Skip to main content
NCBI home page
Elsevier Sponsored Documents logoLink to Elsevier Sponsored Documents
. 2015 Jun 18;22(6):745–754. doi: 10.1016/j.chembiol.2015.05.010

Iterative Mechanism of Macrodiolide Formation in the Anticancer Compound Conglobatin

Yongjun Zhou 1, Annabel C Murphy 1, Markiyan Samborskyy 1, Patricia Prediger 2, Luiz Carlos Dias 3, Peter F Leadlay 1,∗
PMCID: PMC4504003  PMID: 26091168

Summary

Conglobatin is an unusual C2-symmetrical macrodiolide from the bacterium Streptomyces conglobatus with promising antitumor activity. Insights into the genes and enzymes that govern both the assembly-line production of the conglobatin polyketide and its dimerization are essential to allow rational alterations to be made to the conglobatin structure. We have used a rapid, direct in vitro cloning method to obtain the entire cluster on a 41-kbp fragment, encoding a modular polyketide synthase assembly line. The cloned cluster directs conglobatin biosynthesis in a heterologous host strain. Using a model substrate to mimic the conglobatin monomer, we also show that the conglobatin cyclase/thioesterase acts iteratively, ligating two monomers head-to-tail then re-binding the dimer product and cyclizing it. Incubation of two different monomers with the cyclase produces hybrid dimers and trimers, providing the first evidence that conglobatin analogs may in future become accessible through engineering of the polyketide synthase.

Highlights

  • •

    The conglobatin cluster has been cloned using a single-step in vitro procedure

  • •

    The gene cluster in a heterologous strain confers the ability to produce conglobatin

  • •

    A model for ATP-dependent heterocyclization to the oxazole ring is proposed

  • •

    The conglobatin thioesterase catalyzes cyclodimerization by an iterative mechanism

Zhou et al. describe the in vitro cloning from genomic DNA fragments of the intact biosynthetic gene cluster for conglobatin. This provides proof of concept for more convenient recovery of large gene clusters for both known and silent biosynthetic pathways, to enable their detailed analysis.

Introduction

Oxazole-containing polyketides are an intriguing family of natural products with diverse biological activities. Perhaps the best known are the methyloxazole-containing antitumor compound rhizoxin from “Burkholderia rhizoxina,” a symbiont of the fungus Rhizopus microsporus (Partida-Martinez and Hertweck, 2007), the oxazole triene antibiotic oxazolomycin A (Figure 1, 1) from Streptomyces albus, with potent cytotoxic and antibacterial activity (Mori et al., 1985), and the related herbicidal phthoxazolins/inthomycins (Omura et al., 1990; Henkel and Zeeck, 1991; Tanaka et al., 1993; Shiomi et al., 1995). Complex polyketides in bacteria are typically produced on polyketide synthases (PKSs), remarkable assembly-line multienzymes in which each cycle of chain extension is catalyzed by a different module of fatty acid synthase-related enzymatic activities. A similar logic is used in the biosynthesis of antibiotic peptides on non-ribosomal peptide synthetase multienzymes (NRPSs). The oxazole-containing polyketides are among the select group of natural products in which the biosynthetic assembly line contains both PKS and NRPS modules. A canonical PKS module minimally contains a β-ketoacyl synthase (KS) to catalyze C-C bond formation, an acyltransferase (AT) that recruits the appropriate extender unit, and an acylcarrier protein (ACP) domain. In an NRPS module, the condensation (C) domain forges the new amide bond, while the adenylation domain activates the extender amino acid and transfers it to the peptidyl carrier protein (PCP) domain. Optionally, a PKS module may also contain activities that reduce the initially formed β-ketothioester before it is handed on to the following module. These include ketoreductase (KR), dehydratase (DH), and enoylreductase (ER) domains. In many naturally occurring PKSs (including oxazolomycin), the AT activity is not an integrated domain in the multienzyme, but is present on a stand-alone protein which delivers extender units to every module (trans-AT PKSs). Many PKSs follow the assembly-line paradigm but show non-canonical placement and use of domains (non-colinearity), as well as catalyzing the incorporation of more exotic extender units, such as the gem-dimethyl group in oxazolomycin. Elucidating the molecular basis for these variations is an important step in gaining insight into the evolution of PKS systems, as well as laying the foundation for engineering the production of novel polyketides with potentially useful properties.

Figure 1.

Figure 1

Open in a new tab

The Structures of Conglobatin and Related Oxazole or Macrodiolide Natural Products

The 5-substituted oxazole moiety is also found in conglobatin (Figure 1, 2), a 16-membered macrocyclic diolide originally isolated from a polyether-producing strain of Streptomyces conglobatus ATCC 31005 (Westley et al., 1979) and reported at that time to be essentially devoid of antibiotic or antitumor activity. Recently, however, conglobatin (under the pseudonym FW-04-806) has been reinvestigated and reported to inhibit the proliferation of cancer cell lines, to cause G2/M cell-cycle arrest, to induce apoptosis, and to downregulate client oncoproteins of heat-shock protein Hsp90 (Huang et al., 2014). Its mode of action appears to be novel, via direct binding to the N-terminal domain of Hsp90 and disruption of its interaction with co-chaperone Cdc37 (Huang et al., 2015). The related 20-membered symmetrical oxazole macrodiolides known as samroiyotmycins (Figure 1, 3) (Dramae et al., 2013) are reported to be active antimalarial compounds. We have recently investigated the biosynthesis of the macrocyclic diolide elaiophylin (Figure 1) and have shown that the thioesterase of the elaiophylin polyketide synthase acts by an iterative mechanism, first ligating two monomers and then re-loading the linear dimer onto the thioesterase active site for dimerization (Zhou et al., 2015). This same mechanism has been previously demonstrated for the thioesterase/cyclase (TE) domain that catalyzes formation of typical non-ribosomal peptide synthetases (Shaw-Reid et al., 1999; Hoyer et al., 2007; Robbel et al., 2009). The X-ray crystal structures of chain-terminating TE domains from the PKS assembly lines have been determined for both macrocyclic polyketides (Tsai et al., 2002; Giraldes et al., 2006; Akey et al., 2006) and linear polyketides (Scaglione et al., 2010; Gehret et al., 2011), providing a valuable framework for mechanistic investigation. These enzymes belong to the α,β-hydrolase family, and the active site is located in the hydrophobic center of an unusual channel that traverses the entire protein. There are subtle differences in the size, shape, and accessibility of this channel in different structures, but it remains difficult to identify individual enzyme-substrate interactions that determine the outcome (Horsman et al., 2015). The ability of several individual polyketide TE domains to catalyze in vitro macrocyclization of thioester substrates has been demonstrated (Boddy et al., 2003; Wang et al., 2009; Pinto et al., 2012). Given the intriguing newly reported biological properties of conglobatin, we were interested in investigating its biosynthesis and defining the mechanism of dimerization in this pathway. We report here that the enzymology of formation of the oxazole ring in conglobatin closely resembles that for oxazolomycin, and that the conglobatin thioesterase follows an iterative mechanism, which has allowed its use in vitro to produce linear dimers and trimers in which the natural conglobatin seco acid monomer is combined with different polyketide building blocks.

