homologous enzyme NapH3 (napyradiomycin) that does not harbour the same halogenation ability, yet catalyses a mirrored α-hydroxyketone rearrangement. In addition, our investigations identified a PTase, NapT8 (napyradiomycin), responsible for the upstream biosynthetic reaction in the napyradiomycin biosynthetic pathway, which includes prenylation of a halogenated meroterpenoid precursor and further underlines the novelty of biosynthetic transformations within bacterial meroterpenoid biosynthetic pathways.

In a joint effort that involves synthetic chemistry and biochemistry, we set out to exploit the unexpected discovery of this novel enzymatic rearrangement to develop a biosynthesis-inspired synthetic entry into the naphthomevalin class of meroterpenoids. As such, the presented synthesis is designed to follow a blueprint of enzymatic reactions, and is therefore in the strictest sense biomimetic. Moreover, our synthetic endeavour provides additional insight into the enzymatic transformation and lends credence to this rearrangement. We demonstrate that the amalgamation of biochemical investigations into an enzymatic reaction and its cognate synthetic application harbours great potential to address complex questions and advance the fields of both biosynthetic and synthetic chemistry28,29.

Results

Characterization of an additional Mcl24 product

The merochlorins are a group of recently isolated bacterial meroterpenoid antibiotics that include both the C4-prenylated members merochlorin A (5) and B (6), as well as the C3-substituted analogues merochlorin C (not shown) and D (7)1. Produced by a single biosynthetic gene cluster, this system offers a unique opportunity to study the chemistry that leads to the cryptic C3 prenylation observed in merochlorin D (7). We recently reported that the VCPO Mcl24 mediates a chlorination-induced oxidative dearomatization followed by intramolecular cyclization(s) to convert pre-merochlorin (9) into the stereochemically complex natural products 5 and 6 in a single enzymatic reaction (Fig. 2)30,31. However, the formation of the C3-prenylated merochlorin analogues remained elusive. Remarkably, the only PTase in the mcl gene cluster (Mcl23) showed selectivity for the prenylation of THN (8) at C4 to produce 9 exclusively30. Therefore, the possible promiscuity during the prenylation event could be excluded in the biosynthesis of the merochlorins.

Figure 2. Mcl24-mediated reaction of pre-merochlorin (9).

Figure 2

Open in a new tab

a, Reversed-phase HPLC (254 nm) of the standard reaction of Mcl24 with 9 at pH 8.0, pH 6.5, excluding H2O2 and excluding enzyme. When assayed at a lower pH, the major product observed was merochlorin A (5). When assayed at a higher pH, the product distribution changed substantially, with 10 as the major product. Consistent with the reaction requirements of VHPO enzymes, the removal of hydrogen peroxide (or enzyme) resulted in no turnover. All of the products other than 10, including isochloro-merochlorin B (asterisk) and the uncharacterized degradation product of 9 at ~16.5 min, were reported in a previous study31. b, Production of the C3-prenylated THN analogue 10 through an α-hydroxyketone rearrangement of 9. c, The proposed mechanism involves a chlorination-induced enzymatic α-hydroxyketone rearrangement.

During our investigations aimed towards a detailed characterization of Mcl24, we observed that a minor, previously uncharacterized, product of the reported reaction increased in abundance under basic reaction conditions (Fig. 2a and Supplementary Fig. 1). A detailed structural analysis indicated that the newly formed product 10 was a C3-prenylated THN derivative that exhibited dichlorination at C2 (Fig. 2b). In addition, a circular dichroism (CD) spectrum was obtained of 10, and the pronounced features are indicative of an enantiopure chiral molecule, which demonstrates the enantiospecificity of Mcl24 (Supplementary Fig. 2). This finding suggests that 10 is formed through an α-hydroxyketone rearrangement, which accounts for the alkyl migration from C4 to C3 (Fig. 2c). We propose a mechanism wherein Mcl24 first chlorinates 9 at C2 and the C3 phenol to give intermediate 11. Subsequent oxidative dearomatization to forma benzylic carbocation intermediate 12, followed by intramolecular cyclization gives rise to 5 and 6, as reported previously (path a)31. A similar benzylic carbocation intermediate was also proposed in previous total syntheses of both merochlorin A and B (refs 32,33). Alternatively, trapping of the intermediate benzylic carbocation 12 by water gives α-hydroxyketone 13, which undergoes a second chlorination at C2 to give the geminal dichloride 14 (path b). The addition of water is supported by assays conducted in 18OH2, wherein product 10 showed the requisite m/z shift of +2 of the primary ionization peak (Supplementary Fig. 3). The subsequent α-hydroxyketone rearrangement of 14 produces the C3-substituted THN derivative 10, which shares the same carbon skeleton as merochlorin D (7). To the best of our knowledge, this is the first report of a halogenation-mediated α-hydroxyketone rearrangement in nature, and resembles a suggested α-hydroxyketone rearrangement in aurachin biosynthesis34. It is particularly remarkable that a single enzyme, Mcl24, encoded from a 41-open reading frame gene cluster, is responsible for generating a significant portion of the structural complexity and diversity of all of the merochlorin natural products.

