The γ-butyrolactone motif is found in many natural signaling molecules and other specialized metabolites. A prominent example is the potent aquatic phytotoxin cyanobacterin, which has a highly functionalized γ-butyrolactone core structure. The enzymatic machinery that assembles cyanobacterin and structurally related natural products (herein termed furanolides) has remained elusive for decades. Here, we elucidate the biosynthetic process of furanolide assembly. The cyanobacterin biosynthetic gene cluster was identified by targeted bioinformatic screening and validated by heterologous expression in Escherichia coli. Full functional evaluation of the recombinant key enzymes in vivo and in vitro, individually and in concert, provided in-depth mechanistic insights into a streamlined C–C bond-forming cascade that involves installation of compatible reactivity at seemingly unreactive Cα positions of amino acid precursors. Our work extends the biosynthetic and biocatalytic toolbox for γ-butyrolactone formation, provides a general paradigm for furanolide biosynthesis and sets the stage for their targeted discovery, biosynthetic engineering and enzymatic synthesis.
This is a preview of subscription content, access via your institution
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$29.99 / 30 days
cancel any time
Subscribe to this journal
Receive 12 print issues and online access
$259.00 per year
only $21.58 per issue
Rent or buy this article
Prices vary by article type
Prices may be subject to local taxes which are calculated during checkout
Sequence data that support the findings of this study are deposited in GenBank with the accession code BK059219. The sequenced genome of Tolypothrix sp. PCC 9009 is publicly available and was downloaded from the NCBI database using accession number ALWD00000000. All analytical data generated for the present study are available upon request to the corresponding authors.
Mason, C. P. et al. Isolation of chlorine-containing antibiotic from the freshwater cyanobacterium Scytonema hofmanni. Science 215, 400–402 (1982).
Gleason, F. K., Porwoll, J., Flippen-Anderson, J. L. & George, C. X-ray structure determination of the naturally occurring isomer of cyanobacterin. J. Org. Chem. 51, 1615–1616 (1986).
Pignatello, J. J. et al. Structure of the antibiotic cyanobacterin, a chlorine-containing γ-lactone from the freshwater cyanobacterium Scytonema hofmanni. J. Org. Chem. 48, 4035–4038 (1983).
Gleason, F. K. & Paulson, J. L. Site of action of the natural algicide, cyanobacterin, in the blue-green alga, Synechococcus sp. Arch. Microbiol. 138, 273–277 (1984).
Gleason, F. K. & Case, D. E. Activity of the natural algicide, cyanobacterin, on angiosperms. Plant Physiol. 80, 834–837 (1986).
Gleason, F. K., Case, D. E., Sipprell, K. D. & Magnuson, T. S. Effect of the natural algicide, cyanobacterin, on a herbicide-resistant mutant of Anacystis nidulans R2. Plant Sci. 46, 5–10 (1986).
Mallipudi, L. R. & Gleason, F. K. Characterization of a mutant of Anacystis nidulans r2 resistant to the natural herbicide, cyanobacterin. Plant Sci. 60, 149–154 (1989).
Negishi, E. & Kotora, M. Regio- and stereoselective synthesis of γ-alkylidenebutenolides and related compounds. Tetrahedron 53, 6707–6738 (1997).
Yang, X., Shimizu, Y., Steiner, J. R. & Clardy, J. Nostoclide I and II, extracellular metabolites from a symbiotic cyanobacterium, Nostoc sp., from the lichen Peltigera canina. Tetrahedron Lett. 34, 761–764 (1993).
Felder, S. et al. Salimyxins and enhygrolides: antibiotic, sponge-related metabolites from the obligate marine myxobacterium Enhygromyxa salina. ChemBioChem 14, 1363–1371 (2013).
Raju, R., Garcia, R. & Müller, R. Angiolactone, a new butyrolactone isolated from the terrestrial myxobacterium, Angiococcus sp. J. Antibiot. 67, 725–726 (2014).
van Pée, K.-H. & Patallo, E. P. Flavin-dependent halogenases involved in secondary metabolism in bacteria. Appl. Microbiol. Biotechnol. 70, 631–641 (2006).
Duell, E. R. et al. Direct pathway cloning of the sodorifen biosynthetic gene cluster and recombinant generation of its product in E. coli. Microb. Cell Fact. 18, 32 (2019).
Greunke, C. et al. Direct pathway cloning (DiPaC) to unlock natural product biosynthetic potential. Metab. Eng. 47, 334–345 (2018).
D’Agostino, P. M. & Gulder, T. A. M. Direct pathway cloning combined with sequence- and ligation-independent cloning for fast biosynthetic gene cluster refactoring and heterologous expression. ACS Syn. Biol. 7, 1702–1708 (2018).