Results and Discussion

Identification of the Conglobatin Biosynthetic Gene Cluster

Conglobatin is a symmetrical polyketide macrodiolide, and inspection of its structure suggests that its assembly is governed by an NRPS/PKS biosynthetic gene cluster. From the known organization of the NRPS/PKS for oxazolomycin (Zhao et al., 2010) we expected that initiation of conglobatin biosynthesis would involve an N-terminal loading module specific for glycine, followed by four PKS extension modules, and that termination would require a C-terminal chain-terminating TE domain. A high-quality draft genome sequence was generated in-house for the conglobatin-producing strain S. conglobatus ATCC 31005, and screened using as probe the amino acid sequence of the loading module protein of the oxazolomycin NRPS/PKS (OzmO). This search identified a unique locus in the S. conglobatus genome, and detailed in silico analysis of this region strongly supported its identification as the conglobatin gene cluster (Figure 2; Table S1). Experimental confirmation of involvement in conglobatin biosynthesis was obtained after transfer of the entire region into a heterologous host, as described below. We therefore refer to it as the cong locus.

Figure 2.

Figure 2

Open in a new tab

Gene Cluster and Proposed Pathway for Biosynthesis of Conglobatin

(A) Genetic organization of the conglobatin gene cluster (see also Table S1).

(B) Proposed biosynthetic pathway for conglobatin production in S. conglobatus (see the amino acid sequence alignment of each functional domain in Figure S2).

(C) Proposed process of 5-methyloxazole formation on the Cong-PKS, catalyzed by putative cyclodehydratase CongE. Phosphorylation of the amide backbone oxygen (mechanism 1) has been previously shown to promote thiazole/oxazole ring formation in microcin natural products (Dunbar et al., 2014). CongE may instead activate the amide oxygen by adenyltransfer from ATP with release of pyrophosphate. The timing of cyclodehydration with respect to the first polyketide elongation step remains to be established. In alternative mechanism 2, ATP is proposed to activate the other keto group for ring closure.

Sequence Analysis of the Conglobatin Biosynthetic Gene Cluster

Bioinformatic analysis of the cong locus revealed five contiguous open reading frames, spanning 28 kbp, whose gene products could be assigned putative roles in conglobatin biosynthesis. They encode three canonical modular PKSs (CongB, CongC, and CongD), an NRPS module (CongA), and a protein (CongE) of unknown function but with very significant sequence identity (80%) to OzmP from the oxazolomycin biosynthetic gene cluster. These genes appear to be transcribed from a single bidirectional promoter located between congA and congC. Tellingly, congA, congE, and congB are arranged together in exactly the same way as their counterparts ozmO, ozmP, and ozmQ in the oxazolomycin gene cluster. The precise limits of the cluster have not been rigorously defined, but a potential regulatory gene (orfR1) may provide one flank of the gene cluster, while the other may be signaled by orf1, encoding nicotinamide adenine dinucleotide phosphate-dependent glutamate dehydrogenase (Figure 3B; Table S1). However, among orfs 2–10 there are several other putative regulatory genes (Table S1) which may have roles in conglobatin biosynthesis.

Figure 3.

Figure 3

Open in a new tab

One-Step Cloning and Heterologous Expression of the Conglobatin Gene Cluster

(A) A 41-kbp XhoI-EcoRI DNA fragment (black) containing the five genes congA–E is generated by XhoI and EcoRI digestion of total genomic DNA. A 5.3-kbp pSET152 fragment was obtained by PCR amplification using as template the pSET152-derived plasmid pIB139. The resulting linear vector fragment had 39- and 41-bp flanking regions, respectively, identical to the termini of the target DNA (see Experimental Procedures).

(B) Gibson assembly leads to specific cloning of the target fragment, to give the bifunctional E. coli-Streptomyces plasmid pYJ24. The deduced open reading frame functions in the fragment are given in Table S1.

(C) Heterologous expression in S. coelicolor M1154 is confirmed by HPLC-MS and comparison with authentic compound produced by S. conglobatus (see also Figure S1). The mass extraction of m/z 499–500 is used to display the data. The y axis scale of S. conglobatus is 20 times larger than that of pYJ24/M1154 or M1154.

One-Step Cloning of the Conglobatin Gene Cluster and Heterologous Expression of Conglobatin in Streptomyces coelicolor

A 41-kbp XhoI-EcoRI DNA fragment was identified that contains the five genes congA–E. This region includes orfR1 on one side and orf1 to orf10 on the other side (Figure 3B; Table S1). OrfR1 shows 46% sequence identity to FscRI, which has been shown to act as a positive regulator in the biosynthesis of the polyene FR-008 (Zhang et al., 2015). It also includes the three additional potential regulatory genes orf3, orf7, and orf10. Genomic DNA, extracted and purified from 3-day-old mycelia of S. conglobatus, was digested with XhoI and EcoRI (Figure 3A) and fragments smaller than 20 kbp were removed by electrophoresis. To obtain a 5.3-kbp fragment of pSET152 (Bierman et al., 1992) PCR amplification was used with as template the pSET152-derived plasmid pIB139 (Wilkinson et al., 2002). The resulting linear vector fragment had 39- and 41-bp flanking regions, respectively, which are identical with the termini of the 41-kbp target DNA. Vector DNA and the chromosomal fragments were mixed and used for Gibson assembly (Gibson et al., 2009) as described in Experimental Procedures, and the products used in transformation of Escherichia coli (Figure 3B). All 30 transformants tested were found to be positive by using colony PCR. The plasmid extracted from one such colony was confirmed by sequencing to have the desired insert, and was designated pYJ24.

Plasmid pYJ24 was introduced into Streptomyces coelicolor M1154 (Gomez-Escribano and Bibb, 2011) by conjugation. Transformants (pYJ24/M1154) were confirmed by PCR and sequencing. To test for conglobatin expression, TSBYplus medium (see Experimental Procedures) was used as both seed and production medium. After 4 days of growth, heterologous expression of conglobatin in S. coelicolor pYJ24/M1154 was detected using HPLC-MS (high-pressure liquid chromatography-linked mass spectrometry) analysis of extracts. The conglobatin had the same retention time and molecular ion ([M + H]+ = 499.4) as authentic conglobatin produced by S. conglobatus (Figures 3C and S1), and produced the same MS/MS fragments from this molecular ion, with m/z 453.2, 204.2, and 159.3. The production level in S. coelicolor was less than 1% of the level produced by the original strain, which suggests that regulation of conglobatin biosynthesis is far from optimal under these conditions. It may also indicate that additional S. conglobatus genes are required for efficient biosynthesis. Nevertheless, this finding provides important confirmation of the identity of the cloned cong locus. It also paves the way for future analysis and manipulation of the cluster in a convenient heterologous strain. The efficiency with which the cong-bearing genome fragment was cloned from genomic DNA is highly encouraging. It remains to be established whether this direct selection method can be adapted for more general use (and for higher throughput) in whole-cluster transfer into heterologous strains.

Proposed Model of Conglobatin Biosynthesis

Assembly of the conglobatin monomer on the cong NRPS-PKS appears to be for the most part “co-linear,” as indicated in Figure 2B. Chain initiation takes place on the multienzyme CongA, whose adenylation domain is predicted using NRPSpredictor2 (Röttig et al., 2011) to specifically activate the amino acid glycine. The formyltransferase (F) domain shows 23%, 38%, and 59% sequence identity, respectively, to F domains in the initiation modules of the linear gramicidin NRPS (Schönafinger et al., 2006), the PKS-NRPS for the macrodiolide rhizopodin (Pistorius and Müller, 2012), and the oxazolomycin PKS (Zhao et al., 2010). Marahiel’s group (Schönafinger et al., 2006) has demonstrated for linear gramicidin in vitro that the F domain acts, using either N10- or N5-formyltetrahydrofolate as the formyl group donor, on the initiator amino acid tethered by a thioester linkage to the PCP domain, and also that subsequent chain extension does not proceed without prior formylation. The conglobatin F domain is analogously proposed to act on the glycyl residue tethered to the PCP. The formylglycine thioester could conceivably undergo cyclization at this stage, or could first be transferred via KS1 to the apparently redundant ACP domain C terminus of KS1. Alternatively, cyclization may occur after condensation of formylglycine with malonyl-CoA within extension module 1, which is split between multienzymes CongB and CongC.