Chemical chlorination of pre-merochlorin (9) and analogue 17

We previously established a chemical chlorination protocol using a combination of N-chlorosuccinimide (NCS) and i-Pr2NH to mimic the Mcl24 chemistry31. Under these conditions, pre-merochlorin (9) was converted into deschloromerochlorin A (15) and deschloromerochlorin B (16) (Fig. 3). However, we did not observe the formation of a C3-prenylated product resulting from an α-hydroxyketone rearrangement under these conditions. We were curious to probe the influence that a change in the electronic properties of the substrate might have on the rearrangement chemistry. Such a study might be of interest, as some of the proposed intermediates en route to a rearranged α-hydroxyketone product possess polar character and would hence be influenced by such electronic changes in the substrate. Consequently, we subjected the less electron-rich analogue 17 to our chemical chlorination protocol. Interestingly, besides the expected merochlorin A analogue 18, two additional products, 19 and 20, were isolated.

Figure 3. Chemical chlorination of pre-merochlorin (9) or analogue 17.

Figure 3

Open in a new tab

Pre-merochlorin (9) and its less electron-rich analogue 17 were subjected to our previously established NCS/i-Pr2NH chemical chlorination conditions. a, An alternative outcome was observed for the NCS/i-Pr2NH-mediated oxidative dearomatization of 9 and its analogue 17 (Supplementary Table 2 gives the product yields). Whereas 9 converted into the two deschloro-merochlorin analogues 15 and 16, chlorination of substrate 17 produced a mixture of products 18–20. Interestingly, no products that result from an α-hydroxyketone rearrangement were observed under these conditions. b, A possible mechanistic rationale for the formation of products 19 and 20 involves the initial formation of carbocation 22 through oxidation and chloride loss. Intermediate 22 can subsequently be trapped by chloride to provide product 19, or alternatively undergo water addition followed by a C–C bond cleavage to arrive at compound 20. NCS, N-chlorosuccinimide.

We propose that the initial dichlorination of 17 gives intermediate 21, which undergoes oxidative dearomatization to give benzylic carbocation 22, in analogy to the chemistry performed by Mcl24 (Fig. 2c). Cyclization of 22 affords the merochlorin A analogue 18. Alternatively, trapping of the carbocation intermediate 22 by chloride at C4, followed by chlorination of the resultant chloroenol at C2, would deliver product 19. Finally, the addition of water at C4 of carbocation 22 followed by C2 chlorination would give α-hydroxyketone 23, which does not undergo an α-hydroxyketone rearrangement such as that observed in the enzymatic reaction. Instead, intramolecular hemiacetalization to form 24, followed by fragmentation, gives rise to compound 20. It seems probable that the presence of excess base influences the reaction outcome by promoting hemiacetalization at the expense of the desired α-hydroxyketone rearrangement. Furthermore, the different solvent properties of the enzymatic versus the chemical chlorination can influence the reaction outcome, and either prevent or promote the rearrangement chemistry by the impact on the stability of polar intermediates on the reaction path. Our failure to mimic the natural α-hydroxyketone rearrangement in these initial experiments raised questions about the thermodynamic requirements for the enzymatic 1,2-alkyl shift and motivated further synthetic studies.

Biomimetic total synthesis of (±)-naphthomevalin

The novel enzymatic α-hydroxyketone rearrangement discovered in the merochlorin system offered an opportunity to develop a general synthetic entry into the class II meroterpenes (Fig. 1b). A strategy that takes advantage of a biomimetic α-hydroxyketone rearrangement seemed particularly attractive, as it would exploit the predisposed C4-nucleophilic reactivity of THN to build up a precursor molecule efficiently, which would then undergo alkyl migration to arrive at the required C3-substitution pattern. Moreover, despite their promising antibacterial activity, only three total syntheses of naphthomevalin/napyradiomycin meroterpenoids have been reported to date35–37.