Pfeifer, B. A., Admiraal, S. J., Gramajo, H., Cane, D. E. & Khosla, C. Biosynthesis of complex polyketides in a metabolically engineered strain of E. coli. Science 291, 1790–1792 (2001).
Jordan, F. Current mechanistic understanding of thiamin diphosphate-dependent enzymatic reactions. Nat. Prod. Rep. 20, 184–201 (2003).
Balskus, E. P. & Walsh, C. T. Investigating the initial steps in the biosynthesis of cyanobacterial sunscreen scytonemin. J. Am. Chem. Soc. 130, 15260–15261 (2008).
Louie, G. V. et al. Structural determinants and modulation of substrate specificity in phenylalanine–tyrosine ammonia-lyases. Chem. Biol. 13, 1327–1338 (2006).
Moffitt, M. C. et al. Discovery of two cyanobacterial phenylalanine ammonia lyases: kinetic and structural characterization. Biochemistry 46, 1004–1012 (2007).
Balskus, E. P. & Walsh, C. T. An enzymatic cyclopentyl[b]indole formation involved in scytonemin biosynthesis. J. Am. Chem. Soc. 131, 14648–14649 (2009).
Schieferdecker, S. et al. Biosynthesis of diverse antimicrobial and antiproliferative acyloins in anaerobic bacteria. ACS Chem. Biol. 14, 1490–1497 (2019).
Park, J.-S. et al. Identification and biosynthesis of new acyloins from the thermophilic bacterium Thermosporothrix hazakensis SK20-1T. ChemBioChem 15, 527–532 (2014).
Knobloch, K.-H. & Hahlbrock, K. 4-Coumarate: CoA ligase from cell suspension cultures of Petroselinum hortense Hoffm: partial purification, substrate specificity, and further properties. Arch. Biochem. Biophys. 184, 237–248 (1977).
Li, X., Bonawitz, N. D., Weng, J.-K. & Chapple, C. The growth reduction associated with repressed lignin biosynthesis in Arabidopsis thaliana is independent of flavonoids. Plant Cell 22, 1620–1632 (2010).
Joyce, S. A. et al. Bacterial biosynthesis of a multipotent stilbene. Angew Chem. Int. Ed. Engl. 47, 1942–1945 (2008).
Nofiani, R., Philmus, B., Nindita, Y. & Mahmud, T. 3-Ketoacyl-ACP synthase (KAS) III homologues and their roles in natural product biosynthesis. MedChemComm 10, 1517–1530 (2019).
Wei, Y. & Shi, M. Recent advances in organocatalytic asymmetric Morita–Baylis–Hillman/aza-Morita–Baylis–Hillman reactions. Chem. Rev. 113, 6659–6690 (2013).
Chandra Bharadwaj, K. Intramolecular Morita–Baylis–Hillman and Rauhut–Currier reactions. A catalytic and atom economic route for carbocycles and heterocycles. RSC Adv. 5, 75923–75946 (2015).
Basavaiah, D. & Naganaboina, R. T. The Baylis–Hillman reaction: a new continent in organic chemistry—our philosophy, vision and over three decades of research. New J. Chem. 42, 14036–14066 (2018).
Reetz, M. T., Mondière, R. & Carballeira, J. D. Enzyme promiscuity: first protein-catalyzed Morita–Baylis–Hillman reaction. Tetrahedron Lett. 48, 1679–1681 (2007).
López-Iglesias, M., Busto, E., Gotor, V. & Gotor-Fernández, V. Use of protease from Bacillus licheniformis as promiscuous catalyst for organic synthesis: applications in C–C and C–N bond formation reactions. Adv. Synth. Catal. 353, 2345–2353 (2011).
Jiang, L. & Yu, H. An example of enzymatic promiscuity: the Baylis–Hillman reaction catalyzed by a biotin esterase (BioH) from Escherichia coli. Biotechnol. Lett. 36, 99–103 (2014).
Tian, X., Zhang, S. & Zheng, L. First Novozym 435 lipase-catalyzed Morita–Baylis–Hillman reaction in the presence of amides. Enzym. Microb. Technol. 84, 32–40 (2016).
Bjelic, S. et al. Computational design of enone-binding proteins with catalytic activity for the Morita–Baylis–Hillman reaction. ACS Chem. Biol. 8, 749–757 (2013).
Shih, P. M. et al. Improving the coverage of the cyanobacterial phylum using diversity-driven genome sequencing. Proc. Natl Acad. Sci. USA 110, 1053–1058 (2013).
Kearse, M. et al. Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics 28, 1647–1649 (2012).