The mechanism of cyclodehydration to form an oxazole is not known but clearly resembles the equivalent process in oxazolomycin biosynthesis. In many previously studied examples of oxazole and thiazole formation on assembly-line systems, where an amino acid side chain is involved as a nucleophile, a modified C domain, referred to as a heterocyclization domain, catalyzes formation of the oxazoline or thiazoline, and an oxygenase domain catalyzes subsequent flavin-dependent oxidation to the aromatic ring (Roy et al., 1999; Schneider et al., 2003). Neither of these domains is present in the ozm or cong gene clusters. Instead, the enigmatic protein CongE (highly similar to OzmP in the oxazolomycin biosynthetic gene cluster) may be responsible for cyclodehydration. Both these proteins belong to the family of N-type ATP (pyro)phosphohydrolases and contain the signature motif SGGKDS for ATP binding (Bork and Koonin, 1994).

Modeling of CongE using Phyre2 (Kelley and Sternberg, 2009) indicates that it possesses the α-β-α domain fold of this phosphohydrolase family and identifies significant structural homology with enzymes using ATP to activate amides for nucleophilic attack, for example, tRNA thiouridine synthetase TtuA from Pyrococcus horikoshii (Nakagawa et al., 2013). This is intriguing because ATP-dependent cyclodehydratases of the (unrelated) YcaO family have recently been convincingly shown to be responsible for ring formation in the biosynthesis of thiazole/oxazole-modified microcins (Dunbar et al., 2014). It is therefore tempting to propose that CongE and OzmP might phosphorylate (or alternatively transfer AMP to) an amide carbonyl oxygen to promote oxazole formation (Figure 2C). A further mechanistic precedent for CongE to catalyze ATP-dependent cyclodehydration is the reaction catalyzed by PurM, which catalyzes the ATP-dependent heterocyclization of formylglycinamide ribonucleotide to aminoimidazole ribonucleotide as part of purine biosynthesis (Schrimsher et al., 1986; Dunbar et al., 2012). Biochemical studies to test these models for oxazole formation in conglobatin are under way.

One obvious difference between the ozm and cong NRPS-PKS multienzymes is that the Cong multienzyme is a canonical cis-AT system, rather than having the AT as a stand-alone protein operating in trans. All cong KS domains showed the presence of the essential active site cysteine and two histidine residues. Domain KS1 differs significantly from KS2, KS3, and KS4 (Figure S2D). The AT of extension module 1 is predicted to recruit malonyl-CoA to provide the extender unit, while AT2 and AT4 are predicted to select (2S)-methylmalonyl-CoA (Haydock et al., 1995; Yadav et al., 2003), in agreement with the observed structure of conglobatin. However the active site motif GXSXG (where serine is the catalytically essential residue) is replaced in AT3 by GQAVG, so this domain is evidently inactive (Figure S2A). Our working hypothesis is that the KS2 carries out two successive condensations (“stuttering”) before passing the polyketide chain to module 3 for reduction. Such a programmed iteration has previously been invoked for PKS modules of stigmatellin (Gaitatzis et al., 2002), aureothin (He and Hertweck, 2003), borrelidin (Olano et al., 2003), and crocacin (Müller et al., 2014) biosynthesis. If such a mechanism operates here, the second round of chain extension within module 2 must be followed by immediate transfer of the ketoacylthioester to module 3 because the KR of module 2 is predicted to produce the “wrong” configuration for subsequent dehydration and enoylreduction (see below). The cluster contains a DH domain in each of modules 3 and 4, both predicted to be active as required for the conglobatin structure. Likewise, the three KR domains in the PKS all contain the conserved active site residues expected of active enzymes. The stereochemistry of reduction by the KR in module 2 is predicted, from the presence of characteristic active site motifs (Caffrey, 2003; Reid et al., 2003; Keatinge-Clay, 2007), to give rise to a 2R,3S-acyl thioester intermediate, consistent with the observed conglobatin structure (Figure S2B), and similarly a characteristic motif (Kwan and Leadlay, 2010) in the ER domain of extension module 3 correctly predicts the formation of a 2R-acyl thioester intermediate as required for conglobatin (Figure S2C) (Schregenberger and Seebach, 1984).

Reconstitution of Conglobatin Macrodiolide Formation In Vitro

To study the selectivity and mechanism of the C-terminal TE domain of the Cong-PKS, we expressed this domain as a stand-alone protein in E. coli and purified it to homogeneity (Figure S3). The boundaries of the domain were selected on the basis of alignment with homologous TE domains of known structure (Figure S2E). The substrate for dimerization was obtained by alkaline hydrolysis of conglobatin to the monomeric seco acid 4a, and subsequent conversion of this into its N-acetylcysteaminyl (SNAC) thioester 4b (Figure 4A; Experimental Procedures). After 5 hr of incubation of the Cong-TE with 4b at 20°C, the reaction was stopped and the mixture was analyzed using HPLC-MS. Two new peaks were detected in the total ion current, one (at 20.0 min) identical to conglobatin 2 and a second (at 16.3 min) with a mass corresponding to that of the linear dimer as its SNAC thioester 5. In control assays without enzyme, neither 5 nor 2 was detected. When the macrocycle-forming thioesterase from the erythromycin PKS (DEBS-TE) was used instead of Cong-TE, hydrolysis to 4a was observed, not dimerization (Figure 4A).

Figure 4.

Figure 4

Open in a new tab

In Vitro Reconstitution of Conglobatin Macrodiolide Formation Catalyzed by Conglobatin Thioesterase

(A) Conglobatin thioesterase (Cong-TE) catalyzed homodimerization of thioester 4b into 5 and conglobatin 2, analyzed by HPLC-MS. In contrast, the erythromycin PKS thioesterase (DEBS-TE) hydrolyzed 4b to 4a. The structures of 2, 5, 4a, and 4b were determined by high-resolution (HR)-MS (Table S2), and 1D and 2D NMR analysis (see Experimental Procedures and Supplementary Information).

(B) Time course of production of 5 and 2 catalyzed by Cong-TE.

(C) Linear thioester dimer 5, re-incubated with fresh Cong-TE, is converted into conglobatin 2.

This scheme is used to summarize all the conversions. Protein gel is given in Figure S3.

The timing the formation of 5 and 2 was investigated by measuring the area of the extracted ion peak at 2-hr intervals over 12 hr of incubation. The production level of 2 increased continuously while that of 5 reached a plateau by 2 hr (Figure 4B), suggesting that 5 might be an intermediate in diolide formation. To verify this, 5 was isolated from the assay mixture and re-incubated with fresh Cong-TE for 5 hr. Analysis showed significant conversion of 5 into conglobatin 2 (Figure 4C). The Cong-TE assay was scaled up 300-fold and the structures of purified 5 and 2 were elucidated by using 1D and 2D nuclear magnetic resonance (NMR) and high-resolution MS (Table S2; NMR data are shown in Supplemental Information).