As outlined in Fig. 4, our biomimetic synthesis of naphthomevalin (1) commenced with methyl (3,5-dimethoxyphenyl)acetate (25), which was converted into the methyl ketone 26 via Friedel–Crafts acylation38 with Ac2O followed by demethylation using AlCl3. Protection of 26 as a bis-methoxymethyl (MOM) ether followed by base-induced aromatization via a Dieckmann-type cyclization then gave the protected THN derivative 27. Selective C4 geranylation was achieved through a palladium-mediated allylation using ethyl geranyl carbonate (28) as the electrophile30,39. The resulting naphthol 29 was subsequently subjected to Pb(OAc)4-mediated oxidative dearomatization32,40 followed by in situ dichlorination of the intermediate enol at C2 using NCS to give dichlorodiketone 30. Selective removal of one chlorine substituent from C2 of 30 was achieved using a lithium diisopropylamide (LDA)-mediated reduction to give a monochloride that is protected from further dechlorination by enolization41. Subsequent cleavage of the acetate protecting group delivered enol 31, which was then prenylated at C2 to give α-hydroxyketone 32 as a single diastereomer because of the steric hindrance of one face of the enolate derived from 31 by the bulky geranyl substituent. Removal of the MOM protecting groups from 32 under mild acidic conditions then gave α-hydroxyketone 33, and thus set the stage for the key biomimetic α-hydroxyketone rearrangement. Initial attempts to rearrange 33 using acidic or basic conditions failed to give C3-geranylated naphthomevalin (1). However, simply heating 33 in toluene at 110 °C for 16 hours efficiently produced (±)-naphthomevalin (1) in quantitative yield. All of the spectroscopic data for 1 matched the reported data of the natural product (Supplementary Information). The relative configuration of 1, with a trans relationship between the C2-prenyl and C3-geranyl substituents, is supported by its facile conversion into the epoxide natural product A80915G under mild basic conditions17 (Supplementary Information). Further experiments showed that the conversion of 33 into 1 occurred at a significantly lower temperature (60 °C) when water was used as the solvent, and thus represented an unusual example of an ‘on-water’ catalysed α-hydroxyketone rearrangement42.

Figure 4. Biomimetic total synthesis of (±)-naphthomevalin (1) via a thermal α-hydroxyketone rearrangement.

Figure 4

Open in a new tab

This 11-step total synthesis of (±)-naphthomevalin was designed to incorporate several steps of its proposed biosynthesis, which include the C4 geranylation of a THN derivative followed later by an α-hydroxyketone rearrangement to shift the geranyl group to C3. In addition to the natural product, two putative biosynthetic intermediates were synthesized for use in enzymatic studies.

Our synthesis of naphthomevalin (1) is biomimetic as it mirrors several steps of the proposed biosynthesis, namely C4 geranylation, oxidative dearomatization, C2 chlorination, C2 prenylation and α-hydroxyketone rearrangement. Thus, in conjunction with the synthesis of 1, we were able to prepare additional proposed biosynthetic intermediates free of protecting groups, which were instrumental in later enzymatic studies (vide infra).

With a synthesis of naphthomevalin and various analogues in hand, we envisioned further characterization of the thermal α-hydroxyketone rearrangement from a theoretical perspective. As shown in Fig. 5, three α-hydroxyketone derivatives (33, 34 and 35) were subjected to the originally established reaction conditions (110 °C in toluene) for the 1,2-alkyl shift. For each reaction, we also calculated kinetic and thermodynamic parameters using quantum chemical methods (Supplementary Information gives the full details). The calculations support a concerted 1,2-suprafacial-shift mechanism for the α-hydroxyketone rearrangement of both 33 and the dichlorinated derivative 35 (which closely resembles the putative intermediate 14 in the Mcl24 reaction of 9) with an internal proton transfer. The rearrangement of 33 and 35 is thermodynamically favoured, in part by the formation of a more-conjugated, planar structure in products 1 and 36. In contrast, monochlorinated derivative 34 was unreactive, as it possesses a near planar bicyclic structure, which indicates full conjugation of the π system. Clearly, geminal disubstitution at C2 is necessary for the migration to occur. An alternative two-step mechanism that involves a retro-[2,3]-Wittig rearrangement followed by an aromatic Claisen rearrangement has been proposed for a similar α-hydroxyketone rearrangement recently observed in aurachin biosynthesis34,43.

Figure 5. Theoretical and experimental studies on the thermally induced α-hydroxyketone rearrangement.

Figure 5

Open in a new tab

a–c, Attempted 1,2-shift of three α-hydroxyketone substrates (33 (a), 35 (b) and 34 (c)), with ΔG‡soln and ΔG°soln denoting computed free energy changes from the reactant to the transition state and from the reactant to the product, respectively. Structures with and without the ‘‡’ symbol are the computed transition state and reactant geometries, respectively. These studies show that geminal disubstitution at C2 is required for the α-hydroxyketone rearrangement to be thermodynamically favoured. soln, solution.