Naville, M., Ghuillot-Gaudeffroy, A., Marchais, A. & Gautheret, D. ARNold: a web tool for the prediction of Rho-independent transcription terminators. RNA Biol. 8, 11–13 (2011).
D’Agostino, P. M., Song, X., Neilan, B. A. & Moffitt, M. C. Proteogenomics of a saxitoxin-producing and non-toxic strain of Anabaena circinalis (cyanobacteria) in response to extracellular NaCl and phosphate depletion. Environ. Microbiol. 18, 461–476 (2016).
We thank C. Chapple (Purdue University) and P. Wang (Willow Biosciences) for providing the pET30a(+)::At4CL1 vector. We also thank L. Wolf and D. Weuster-Botz (Technical University of Munich) for technical support. P.M.D. thanks the Marie Skłodowska-Curie Actions Individual Fellowship (project ID 745435) for funding. C.J.S. and X.J. thank the DBU (Deutsche Bundesstiftung Umwelt; grant 20015/400) and the CSC (China Scholarship Council), respectively, for their PhD fellowships. Research in the T.A.M.G. (DFG GU 1233/1-1) and T.G. (DFG GU 1134/3-1) laboratories is generously funded by the German Research Foundation.
The authors declare no competing interests.
Peer review information
Nature Chemical Biology thanks Toshiyuki Wakimoto and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Biomass of Tolypothrix sp. PCC 9009 was collected by centrifugation and the pellet was extracted three times with diethyl ether. The extracted organic phase was desiccated in vacuo and resuspended in methanol for analysis by high-resolution LCMS. A) Total ion count (top) and extracted mass specific for cyanobacterin (m/z = 430.5–431.5, bottom). A single peak can be observed within the extracted mass range. B) Mass spectrum of peak at RT 5.7 min shows the identification of cyanobacterin (m/z calculated for [C23H24ClO6]+ ([M + H]+) 431.1256, found: 431.1257) with the characteristic chlorine isotopic pattern.
Extended Data Fig. 2 1H-NMR spectra and HPLC chromatograms for the analysis of isolated E-7 (upper) and full interconversion to the thermodynamically more stable isomer Z-7 (lower).
The NMR was recorded in MeOD. The upper 1H-NMR spectrum (blue) was recorded directly after extraction and purification and contains a mixture of both isomers. Repeated measuring of the 1H-NMR one day after isolation (the compound was kept in solution: MeOD) reveals full interconversion of E-7 to the thermodynamically more stable Z-7. The reinjection experiment (HPLC) likewise confirmed this interconversion.
Extended Data Fig. 3 MS and 13C-NMR analysis of precyanobacterin I (Z-7) isolated upon supplementation of the expression cultures with 13C5-l-valine.
A) Structure of precyanobacterin I (E-7) indicating the positions with 13C-enrichment derived from 13C5-l-valine (blue dots). B) MS analysis of E-7 in cyb heterologous expression extracts after feeding unlabelled l-valine (top) and 13C5-l-valine (bottom). The incorporation of 4 13C atoms derived from l-valine is evident by the +4 shift. C) Comparison of 13C-NMR spectra of precyanobacterin I (Z-7) without (top, black) and with 13C-labelling (bottom, blue) derived from 13C5-l-valine, with expansion of the signals of positions I-III with the expected 13C,13C-coupling pattern.
Extended Data Fig. 4 MS and 13C-NMR analysis of precyanobacterin I (Z-7) isolated upon supplementation of the expression cultures with 13C9-l-tyrosine.
A) Structure of precyanobacterin I (E-7) indicating the positions with 13C-enrichment derived from 13C9-l-tyrosine (green and red dots). B) MS analysis of E-7 in cyb heterologous expression extracts after feeding unlabelled l-tyrosine (top) and 13C9-l-tyrosine (bottom). The incorporation of 17 carbon atoms derived of 2 13C9-l-tyrosine is evident by the +8 (from first 13C9-l-tyrosine), +9 (from second 13C9-l-tyrosine), and +17 (double-labelling with two 13C9-l-tyrosine units) mass shifts. C) Comparison of 13C-NMR spectra of precyanobacterin I (Z-7) without (top, black) and with 13C-labelling derived from 13C9-l-tyrosine (bottom, red).
Extended Data Fig. 5 13C-NMR analysis to identify acyloin intermediate 9 utilizing labelling experiments with 13C9-l-tyrosine and 13C5-l-valine.
Acyloins such as 9 are notoriously difficult to isolate and characterize or to be detected by mass spectrometry9. Thus, to determine the structure of 9, labelled amino acid precursors were fed to expression cultures and the raw extract was analysed by 13C-NMR. Three separate experiments a-c were conducted, applying (a) both 13C9-l-tyrosine and 13C5-l-valine (top, black), (b) 13C9-l-tyrosine (middle, red), or 13C5-l-valine (bottom, blue). Comparison of expression cultures allowed assignment of the proposed decarboxylated acyloin intermediate 9.