The chemical competence of 5 as an intermediate in conglobatin formation in vitro strongly supports an iterative mechanism for TE catalysis, as we recently proposed for biosynthesis of the symmetrical 16-membered diolide elaiophylin (Zhou et al., 2015). In this mechanism, the full-length conglobatin monomer attached to the TE active site is attacked by the distal hydroxy group of an identical monomer on the adjacent ACP. The linear dimer produced then re-acylates the TE active site, and finally macrocyclization is catalyzed (Figure 2B). For the elaiophylin TE (ElaTE) in vitro, it was shown that a pentaketide substrate analog 6 gave rise to a symmetrical 16-membered decaketide diolide analog of elaiolide (Zhou et al., 2015). More surprisingly the tetraketide analog 7, although not itself a substrate, when mixed with 6 and presented to the ElaTE, gave rise to an additional compound, an asymmetrical nonaketide diolide. These results are readily accommodated by the iterative mechanism, since the various acyl transfers to and from the TE may well have different substrate preferences. We were therefore interested in the present work to discover whether 6 or 7 (neither of which is itself a substrate for the Cong-TE) could similarly intervene in the formation of novel dimers, when mixed with the natural seco acid analog 4b.

Formation of Hybrid Polyketides Using Cong-TE In Vitro

Assays with mixtures of substrates were carried out using the same conditions as for reaction with 4b alone, except that incubation was for 12 hr. We expected that the analog of the natural substrate for Cong-TE would preferentially acylate the TE but that we might see “cross-coupled” dimers through the nucleophilic attack of 6 on the acyl-enzyme in competition with a second molecule of 4b. To our initial surprise, co-incubation of 4b with 6 produced, in addition to the expected 5 and 2, a new compound whose structure was determined as the SNAC thioester of a hybrid linear dimer 8 (Figures 5A and S4A). This was accompanied by the corresponding carboxylic acid 9. These cross-coupled products have clearly been formed by nucleophilic attack of 4b on the 6-derived acyl-TE intermediate. The hybrid molecule 8 is able to re-acylate the TE and undergoes hydrolysis to 9 in preference to macrocyclization. Surprisingly, a further product was identified as the linear trimeric molecule 10 (Figures 5A and S4C). To confirm the likely mechanism for production of 10, 5 was purified and co-incubated with Cong-TE in the presence of 6. As shown in Figure 5B, 5 was converted into a mixture of conglobatin 2 and trimer 10. The formation of 10 is evidently caused by nucleophilic attack of 5 on the 6-derived acyl-TE intermediate in competition with macrocyclization.

Figure 5.

Figure 5

Open in a new tab

Hybrid Polyketide Dimers and Trimers Generated by Cong-TE from Mixed Substrates

(A) Conglobatin thioesterase (Cong-TE) in the presence of thioesters 4b and 6 catalyzed homodimerization of 4b into 5 and conglobatin 2, and heterodimerization of 4b and 6 into both linear heterodimer 8 (and its hydrolysis product 9) and linear heterotrimer 10.

(B) Linear thioester dimer 5, re-incubated with fresh Cong-TE in the presence of 6, is converted into conglobatin 2 and into heterotrimer 10.

(C) Linear thioester dimer 5, re-incubated with fresh Cong-TE in the presence of 7, is converted into 2 and into heterotrimer 11 (and its hydrolysis product 12). Details of the MS/MS and HR-MS analysis are given in Figure S4 and Table S2, respectively.

We extended these findings by co-incubation of the linear homodimeric SNAC thioester 5 with the tetraketide 7. Analysis of this mixture after 12 hr showed formation of trimer thioester 11 as well as the product of thioester hydrolysis of this compound, 12 (Figures 5C and S4D). We conclude from these findings that the elaiophylin-related thioesters 6 and 7 are both reactive enough (and the Cong-TE flexible enough) to compete successfully with 4b for acylation of the Cong-TE. However, deacylation is selective, hinting at the presence of a specific binding site on the protein for the incoming nucleophile. Unable to exploit this, nucleophilic attack by water (hydrolysis) or a second molecule of 6 or 7 is slow. In contrast, the conglobatin precursor analogs 4b and 5 are effective in deacylation, leading to the products described in Figure 5. These results reveal the subtlety of the iterative mechanism operated by macrodiolide TE domains. Meanwhile, the availability of the conglobatin gene sequence will enable engineering of the pathway to produce non-natural conglobatins as potential antitumor compounds. Given their close structural similarity (Figure 1), the conglobatin pathway might also be engineered to provide analogs of the 20-membered antimalarial oxazole macrodiolides 3, known as samroiyotmycins.

Significance

The C-terminal chain-terminating thioesterase domains of modular polyketide synthase assembly lines have attracted great interest because they exert a decisive influence on whether polyketide products are released in linear or cyclic form, and on the mode of any cyclization. An understanding of their specificity is also central to the success of attempts to manipulate assembly-line polyketide synthases to synthesize altered products. The thioesterase domains that lead to formation of conglobatin, elaiophylin, and other C2-symmetrical macrocyclic dilactones are especially intriguing because they apparently bring about the head-to-tail dimerization of identical chains using a single thioesterase active site. We have shown that a recombinant thioesterase can catalyze dimerization to conglobatin in vitro, and that it does so by an iterative mechanism, as shown recently for elaiophylin. Further evidence for the iterative operation of the thioesterase has been provided by finding that the recombinant enzyme can combine two different monomers into hybrid linear dimers or even trimers, suggesting its use as a means of producing diverse complex polyketides from relatively simple building blocks. Meanwhile, the discovery of the cong gene cluster, its direct one-step cloning, and the successful demonstration of conglobatin production in a convenient heterologous host together pave the way to deconvolution of the detailed enzymology of oxazole ring formation on the polyketide synthase, and to engineered production of novel conglobatins as potential antitumor compounds. Finally, this work provides proof of concept for using DNA fragments bearing an intact target gene cluster for rapid in vitro capture by Gibson assembly and direct cloning in E. coli. Converted to high throughput, this could be a valuable part of a pipeline for identifying the products of silent or cryptic gene clusters by controlled expression in heterologous host strains.

Experimental Procedures

Bacterial Strains and Culture Conditions

S. conglobatus ATCC 31005 (Westley et al., 1979) was grown in TSBY medium (3% tryptone soy broth, 10.3% sucrose, 0.5% yeast extract) at 30°C and 200 rpm to produce mycelium for genomic DNA extraction. For conglobatin production, the seed medium was 3% soybean flour (the supernatant after first autoclaving was used), 5% glucose (autoclaved separately), 0.5% CaCO3, 5 mg/l CoCl2·6H2O, and 0.2% (v/v) anti-foam. For the production medium, CoCl2·6H2O was omitted. Fermentation was carried out by inoculating 50 ml medium in a 250-ml conical flask fitted with a metal spring, with 10% (v/v) of a 3-day seed culture, then incubating at 30°C, 200 rpm for 5 days.

E. coli DH10B was used for DNA manipulation. E. coli ET12567 (pUZ8002) was used for intergeneric conjugation. E. coli BL21 CodonPlus (DE3) and BL21 (DE3) plysS were used for protein expression.

DNA Manipulation

Restriction endonucleases and T4 DNA ligase were purchased from New England Biolabs. Chemicals were purchased from Sigma-Aldrich. Plasmid DNA was isolated from an overnight culture using the Plasmid Mini Kit I (Omega BioTek) according to the manufacturer's protocol. PCR amplification was carried out using Phusion High-Fidelity PCR Master Mix from New England Biolabs (for cloning), or BioMix Red from Bioline (for screening purposes). Genomic DNA isolated from 3-day mycelium was used as a template for PCR. DNA sequencing was carried out by the DNA Sequencing Facility in the Department of Biochemistry, University of Cambridge.