Characterization of an α-hydroxyketone rearrangement in napyradiomycin biosynthesis

After unravelling the origin of C3 prenylation in merochlorin biosynthesis coupled with the successful application of the key α-hydroxyketone rearrangement in the total synthesis of naphthomevalin (1), it remained unclear if this rearrangement can, indeed, serve as a general rule for the C3-prenylation pattern in all naphthoquinone-derived meroterpenoids. We therefore set out to explore whether an analogous α-hydroxyketone rearrangement is involved in the biosynthesis of naphthomevalin (1) using the proposed biosynthetic intermediates generated in our biomimetic synthesis as substrates for enzymatic studies.

In 2007, we characterized the naphthomevalin-related napyradiomycin biosynthetic gene cluster44 and identified genes that encoded two PTases and three vanadium-dependent haloperoxidases (VHPOs). The VCPO NapH1 was subsequently shown to catalyse the stereoselective chlorination–cyclization of the dimethylallyl group of a methylated analogue of naphthomevalin (1)45. Therefore, we set out to investigate the function of the putative ABBA aromatic PTases10,11 encoded by napT8 and napT9. We surmised that these genes encode enzymes responsible for the sequential addition of the dimethylallyl and geranyl moieties to THN at the electron-rich C2 and C4 positions (and not, as originally suggested, at the electron-deficient C3), which results in an intermediate primed for an α-hydroxyketone rearrangement. Indeed, NapT8 was cloned, heterologously expressed, purified and found to catalyse the addition of dimethylallyl pyrophosphate (DMAPP) to the C2 position of 34 to generate 33 (Fig. 6a). Furthermore, NapT8 was only active when the isoprenoid DMAPP was utilized and activity was stimulated by the presence of exogenous MgCl2 (Supplementary Fig. 7). To the best of our knowledge, this represents the first example of a PTase that catalyses the prenylation of a halogenated carbon centre.

Figure 6. NapT8-, NapH3- and NapH1-coupled assays with synthetic 34.

Figure 6

Open in a new tab

a, Reversed-phase HPLC chromatogram (254 nm) of the standard reaction of racemic 34 with: I, no enzyme (control); II, NapT8; IV, NapT8 then NapH3; VI, NapT8, then NapH3, then NapH1; and in addition to standard compounds: III, 33; V, naphthomevalin (1); VII, napyradiomycin A1 (2) (Supplementary Figs 4–6 give the complete ultraviolet/visible light and mass spectrometry characterizations and Supplementary Information gives the assay details). As evidenced through the in vitro assays on isolated reaction products, the sequential actions of NapT8, NapH3 and NapH1 are required for the formation of napyradiomycin A1 (2) from the synthetic substrate 34. b, Schematic of the pathway from 34 to napyradiomycin A1 (2). The pathway also proceeds through an α-hydroxyketone rearrangement of a substrate that possesses disubstitution at C2, which parallels the reaction catalysed by Mcl24.

We next turned our attention to the identification of an enzyme that could potentially catalyse an analogous α-hydroxyketone rearrangement of 33 to give naphthomevalin (1). We immediately focused on the orphan napyradiomycin VHPO homologues encoded by napH3 and napH4 that share about a 57% pairwise amino acid sequence identity with the previously characterized NapH1 (ref. 46). Although NapH4 proved recalcitrant to all of the attempts at recombinant expression, we were able to clone, heterologously express and purify NapH3. Recombinant NapH3 was subjected to activity assays with putative synthetic pre-napyradiomycin substrates (Supplementary Figs 8 and 9), along with the enzymatic product of the NapT8 reaction (33). Of all of the substrates evaluated, NapH3 exhibited an efficient capacity to mediate the C4-to-C3 α-hydroxyketone rearrangement of the geranyl moiety only on incubation with both synthetic and NapT8-produced 33, to form naphthomevalin (1) (Fig. 6a and Supplementary Fig. 8). Incubation of 33 with the VHPOs NapH1 and Mcl24, on the other hand, did not catalyse the formation of naphthomevalin (1) (Supplementary Fig. 10), although we could measure some slow, non-enzymatic conversion, as similarly observed in the synthetic conversion and NapT8 assays. To compare the non-enzymatic rate with that of the NapH3-catalysed reaction, the initial velocities under substrate-saturating conditions were measured (Supplementary Fig. 11). Under these conditions, NapH3 increased the rate nearly tenfold (0.374 ± 0.021 µM min−1 for the NapH3-catalysed reaction compared with 0.041 ± 0.002 µM min−1 for the non-enzymatic reaction) when assayed at a 10 µM enzyme concentration. The turnover number (kcat) for NapH3 is 0.038 ± 0.002 min−1, and remained constant when assayed with either synthetic or enantiopure (NapT8-produced) 33. To ascertain that naphthomevalin (1) is, indeed, a biosynthetic intermediate, we further subjected racemic 1 to NapH1 to yield the natural product napyradiomycin A1 (2) (Fig. 6a). To gain insight into the stereochemistry of the enzymatic reactions of NapT8 and NapH3, CD spectra were measured for the enzymatically produced 33 (Supplementary Fig. 12) and 1 (Supplementary Fig. 13). The CD spectrum of 33 formed from racemic 34 and NapT8 demonstrated the presence of an enantiopure chiral molecule, that is, a kinetic resolution had taken place. The CD spectrum of 1 generated by NapH3 catalysis on racemic 33 also indicated an enantiopure chiral molecule. Furthermore, unreacted 33 re-isolated from the NapH3 reaction has the opposite CD spectrum to that of 33 generated from prenylation of 34 with NapT8. These observations prove conclusively that the prenylation and α-hydroxyketone steps are under strict enzymatic stereocontrol, dictated by NapT8 and NapH3, respectively.