In vitro assay of At4CL1. A) HPLC analysis of in vitro assay of At4CL1 which successfully converted 4-coumaric acid (black) to 4-coumaroyl-CoA (blue). Conversion was unsuccessful in the negative control (minus enzyme; blue). B) HR-LCMS analysis of in vitro conversion of 4-coumaric acid to 4-coumaroyl-CoA. Presence of 4-coumaroyl-CoA is confirmed by MS1 (top) and MS2 (bottom).
Extended Data Fig. 7 1H-NMR analysis of 1,4-hydride shift experiments enabled by one-pot enzymatic synthesis in D2O.
Compared to the control performed in H2O (top), the experiment in D2O lead to deuterium incorporation at three positions, at two of which by keto/enol-tautomerism (blue and green dots) and one by 1,4-deuterium-shift (yellow).
Extended Data Fig. 8 Evaluation of CybE substrate preference by exchange of the aliphatic (β) building block.
Individual assays with At4CL1, CybE, and CybF with the alternative aliphatic building blocks 3-methyl-2-oxopentanoic acid (derived from isoleucine) and 4-methyl-2-oxopentanoic acid were conducted in comparison to the cyanobacterin precursor 3-methyl-2-oxobutanoic acid. Strong preference of CybF towards 3-methyl-2-oxopentanoic acid leading to 8 is indicated by significantly higher amounts of product 8 when compared to 24 formation. In addition, competition assays with both alternative precursors lead to exclusive formation of 8. Depicted assays with substrates: 3-methyl-2-oxobutanoic acid (i); 3-methyl-2-oxopentanoic acid (ii); 4-methyl-2-oxopentanoic acid (iii). competition assay simultaneously using 500 µM 3-methyl-2-oxopentanoic acid and 500 µM with 4-methyl-2-oxopentanoic acid (iv). To exclude failed detection of S1 in assay (iv) due to potentially insufficient chromatographic separation of 8 and S1, samples (iii) and (iv) were mixed in a 1:1 ratio and reanalyzed by HPLC (v). As can be seen (box) there is separation of 8 and 24 and hence the absence of the 24 peak in chromatogram (iv) is due to substrate preference of CybE for 3-methyl-2-oxopentanoic acid over with 4-methyl-2-oxopentanoic acid. Structures of all compounds are provided at the top of the figure.
Extended Data Fig. 9 Evaluation of CybE substrate preference by exchange of the aromatic (γ) building block.
Individual assays with At4CL1, CybE and CybF with the alternative building block phenylpyruvate (derived from phenylalanine) were conducted in comparison to the cyanobacterin precursor 4-hydroxphenylpyruvate. Strong preference of CybE towards 4-hydroxyphenylpyruvate is demonstrated by the significantly higher amount of product 7 when compared to 25 formation. In addition, competition experiments using both substrates almost exclusively lead to formation of 7 with only minute amounts of 25. These assays also confirm the proposed structure of compound * as 25 from E. coli heterologous extracts (see Fig. 2). Depicted assays with substrates: 4-hydroxyphenylpyruvate (i); phenylpyruvate (ii); competition assay using 500 µM hydroxyphenylpyruvate and 500 µM phenolpyruvate (iii). Structures of all compounds are provided at the top of the figure.
Extended Data Fig. 10 Evaluation of CybF substrate preference by exchange of the -CoA (α) building block and time course assay.
A) A one-pot reaction consisting of At4CL1, CybE and CybF was able to efficiently activate both cinnamic acid and 4-methoxycinnamic acid derived unnatural derivatives which as substituted at the α position (derived from CoA substrate). Control one-pot assay using 4-coumaric acid (i); One-pot assay with 4-methoxy cinnamic acid as the CoA substrate (ii); One-pot assay with cinnamic acid as the CoA substrate (iii). B) Time course assays performed by CybF from 0–240 minutes while utilizing various –CoA substrates such as coumaroyl-CoA (red), 4-methoxy cinnamoyl-CoA (green) and cinnamoyl-CoA (blue). Structures of all compounds are provided at the top of the figure. Replicates were performed (n = 4 biologically independent samples). Data are represented as mean values ± SD.
About this article
Cite this article
D’Agostino, P.M., Seel, C.J., Ji, X. et al. Biosynthesis of cyanobacterin, a paradigm for furanolide core structure assembly. Nat Chem Biol 18, 652–658 (2022). https://doi.org/10.1038/s41589-022-01013-7