Plasmid Construction

The DNA fragment encoding Cong-TE was amplified from genomic DNA of S. conglobatus ATCC 31005 by PCR using oligonucleotides Cong-TE-S (5′-ATTATCATATGAGCACGGGCCTGTGCCGGCACCT-3′ [NdeI]) and Cong-TE-A (5′-ATTATCTCGAGCCGGCGGTCCGCCGGAGCGT-3′ [XhoI]). The PCR product was digested with NdeI and XhoI before introduction into the corresponding sites of pET29b(+). The resulting plasmid was designated pYJ41. General procedures for E. coli manipulation were carried out according to Sambrook and Russell (2001).

Single-Step Cloning of the Conglobatin Gene Cluster

S. conglobatus genomic DNA was digested with XhoI and EcoRI (FastDigest, Thermo Scientific) at 37°C for 3.5 hr. The digested DNA was fractionated by electrophoresis using 0.6% agarose gel, 40 V, for 15 hr. The gel fraction containing fragments larger than 20 kbp was recovered, the agarose was melted at 50°C, and 200 μl of phenol/chloroform mixture was added and mixed. After centrifugation, the supernatant was moved to another tube and mixed with 0.6 volumes of isopropanol. The mixture was incubated at −20°C for 10 min before precipitating the DNA by centrifugation for 10 min. After washing with 80% ethanol, the DNA was dissolved in 20 μl of water, giving a concentration of 67 ng/μl.

The vector DNA, a 5.3-kbp pSET152 fragment (Bierman et al., 1992), was obtained by PCR amplification using as template the pSET152-derived plasmid pIB139 (Wilkinson et al., 2002) linearized by NdeI and EcoRV, and oligonucleotides pSET152-cong-S (5′-GGCGGAGGCGGCGAGGTCGCGTCACCCGACGGCGGTGCCCAATTCCACACAACATACGAG-3′) and pSET152-cong-A (5′-TGCCGGACACTGGTGGATCATGCAGGACCCGGAAGGCAACGAACTTCTCGACAGACGTAGATC-3′), which contain 39 and 41 bp (underlined) and overlap respectively with the ends of the 41-kbp target DNA. The 5,271-bp PCR product was purified and concentrated to 202 ng/μl. For DNA assembly, 0.5 μl of vector DNA and 4.5 μl of genomic DNA were added to 15 μl of Gibson DNA assembly solution (Gibson et al., 2009). The reaction was carried out at 50°C for 1 hr, then 10 μl of the reaction was used for calcium-assisted transformation of DH10B. Around 100 transformants were obtained, 30 of which were confirmed as positive by using colony PCR with the primers ConConfir-S (5′-AGGACCTCACCACCTGGGAAAC-3′) and ConConfir-A (5′-TAGGTCCGCAGGGTCTGAGGCA-3′). The identity of the 509-bp PCR product was confirmed by sequencing.

Heterologous Expression of the Conglobatin Pathway

The resulting plasmid pYJ24 was introduced into S. coelicolor M1154 (Gomez-Escribano and Bibb, 2011) by conjugation. For selection of Streptomyces transformants, apramycin and nalidixic acid were used at concentrations of 25 μg/ml. Transformants (pYJ24/M1154) were grown up in TSBY medium supplemented with 25 μg/ml apramycin for genomic DNA extraction, PCR, and sequencing confirmation. To check conglobatin expression, TSBY plus medium (1 l of TSBY medium supplemented with 5 mM MgCl2 and 2 ml of trace element solution as used for R2YE medium [Kieser et al., 2000]) was used as both seed and production medium. For production, a 3-day-old seed culture (supplemented with 25 μg/ml apramycin) was inoculated into 10 volumes of fresh medium. After 4 days, compound 2 was extracted from 1 ml of broth by incubation with 0.5 ml of ethyl acetate for 30 min at 50°C. The organic phase was separated and the solvent was removed, the residue was dissolved in 100 μl of methanol, and 50 μl of this was subjected to LC-MS analysis.

Protein Expression and Purification

Plasmid pYJ41 was introduced into E. coli BL21 CodonPlus (DE3) for Cong-TE protein expression, and pKJW63 (Tran et al., 2008) was introduced into E. coli BL21 (DE3) plysS for DEBS-TE protein expression. A single colony was inoculated into 5 ml of LB medium containing 50 μg/ml kanamycin and grown overnight at 37°C. 2 ml of the overnight culture was inoculated into 1 l of LB medium containing 50 μg/ml kanamycin and incubated at 37°C, 200 rpm until A600 reached 0.6–0.8 before adding 200 μl of 1 M isopropyl-β-D-thiogalactopyranoside and incubating at 22°C for 15 hr to induce protein expression. Cells were pelleted at 11,325 × g for 5 min, resuspended in lysis buffer (50 mM Tris-HCl, 0.3 M NaCl [pH 7.2]), and lysed by sonication. The total lysate was centrifuged at 34,925 × g for 25 min, and the supernatant was passed through a 0.45-μm filter before loading onto a His-Bind affinity column (1 ml bed volume). The column was washed with 10 column volumes of lysis buffer. Bound proteins were eluted by stepwise increases in the concentration of imidazole (up to 500 mM). Cong-TE and DEBS-TE were eluted from their respective columns at imidazole concentrations of 100 and 80 mM. The proteins were concentrated and buffer was exchanged into 100 mM potassium phosphate buffer (pH 8.2) using Amicon Ultra-4 concentrators (Millipore) fitted with a filter of 10-kDa cutoff. The yield of Cong-TE and DEBS-TE was 6 mg/l and 0.5 mg/l, respectively. The purified proteins were analyzed by 4%–12% Bis-Tris Gel (Novex) SDS-PAGE. Protein concentrations were measured using a NanoDrop 1000 spectrophotometer.

Enzyme Assays

In vitro assays contained, in a total volume of 50 μl, 40 μM Cong-TE or DEBS-TE, 3 mM substrate, and 10% (v/v) DMSO 100 mM potassium phosphate (pH 8.2). After either 5 or 12 hr of incubation at 20°C, the reaction was stopped by adding 200 μl of acetonitrile. After centrifugation, 40 μl of supernatant was injected for HPLC-MS analysis. To accumulate a sufficient amount of 2 and 5 for NMR analysis, the assay was scaled up to 15-ml system with 50 μM Cong-TE, 6 mM substrate 4b, and 4 hr of incubation at 20°C.

HPLC-MS Procedures

HPLC-MS analysis was performed using an HPLC (Agilent Technologies 1200) coupled to a Thermo Fisher LTQ mass spectrometer fitted with an electrospray ionization (ESI) source. The HPLC was fitted with a Prodigy 5μ C18 column (4.6 × 250 mm, Phenomenex). A solvent system of acetonitrile and water both containing 0.1% formic acid (v/v) was used. Samples were eluted at a flow rate of 0.7 ml min−1 with a linear gradient of 40%–100% acetonitrile over 30 min. The mass spectrometer was run in positive ionization mode, scanning from m/z 200 to 2000 using a normalized collision energy of 35%. Preparative HPLC purification was performed on an Agilent Technologies 1200 apparatus, using a C18 column (100 Å, 250 × 21.20 mm, 10 μm; Phenomenex) at a flow rate of 15 ml min−1. Sample injection volume was 200 or 500 μl. ESI high-resolution MS was carried out on a Thermo Fisher Orbitrap with 30,000 resolution.

NMR Analysis

NMR spectra were recorded on a Bruker 500-MHz DCH Cryoprobe Spectrometer except for compound 5, which was recorded on a Bruker 500-MHz TCI Cryoprobe Spectrometer, and compound 4a, which was recorded on a Bruker 400-MHz Avance III HD Spectrometer. Chemical shifts are expressed in parts per million on the δ scale, referenced to CHCl3 at δH 7.26 (1H) and CHCl3 at δ C 77.0 (13C). Where 1H signals were obscured by other signals or contaminants, they were obtained from heteronuclear single quantum coherence experiments (indicated by “obs”). All spectra are provided in the Supplemental Information.