In contrast to the standard assay conditions required for the halogenation activity of VHPOs47, H2O2 or exogenous Na3VO4 were not required for NapH3 activity. Furthermore, this enzyme did not show chlorination or bromination activity when assayed with monochlorodimedone (MCD), a commonly used substrate to assess the capacity of VHPOs to form hypohalous acid48,49. The observation that NapH3 catalyses the α-hydroxyketone rearrangement on an already chlorinated/dearomatized intermediate further supports the lack of halogenation ability. A multiple-sequence alignment of the VHPO homologues from both the merochlorin and napyradiomycin gene clusters revealed a conserved serine residue that is present as a phenylalanine residue (F378) only in NapH3 (Supplementary Fig. 14). This residue was found to be critical for the turnover in NapH1, wherein the mutation to histidine abolished the native activity45. To test if this single mutation would be sufficient to confer halogenation ability, the NapH3 F378S variant was expressed, purified and assayed using the standard MCD assay; however, no activity was observed (data not shown).

When taken together, the general biosynthetic transformation for both pathways first employs chlorination-mediated oxidative dearomatization, followed by subsequent α-hydroxyketone rearrangement through a presumed acid/base catalysis. There are very few examples of a rearrangement of a long-chain carbon moiety in natural product biosynthesis; it has only been observed as a reaction that results in a shunt product in hapalindole/ambiguine biosynthesis29 and it is suggested in the biosynthesis of aurachin34,43. In merochlorin biosynthesis, both reactions are catalysed by a single enzyme, Mcl24. The increased production of 10 when the pH is increased from 6.5 to 8.0 implies a specific catalytic base residue with a pKa of ~7, such as a histidine residue. In napyradiomycin biosynthesis, NapH3 catalyses the ‘second half ’ of the Mcl24 reaction and presumably participates in a solely acid/base catalysis to promote this novel α-hydroxyketone rearrangement. Yet the requirement for disubstitution of C2 remains in both parallel enzymatic transformations, which is in line with our previous synthetic and theoretical observations. Studies are ongoing to identify the origin of the initial chlorination event in napyradiomycin biosynthesis.

Discussion and conclusions

The merochlorin and napyradiomycin families of natural products have proved valuable model biosynthetic pathways to address the long-standing question of how a counterintuitive substitution pattern is generated in the largest subclass of bacterial meroterpenoids. We identified two enzymes with a highly conserved primary sequence, capable of mediating an α-hydroxyketone rearrangement that leads to a non-standard prenyl substitution of the parent THN core. Remarkably, these enzymes invoke similar, yet in one instance disjointed, chemistry. In the biosynthesis of the merochlorins, the VHPO Mcl24 acts as a multitasking halogenating enzyme that can catalyse both the oxidative dearomatization of a simple precursor molecule and the movement of a prenyl appendage. In contrast, its homologue in the napyradiomycin biosynthetic pathway, NapH3, requires a C2-chlorinated substrate that is predisposed for the rearrangement. Despite the close relationship between the two VHPO homologues, NapH3 appears to have lost its capability as a haloperoxidase and has evolved into an exclusive and selective catalyst for the requisite rearrangement chemistry. Furthermore, we have identified the PTase NapT8, which is capable of prenylating a halogenated carbon centre on the THN core, and thereby primes the substrate for the α-hydroxyketone rearrangement catalysed by NapH3.

Through our synergistic approach of coordinating synthetic chemistry with biochemistry, we were able to discover and probe a novel biosynthetic reaction. Based on the presented results, we propose a new paradigm for the biosynthesis of an entire class of natural product compounds. We envision that all bacterial THN-derived C3-prenylated (class II) meroterpenoids proceed through an equivalent α-hydroxyketone rearrangement catalysed by a VHPO homologue, and thereby unify the class I and II meroterpenoid natural products.

Data availability

All data and any associated accession codes and references are available in the online version of this paper.

Supplementary Material

SI

Acknowledgments

We are grateful to our University of California San Diego colleagues B. Duggan for assistance with NMR measurements and X. Tang for helpful discussions. We are also grateful to M. Ghadiri and L. J. Leman at The Scripps Research Institute for their help in the collection of CD measurements. S.D. acknowledges the Swiss National Science Foundation for a postdoctoral fellowship. This research was supported by the US National Institutes of Health (R01-AI047818) and the Australian Research Council (DP160103393), and was undertaken with the assistance of resources from the National Computational Infrastructure, which is supported by the Australian Government.

Footnotes

Author contributions

Z.D.M., S.D., H.P.P., J.H.G. and B.S.M. designed the study and wrote the manuscript with input from all of the authors. Z.D.M., S.D. and H.P.P. performed the experiments. D.M.H. performed the computational studies.

Additional information

Supplementary information and chemical compound information are available in the online version of the paper.

Competing financial interests

The authors declare no competing financial interests.

References

  • 1.Kaysser L, et al. Merochlorins A–D, cyclic meroterpenoid antibiotics biosynthesized in divergent pathways with vanadium-dependent chloroperoxidases. J. Am. Chem. Soc. 2012;134:11988–11991. doi: 10.1021/ja305665f. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Wu Z, et al. Antibacterial and cytotoxic new napyradiomycins from the marine-derived Streptomyces sp. SCSIO 10428. Mar. Drugs. 2013;11:2113–2125. doi: 10.3390/md11062113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Cheng Y-B, Jensen PR, Fenical W. Cytotoxic and antimicrobial napyradiomycins from two marine-derived Streptomyces strains. Eur. J. Org. Chem. 2013;2013:3751–3757. doi: 10.1002/ejoc.201300349. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Shiomi K, et al. Novel antibiotics napyradiomycins. Production, isolation, physico-chemical properties and biological activity. J. Antibiot. 1986;39:487–493. doi: 10.7164/antibiotics.39.487. [DOI] [PubMed] [Google Scholar]
  • 5.Haste NM, et al. Bactericidal kinetics of marine-derived napyradiomycins against contemporary methicillin-resistant Staphylococcus aureus. Mar. Drugs. 2011;9:680–689. doi: 10.3390/md9040680. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Soria-Mercado IE, Prieto-Davó A, Jensen PR, Fenical W. Antibiotic terpenoid chloro-dihydroquinones from a new marine actinomycete. J. Nat. Prod. 2005;68:904–910. doi: 10.1021/np058011z. [DOI] [PubMed] [Google Scholar]
  • 7.Farnaes L, et al. Napyradiomycin derivatives, produced by a marine-derived actinomycete, illustrate cytotoxicity by induction of apoptosis. J. Nat. Prod. 2014;77:15–21. doi: 10.1021/np400466j. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Shin-ya K, et al. Isolation and structural elucidation of an antioxidative agent, naphterpin. J. Antibiot. 1990;43:444–447. doi: 10.7164/antibiotics.43.444. [DOI] [PubMed] [Google Scholar]
  • 9.Izumikawa M, Nagai A, Hashimoto J, Takagi M, Shin-ya K. Isolation of 2 new naphthablin analogs, JBIR-79 and JBIR-80, from Streptomyces sp. RI24. J. Antibiot. 2010;63:729–731. doi: 10.1038/ja.2010.120. [DOI] [PubMed] [Google Scholar]
  • 10.Heide L. Prenyl transfer to aromatic substrates: genetics and enzymology. Curr. Opin. Chem. Biol. 2009;13:171–179. doi: 10.1016/j.cbpa.2009.02.020. [DOI] [PubMed] [Google Scholar]
  • 11.Tello M, Kuzuyama T, Heide L, Noel JP, Richard SB. The ABBA family of aromatic prenyltransferases: broadening natural product diversity. Cell. Mol. Life Sci. 2008;65:1459–1463. doi: 10.1007/s00018-008-7579-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Bach TJ, Rohmer M. Isoprenoid Synthesis in Plants and Microorganisms: New Concepts and Experimental Approaches. Springer; 2012. [Google Scholar]
  • 13.Matsuda Y, Abe I. Biosynthesis of fungal meroterpenoids. Nat. Prod. Rep. 2015;33:26–53. doi: 10.1039/c5np00090d. [DOI] [PubMed] [Google Scholar]