Isolation of Conglobatin

To obtain conglobatin 2, 1 l of broth was extracted three times with 200 ml of ethyl acetate. The combined extracts were evaporated to dryness, re-dissolved in 300 ml of methanol, and extracted twice with 100 ml of hexane. The methanol phase was then evaporated and the residue was dissolved in 5 ml of methanol for further purification by preparative HPLC using water and methanol as solvents, both containing 0.1% formic acid (v/v). The sample was eluted using a gradient of 50%–98% methanol over 30 min. Conglobatin-containing fractions were combined and evaporated to remove methanol. The remaining aqueous phase was then extracted three times with 0.5 volume of ethyl acetate, and evaporation of the ethyl acetate yielded 382 mg of conglobatin as a pale yellow solid. The structure of 2 was confirmed by 1D and 2D NMR (for this and other compounds discussed below, spectra and assignments are shown in the Supplemental Information).

Synthesis of Conglobatin Seco Acid 4a

To generate 4a, 191 mg of conglobatin was hydrolyzed in 200 ml of 1.2 M NaOH containing 40% methanol at 65°C, and stirred vigorously for 2 hr. The reaction mixture was extracted twice with 60 ml of diethyl ether/hexane (3:1, v/v). The aqueous phase was evaporated to remove methanol, the pH was adjusted to 3.0, and the aqueous phase was extracted three times with 60 ml of ethyl acetate to recover 4a. After evaporation of the ethyl acetate, the residue was dissolved in 5 ml of methanol and 4a was purified by preparative HPLC using water and acetonitrile as solvents, both containing 0.1% formic acid (v/v). Elution was carried out with a gradient of 30%–66% acetonitrile over 15 min. Fractions containing 4a were combined and evaporated to remove acetonitrile. The aqueous solution was then extracted three times with 40 ml of ethyl acetate. Evaporation of the ethyl acetate gave 105 mg of 4a as a pale yellow oil. The structure of 4a was confirmed by 1D and 2D NMR.

Synthesis of Conglobatin Seco Acid-SNAC 4b

To a solution of 4a (60 mg), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC-HCl, 48 mg) and dimethylaminopyridine (2.5 mg) in CH2Cl2 (5 ml) was added 65 μl of N-acetylcysteamine, and the mixture was stirred at room temperature for 1.5 hr. The reaction mixture was then diluted with 50 ml of CH2Cl2 and washed twice with 10 ml of 0.01 M HCl, then dried under reduced pressure. The crude product was re-dissolved in methanol and purified by preparative HPLC using the same conditions as used for 4a. Fractions containing 4b were combined and evaporated to remove acetonitrile. The resulting aqueous phase was then extracted three times with 0.5 volume ethyl acetate, and after removal of the ethyl acetate 40 mg of 4b was obtained as a pale yellow oil. The structure of 4b was confirmed by 1D and 2D NMR.

Purification of 2 and 5 from Cong-TE Assay Mixtures

Cong-TE protein was removed from the assay solution by adding 4 volumes of acetonitrile, then centrifuged. The supernatant was evaporated to remove acetonitrile, and the remaining aqueous phase was extracted three times with 0.5 volume of ethyl acetate. The organic phase was evaporated to dryness and dissolved in methanol for purification by preparative HPLC. Water and acetonitrile were used for elution with a gradient of 40%–98% acetonitrile over 30 min. Product-containing fractions were combined and evaporated to remove acetonitrile. The remaining aqueous solution was then extracted three times with 0.5 volume of ethyl acetate. Approximately 0.3 mg of 5 and 0.6 mg of 2 were obtained after removal of the organic extract. The identity of 2 was confirmed by comparison of 1D and 2D NMR data with data obtained for the authentic sample isolated from S. conglobatus. The structure of 5 was confirmed by 1D and 2D NMR.

Acknowledgments

We gratefully acknowledge BBSRC (project grant BB/J007250/1 to P.F.L.), the European Commission (Marie Curie Fellowship to Y.Z.), and the University of Cambridge (Herchel Smith Research Fellowship to A.C.M.), and Ms. Asha Boodhun (Department of Chemistry, University of Cambridge) for help in HR-MS analysis. L.C.D. acknowledges the support of Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP, Proc. 2012/04616-3 and 2012/02230-0). P.F.L. is an International Research Awardee of the Alexander von Humboldt Foundation.

Published: June 18, 2015

Footnotes

This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Supplemental Information includes two tables, four figures, and 18 NMR spectra and can be found with this article online at http://dx.doi.org/10.1016/j.chembiol.2015.05.010.

Accession Numbers

The sequence of the conglobatin biosynthetic gene cluster of S. conglobatus ATCC 31005 and the flanking regions has been submitted to the European Nucleotide Archive under accession number European Nucleotide Archive: LN_849060.

Supplemental Information

Document S1. Tables S1 and S2, Figures S1–S4, and NMR Data Assignment and the Spectra
mmc1.pdf (4MB, pdf)
Document S2. Article plus Supplemental Information
mmc2.pdf (5.3MB, pdf)

References

  1. Akey D.L., Kittendorf J.D., Giraldes J.W., Fecik R.A., Sherman D.H., Smith J.L. Structural basis for macrolactonization by the pikromycin thioesterase. Nat. Chem. Biol. 2006;2:537–542. doi: 10.1038/nchembio824. [DOI] [PubMed] [Google Scholar]
  2. Bierman M., Logan R., O'Brien K., Seno E.T., Rao R.N., Schoner B.E. Plasmid cloning vectors for the conjugal transfer of DNA from Escherichia coli to Streptomyces spp. Gene. 1992;116:43–49. doi: 10.1016/0378-1119(92)90627-2. [DOI] [PubMed] [Google Scholar]
  3. Boddy C.N., Scheider T.L., Hotta K., Walsh C.T., Khosla C. Epothilone C macrolactonization and hydrolysis are catalyzed by the isolated thioesterase domain of epothilone polyketide synthase. J. Am. Chem. Soc. 2003;125:3428–3429. doi: 10.1021/ja0298646. [DOI] [PubMed] [Google Scholar]
  4. Bork P., Koonin E.V. A P-loop-like motif in a widespread ATP pyrophosphatase domain: implications for the evolution of sequence motifs and enzyme activity. Proteins. 1994;20:347–355. doi: 10.1002/prot.340200407. [DOI] [PubMed] [Google Scholar]
  5. Caffrey P. Conserved amino acid residues correlating with ketoreductase stereospecificity in modular polyketide synthases. ChemBioChem. 2003;4:649–662. doi: 10.1002/cbic.200300581. [DOI] [PubMed] [Google Scholar]
  6. Dramae A., Nithithanasilp S., Choowong W., Rachtawee P., Prabpai S., Kongsaeree P., Pittayakhajonwut P. Antimalarial 20-membered macrolides from Streptomyces sp BCC33756. Tetrahedron. 2013;69:8205–8208. [Google Scholar]
  7. Dunbar K.L., Melby J.O., Mitchell D.A. YcaO domains utilize ATP to activate amide backbones during peptide cyclodehydrations. Nat. Chem. Biol. 2012;8:569–575. doi: 10.1038/nchembio.944. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dunbar K.L., Chekan J.R., Cox C.L., Burkhart B.J., Nair S.K., Mitchell D.A. Discovery of a new ATP-binding motif involved in peptidic azoline biosynthesis. Nat. Chem. Biol. 2014;10:823–829. doi: 10.1038/nchembio.1608. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Gaitatzis N., Silakowski B., Kunze B., Nordsiek G., Blöcker H., Höfle G., Müller R. The biosynthesis of the aromatic myxobacterial electron transport inhibitor stigmatellin is directed by a novel type of modular polyketide synthase. J. Biol. Chem. 2002;277:13082–13090. doi: 10.1074/jbc.M111738200. [DOI] [PubMed] [Google Scholar]
  10. Gehret J.J., Gu L., Gerwick W.H., Wipf P., Sherman D.H., Smith J.L. Terminal alkene formation by the thioesterase of curacin A biosynthesis: structure of a decarboxylating thioesterase. J. Biol. Chem. 2011;286:14445–14454. doi: 10.1074/jbc.M110.214635. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Gibson D.G., Young L., Chuang R.Y., Venter J.C., Hutchison C.A., 3rd, Smith H.O. Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat. Methods. 2009;6:343–345. doi: 10.1038/nmeth.1318. [DOI] [PubMed] [Google Scholar]
  12. Giraldes J.W., Akey D.L., Kittendorf J.D., Sherman D.H., Smith J.L., Fecik R.A. Structural and mechanistic insights into polyketide macrolactonization from polyketide-based affinity labels. Nat. Chem. Biol. 2006;2:531–536. doi: 10.1038/nchembio822. [DOI] [PubMed] [Google Scholar]
  13. Gomez-Escribano J.P., Bibb M.J. Engineering Streptomyces coelicolor for heterologous expression of secondary metabolite gene clusters. Microb. Biotechnol. 2011;4:207–215. doi: 10.1111/j.1751-7915.2010.00219.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Haydock S.F., Aparicio J.F., Molnár I., Schwecke T., Khaw L.E., König A., Marsden A.F., Galloway I.S., Staunton J., Leadlay P.F. Divergent sequence motifs correlated with the substrate specificity of (methyl)malonyl-CoA:acyl carrier protein transacylase domains in modular polyketide synthases. FEBS Lett. 1995;374:246–248. doi: 10.1016/0014-5793(95)01119-y. [DOI] [PubMed] [Google Scholar]
  15. He J., Hertweck C. Iteration as programmed event during polyketide assembly; molecular analysis of the aureothin biosynthesis gene cluster. Chem. Biol. 2003;10:1225–1232. doi: 10.1016/j.chembiol.2003.11.009. [DOI] [PubMed] [Google Scholar]
  16. Henkel T., Zeeck A. Sekundärstoffe aus dem chemischen Screening, 16. Inthomycine, neue Oxazol-triene aus Streptomyces sp. Liebigs Ann. Chem. 1991;1991:367–373. [Google Scholar]
  17. Horsman M.E., Hari T.P., Boddy C.N. Polyketide synthase and non-ribosomal peptide synthetase thioesterase selectivity: logic gate or a victim of fate? Nat. Prod. Rep. 2015 doi: 10.1039/c4np00148f. [Epub ahead of print] [DOI] [PubMed] [Google Scholar]
  18. Hoyer K.M., Mahlert C., Marahiel M.A. The iterative gramicidin S thioesterase catalyzes peptide ligation and cyclization. Chem. Biol. 2007;14:13–22. doi: 10.1016/j.chembiol.2006.10.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Huang W., Ye M., Zhang L.R., Wu Q.D., Zhang M., Xu J.H., Zheng W. FW-04-806 inhibits proliferation and induces apoptosis in human breast cancer cells by binding to N-terminus of Hsp90 and disrupting Hsp90-Cdc37 complex formation. Mol. Cancer. 2014;13:150–162. doi: 10.1186/1476-4598-13-150. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Huang W., Wu Q., Zhang M., Kong Y., Cao P., Zheng W., Xu J., Ye M. Novel Hsp90 inhibitor FW-04-806 displays potent antitumor effects in HER2-positive breast cancer cells as a single agent or in combination with lapatinib. Cancer Lett. 2015;356:862–871. doi: 10.1016/j.canlet.2014.10.040. [DOI] [PubMed] [Google Scholar]
  21. Keatinge-Clay A.T. A tylosin ketoreductase reveals how chirality is determined in polyketides. Chem. Biol. 2007;14:898–908. doi: 10.1016/j.chembiol.2007.07.009. [DOI] [PubMed] [Google Scholar]
  22. Kelley L.A., Sternberg M.J.E. Protein structure prediction on the web: a case study using the Phyre server. Nat. Protoc. 2009;4:363–371. doi: 10.1038/nprot.2009.2. [DOI] [PubMed] [Google Scholar]
  23. Kieser T., Bibb M.J., Buttner M.J., Chater K.F., Hopwood D.A. John Innes Foundation; 2000. Practical Streptomyces Genetics. [Google Scholar]
  24. Kwan D.H., Leadlay P.F. Mutagenesis of a modular polyketide synthase enoylreductase domain reveals insights into catalysis and stereospecificity. ACS Chem. Biol. 2010;59:829–838. doi: 10.1021/cb100175a. [DOI] [PubMed] [Google Scholar]
  25. Mori T., Takahashi K., Kashiwabara M., Uemura D., Katayama C., Iwadare S., Shizuri Y., Mitomo R., Nakano F., Matsuzaki A. Structure of oxazolomycin, a novel β-lactone antibiotic. Tetrahedron Lett. 1985;26:1073–1076. [Google Scholar]
  26. Müller S., Rachid S., Hoffmann T., Surup F., Volz C., Zaburannyi N., Müller R. Biosynthesis of crocacin involves an unusual hydrolytic release domain showing similarity to condensation domains. Chem. Biol. 2014;21:855–865. doi: 10.1016/j.chembiol.2014.05.012. [DOI] [PubMed] [Google Scholar]
  27. Nakagawa H., Kuratani M., Goto-Ito S., Ito T., Katsura K., Terada T., Shirouzu M., Sekine S., Shigi N., Yokoyama S. Crystallographic and mutational studies on the tRNA thiouridine synthetase TtuA. Proteins. 2013;81:1232–1244. doi: 10.1002/prot.24273. [DOI] [PubMed] [Google Scholar]
  28. Olano C., Wilkinson B., Moss S.J., Braña A.F., Méndez C., Leadlay P.F., Salas J.A. Evidence from engineered gene fusions for the repeated use of a module in a modular polyketide synthase. Chem. Commun. 2003;2003:2780–2782. doi: 10.1039/b310648a. [DOI] [PubMed] [Google Scholar]