  • 14.Xu Z, Baunach M, Ding L, Hertweck C. Bacterial synthesis of diverse indole terpene alkaloids by an unparalleled cyclization sequence. Angew. Chem. Int. Ed. 2012;51:10293–10297. doi: 10.1002/anie.201204087. [DOI] [PubMed] [Google Scholar]
  • 15.Yamada Y, et al. Terpene synthases are widely distributed in bacteria. Proc. Natl Acad. Sci. USA. 2015;112:857–862. doi: 10.1073/pnas.1422108112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Baunach M, Franke J, Hertweck C. Terpenoid biosynthesis off the beaten track: unconventional cyclases and their impact on biomimetic synthesis. Angew. Chem. Int. Ed. 2014;54:2604–2626. doi: 10.1002/anie.201407883. [DOI] [PubMed] [Google Scholar]
  • 17.Henkel T, Zeeck A. Secondary metabolites by chemical screening, 15. Structure and absolute configuration of naphthomevalin, a new dihydro-naphthoquinone antibiotic from Streptomyces sp. J. Antibiot. 1991;44:665–669. doi: 10.7164/antibiotics.44.665. [DOI] [PubMed] [Google Scholar]
  • 18.Shiomi K, et al. New antibiotic napyradiomycins A2 and B4 and stereochemistry of napyradiomycins. J. Antibiot. 1987;40:1213–1219. doi: 10.7164/antibiotics.40.1213. [DOI] [PubMed] [Google Scholar]
  • 19.Shiomi K, et al. Structures of new antibiotics: napyradiomycins. J. Antibiot. 1986;39:494–501. doi: 10.7164/antibiotics.39.494. [DOI] [PubMed] [Google Scholar]
  • 20.Kalaitzis JA, Hamano Y, Nilsen G, Moore BS. Biosynthesis and structural revision of neomarinone. Org. Lett. 2003;5:4449–4452. doi: 10.1021/ol035748b. [DOI] [PubMed] [Google Scholar]
  • 21.Shin-ya K, Furihata K, Hayakawa Y, Seto H. Biosynthetic studies of naphterpin, a terpenoid metabolite of Streptomyces. Tetrahedron Lett. 1990;31:6025–6026. [Google Scholar]
  • 22.Pepper H, George J. The biosynthesis and biomimetic synthesis of merochlorins A and B. Synlett. 2015;26:2485–2490. [Google Scholar]
  • 23.Funayama S, Ishibashi M, Komiyama K, Omura S. Biosynthesis of furaquinocins A and B. J. Org. Chem. 2002;55:1132–1133. [Google Scholar]
  • 24.Funa N, et al. A new pathway for polyketide synthesis in microorganisms. Nature. 1999;400:897–899. doi: 10.1038/23748. [DOI] [PubMed] [Google Scholar]
  • 25.Kuzuyama T, Noel JP, Richard SB. Structural basis for the promiscuous biosynthetic prenylation of aromatic natural products. Nature. 2005;435:983–987. doi: 10.1038/nature03668. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Kumano T, Tomita T, Nishiyama M, Kuzuyama T. Functional characterization of the promiscuous prenyltransferase responsible for furaquinocin biosynthesis: identification of a physiological polyketide substrate and its prenylated reaction products. J. Biol. Chem. 2010;285:39663–39671. doi: 10.1074/jbc.M110.153957. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Leipoldt F, et al. Diversity of ABBA prenyltransferases in marine Streptomyces sp. CNQ-509: promiscuous enzymes for the biosynthesis of mixed terpenoid compounds. PLoS ONE. 2015;10:e0143237. doi: 10.1371/journal.pone.0143237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Sydor PK, et al. Regio- and stereodivergent antibiotic oxidative carbocyclizations catalysed by Rieske oxygenase-like enzymes. Nat. Chem. 2011;3:388–392. doi: 10.1038/nchem.1024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Li S, et al. Hapalindole/ambiguine biogenesis is mediated by a Cope rearrangement, C–C bond-forming cascade. J. Am. Chem. Soc. 2015;137:15366–15369. doi: 10.1021/jacs.5b10136. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Teufel R, et al. One-pot enzymatic synthesis of merochlorin A and B. Angew. Chem. Int. Ed. 2014;53:11019–11022. doi: 10.1002/anie.201405694. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Diethelm S, Teufel R, Kaysser L, Moore BS. A multitasking vanadium-dependent chloroperoxidase as an inspiration for the chemical synthesis of the merochlorins. Angew. Chem. Int. Ed. 2014;53:11023–11026. doi: 10.1002/anie.201405696. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Pepper HP, George JH. Biomimetic total synthesis of (±)-merochlorin A. Angew. Chem. Int. Ed. 2013;52:12170–12173. doi: 10.1002/anie.201307200. [DOI] [PubMed] [Google Scholar]