  29. Omura S., Tanaka Y., Kanaya I., Shinose M., Takahashi Y. Phthoxazolin, a specific inhibitor of cellulose biosynthesis, produced by a strain of Streptomyces sp. J. Antibiot. 1990;43:1034–1036. doi: 10.7164/antibiotics.43.1034. [DOI] [PubMed] [Google Scholar]
  30. Partida-Martinez L.P., Hertweck C. A gene cluster encoding rhizoxin biosynthesis in “Burkholderia rhizoxina”, the bacterial symbiont of the fungus Rhizopus microsporus. ChemBioChem. 2007;8:41–45. doi: 10.1002/cbic.200600393. [DOI] [PubMed] [Google Scholar]
  31. Pinto A., Wang M., Horsman M., Boddy C.N. 6-deoxyerythronolide B synthase thioesterase-catalyzed macrocyclization is highly stereoselective. Org. Lett. 2012;14:2278–2281. doi: 10.1021/ol300707j. [DOI] [PubMed] [Google Scholar]
  32. Pistorius D., Müller R. Discovery of the rhizopodin biosynthetic gene cluster in Stigmatella aurantiaca Sg a15 by genome mining. ChemBioChem. 2012;13:416–426. doi: 10.1002/cbic.201100575. [DOI] [PubMed] [Google Scholar]
  33. Reid R., Piagentini M., Rodriguez E., Ashley G., Viswanathan N., Carney J., Santi D.V., Hutchinson C.R., McDaniel R. A model of structure and catalysis for ketoreductase domains in modular polyketide synthases. Biochemistry. 2003;42:72–79. doi: 10.1021/bi0268706. [DOI] [PubMed] [Google Scholar]
  34. Robbel L., Hoyer K.M., Marahiel M.A. TioS T-TE—a prototypical thioesterase responsible for cyclodimerization of the quinoline- and quinoxaline-type class of chromodepsipeptides. FEBS J. 2009;276:1641–1653. doi: 10.1111/j.1742-4658.2009.06897.x. [DOI] [PubMed] [Google Scholar]
  35. Röttig M., Medema M.H., Blin K., Weber T., Rausch C., Kohlbacher O. NRPSpredictor2—a web server for predicting NRPS adenylation domain specificity. Nucleic. Acids Res. 2011;39:W362–W367. doi: 10.1093/nar/gkr323. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Roy R.S., Gehring A.M., Milne J.C., Belshaw P.J., Walsh C.T. Thiazole and oxazole peptides: biosynthesis and molecular machinery. Nat. Prod. Rep. 1999;16:249–263. doi: 10.1039/a806930a. [DOI] [PubMed] [Google Scholar]
  37. Sambrook J., Russell D.W. Cold Spring Harbor Laboratory Press; 2001. Molecular Cloning: A Laboratory Manual. [Google Scholar]
  38. Scaglione J.B., Akey D.L., Sullivan R., Kittendorf J.D., Rath C.M., Kim E.S., Smith J.L., Sherman D.H. Biochemical and structural characterization of the tautomycetin thioesterase: analysis of a stereoselective polyketide hydrolase. Angew. Chem. Int. Ed. Engl. 2010;49:5726–5730. doi: 10.1002/anie.201000032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Schneider T.L., Shen B., Walsh C.T. Oxidase domains in epothilone and bleomycin biosynthesis: thiazoline to thiazole oxidation during chain elongation. Biochemistry. 2003;42:9722–9730. doi: 10.1021/bi034792w. [DOI] [PubMed] [Google Scholar]
  40. Schönafinger G., Schracke N., Linne U., Marahiel M.A. Formylation domain: an essential modifying enzyme for the nonribosomal biosynthesis of linear gramicidin. J. Am. Chem. Soc. 2006;128:7406–7407. doi: 10.1021/ja0611240. [DOI] [PubMed] [Google Scholar]
  41. Schregenberger C., Seebach D. Synthesis and determination of the absolute configuration of the macrodiolide (+)-conglobatin. Tetrahedron Lett. 1984;25:5881–5884. [Google Scholar]
  42. Schrimsher J.L., Schenmdel F.J., Stubbe J., Smith J.M. Purification and characterization of aminoimidazole ribonucleotide synthetase from Escherichia coli. Biochemistry. 1986;25:4366–4371. doi: 10.1021/bi00363a028. [DOI] [PubMed] [Google Scholar]
  43. Shaw-Reid C.A., Kelleher N.L., Losey H.C., Gehring A.M., Berg C., Walsh C.T. Assembly line enzymology by multimodular nonribosomal peptide synthetases: the thioesterase domain of E. coli EntF catalyzes both elongation and cyclolactonization. Chem. Biol. 1999;6:385–400. doi: 10.1016/S1074-5521(99)80050-7. [DOI] [PubMed] [Google Scholar]
  44. Shiomi K., Arai N., Shinose M., Takahashi Y., Yoshida H., Iwabuchi J., Tanaka Y., Omura S. New antibiotics phthoxazolins B, C and D produced by Streptomyces sp. KO-7888. J. Antibiot. 1995;48:714–719. doi: 10.7164/antibiotics.48.714. [DOI] [PubMed] [Google Scholar]
  45. Tanaka Y., Kanaya I., Shiomi K., Tanaka H., Omura S. Phthoxazolin A, a specific inhibitor of cellulose biosynthesis from microbial origin I. J. Antibiot. 1993;46:1214–1218. doi: 10.7164/antibiotics.46.1214. [DOI] [PubMed] [Google Scholar]
  46. Tran L., Tosin M., Spencer J.B., Leadlay P.F., Weissman K.J. Covalent linkage mediates communication between ACP and TE domains in modular polyketide synthases. ChemBioChem. 2008;9:905–915. doi: 10.1002/cbic.200700738. [DOI] [PubMed] [Google Scholar]
  47. Tsai S.-C., Lu H., Cane D.E., Khosla C., Stroud R.M. Insights into channel architecture and substrate specificity from crystal structures of two macrocycle-forming thioesterases of modular polyketide synthases. Biochemistry. 2002;41:12598–12606. doi: 10.1021/bi0260177. [DOI] [PubMed] [Google Scholar]
  48. Wang M., Zhou H., Wirz M., Tang Y., Boddy C.N. A thioesterase from an iterative fungal polyketide synthase shows macrocyclization and cross coupling activity and may play a role in controlling iterative cycling through product offloading. Biochemistry. 2009;48:6288–6290. doi: 10.1021/bi9009049. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Westley J.W., Liu C.M., Evans R.H., Blount J.F. Conglobatin, a novel macrolide dilactone from Streptomyces conglobatus ATCC 31005. J. Antibiot. 1979;32:874–877. doi: 10.7164/antibiotics.32.874. [DOI] [PubMed] [Google Scholar]
  50. Wilkinson C.J., Hughes-Thomas Z.A., Martin C.J., Böhm I., Mironenko T., Deacon M., Wheatcroft M., Wirtz G., Staunton J., Leadlay P.F. Increasing the efficiency of heterologous promoters in actinomycetes. J. Mol. Microbiol. Biotechnol. 2002;4:417–426. [PubMed] [Google Scholar]
  51. Yadav G., Gokhale R.S., Mohanty D. Computational approach for prediction of domain organization and substrate specificity of modular polyketide synthases. J. Mol. Biol. 2003;328:335–363. doi: 10.1016/s0022-2836(03)00232-8. [DOI] [PubMed] [Google Scholar]
  52. Zhang P., Zhao Z., Li H., Chen X.L., Deng Z., Bai L., Pang X. Production of the antibiotic FR-008/candicidin in Streptomyces sp. FR-008 is co-regulated by two regulators, FscRI and FscRIV, from different transcription factor families. Microbiology. 2015;161:539–552. doi: 10.1099/mic.0.000033. [DOI] [PubMed] [Google Scholar]
  53. Zhao C., Coughlin J.M., Ju J., Zhu D., Wendt-Pienkowski E., Zhou X., Wang Z., Shen B., Deng Z. Oxazolomycin biosynthesis in Streptomyces albus JA3453 featuring an “acyltransferase-less” type I polyketide synthase that incorporates two distinct extender units. J. Biol. Chem. 2010;285:20097–20108. doi: 10.1074/jbc.M109.090092. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Zhou Y., Prediger P., Dias L.C., Murphy A.C., Leadlay P.F. Macrodiolide formation by the thioesterase of a modular polyketide synthase. Angew. Chem. Int. Ed. Engl. 2015;54:5232–5235. doi: 10.1002/anie.201500401. [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

Document S1. Tables S1 and S2, Figures S1–S4, and NMR Data Assignment and the Spectra
mmc1.pdf (4MB, pdf)
Document S2. Article plus Supplemental Information
mmc2.pdf (5.3MB, pdf)

RESOURCES