  • 33.Meier R, Strych S, Trauner D. Biomimetic synthesis of (±)-merochlorin B. Org. Lett. 2014;16:2634–2637. doi: 10.1021/ol500800z. [DOI] [PubMed] [Google Scholar]
  • 34.Katsuyama Y, Harmrolfs K, Pistorius D, Li Y, Müller R. A semipinacol rearrangement directed by an enzymatic system featuring dual-function FAD-dependent monooxygenase. Angew. Chem. Int. Ed. 2012;51:9437–9440. doi: 10.1002/anie.201204138. [DOI] [PubMed] [Google Scholar]
  • 35.Tatsuta K, Tanaka Y, Kojima M, Ikegami H. The first total synthesis of (±)-napyradiomycin A1. Chem. Lett. 2002:14–14. [Google Scholar]
  • 36.Snyder SA, Tang Z-Y, Gupta R. Enantioselective total synthesis of (−)-napyradiomycin A1 via asymmetric chlorination of an isolated olefin. J. Am. Chem. Soc. 2009;131:5744–5745. doi: 10.1021/ja9014716. [DOI] [PubMed] [Google Scholar]
  • 37.Takemura S, et al. A concise total synthesis of (±)-A80915G, a member of the napyradiomycin family of antibiotics. Tetrahedron Lett. 1999;40:7501–7505. [Google Scholar]
  • 38.Beekman AM, Castillo Martinez E, Barrow RA. First syntheses of the biologically active fungal metabolites pestalotiopsones A, B, C and F. Org. Biomol. Chem. 2013;11:1109–1115. doi: 10.1039/c2ob26904j. [DOI] [PubMed] [Google Scholar]
  • 39.Kimura M, Fukasaka M, Tamaru Y. Palladium-catalyzed, triethylborane-promoted C-allylation of naphthols and benzene polyols by direct use of allyl alcohols. Synthesis. 2006;2006:3611–3616. [Google Scholar]
  • 40.Roche SP, Porco JA., Jr Dearomatization strategies in the synthesis of complex natural products. Angew. Chem. Int. Ed. 2011;50:4068–4093. doi: 10.1002/anie.201006017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Essa AH, et al. Reduction of 2,2,2-trichloro-1-arylethanones by RMgX: mechanistic investigation and the synthesis of substituted α,α-dichloroketones. Chem. Commun. 2013;49:2756–2758. doi: 10.1039/c3cc39147g. [DOI] [PubMed] [Google Scholar]
  • 42.Narayan S, et al. ‘On water’: unique reactivity of organic compounds in aqueous suspension. Angew. Chem. Int. Ed. 2005;44:3275–3279. doi: 10.1002/anie.200462883. [DOI] [PubMed] [Google Scholar]
  • 43.Katsuyama Y, Li X-W, Müller R, Nay B. Chemically unprecedented biocatalytic (AuaG) retro-[2,3]-Wittig rearrangement: a new insight into aurachin B biosynthesis. Chembiochem. 2014;15:2349–2352. doi: 10.1002/cbic.201402373. [DOI] [PubMed] [Google Scholar]
  • 44.Winter JM, et al. Molecular basis for chloronium-mediated meroterpene cyclization: cloning, sequencing, and heterologous expression of the napyradiomycin biosynthetic gene cluster. J. Biol. Chem. 2007;282:16362–16368. doi: 10.1074/jbc.M611046200. [DOI] [PubMed] [Google Scholar]
  • 45.Bernhardt P, Okino T, Winter JM, Miyanaga A, Moore BS. A stereoselective vanadium-dependent chloroperoxidase in bacterial antibiotic biosynthesis. J. Am. Chem. Soc. 2011;133:4268–4270. doi: 10.1021/ja201088k. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Winter JM, Moore BS. Exploring the chemistry and biology of vanadium-dependent haloperoxidases. J. Biol. Chem. 2009;284:18577–18581. doi: 10.1074/jbc.R109.001602. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Agarwal V, et al. Enzymatic halogenation and dehalogenation reactions: pervasive and mechanistically diverse. Chem. Rev. 2017;117:5619–5674. doi: 10.1021/acs.chemrev.6b00571. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Soedjak HS, Walker JV, Butler A. Inhibition and inactivation of vanadium bromoperoxidase by the substrate hydrogen peroxide and further mechanistic studies. Biochemistry. 1995;34:12689–12696. doi: 10.1021/bi00039a027. [DOI] [PubMed] [Google Scholar]
  • 49.de Boer E, Wever R. The reaction mechanism of the novel vanadium-bromoperoxidase. A steady-state kinetic analysis. J. Biol. Chem. 1988;263:12326–12332. [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

SI
NIHMS967051-supplement-SI.pdf (11.4MB, pdf)

Data Availability Statement

All data and any associated accession codes and references are available in the online version of this paper.

RESOURCES