Letter pubs.acs.org/OrgLett
Selected Mutations Revealed Intermediates and Key Precursors in the Biosynthesis of Polyketide−Terpenoid Hybrid Sesquiterpenyl Epoxy-cyclohexenoids Lin-Lin Teng,†,‡ Tian-Yang Song,† Zi-Fei Xu,† Xiao Liu,† Rong Dai,† Yong-Hong Chen,† Sheng-Hong Li,*,‡ Ke-Qin Zhang,*,† and Xue-Mei Niu*,† †
State Key Laboratory for Conservation and Utilization of Bio-Resources & Key Laboratory for Microbial Resources of the Ministry of Education, School of Life Sciences, Yunnan University, Kunming 650091, P. R. China ‡ State Key Laboratory of Phytochemistry and Plant Resources in West China, Kunming Institute of Botany, Chinese Academy of Sciences, Kunming 650204, P. R. China S Supporting Information *
ABSTRACT: Sesquiterpenyl epoxy-cyclohexenoids (SECs) show impressive biological activities. However, the key pathways for SECs still remain unambiguous. Unexpectedly, 11 new SECs and derivatives with diverse oxidation patterns were isolated after the deletion of gene 274. A high accumulation of toluquinol and its new glycosides in mutant Δ276 and further isolation of the most crucial precursors farnesyl hydroquinone, farnesyl quinone, and three new derivatives from mutant Δ278 confirm that farnesylation at toluquinol is the key step for SECs.
N
and contained three extra genes including YanG encoding farnesyltransferase (prenyltransferase). However, no biosynthetic intermediate has been obtained from mutational synthesis and heterologous expressions of yanuthone biosynthesis. Only two derivatives of 6-MSA, m-cresol and toluquinol, have been assumed to be involved in the biosynthesis of yanuthones through the combination of isotope-feeding experiments and bioinformatics.5 Recently, we identified a gene cluster AOL_s00215g containing the 6-MSA synthase gene 283 involved in the biosynthesis of morphological regulatory arthrosporols in Ar. oligospora (Figure 2).6 The cluster AOL_s00215g shared all the analogues (274, 276, 277, 279−283) of the eight biosynthetic genes in the Yan cluster and possessed one more CYP450 gene 278. Deletion of the YanB-like decarboxylase gene 281 and YanC-like P450 gene 282, respectively, resulted in high production of 6-MSA in mutant Δ281 and m-cresol in mutant Δ282.7 These findings suggest that Ar. oligospora is an alternative system to study the biogenesis of the SECs. Our latest mutational analysis of arthrosporol biosynthesis showed somewhat surprisingly that a class of three novel SEC derivatives 16−18 with novel carbon skeletons have been obtained from mutant Δ280.8 Interestingly, 16−18 shared the same oxygenated pattern around EC with ketone at C-4, which did not exist in all the known SECs from nematode-trapping fungi but often appeared in yunathones with ketone at C-4. The amino acid sequence of 280 shares 59% similarity to
ematode-trapping fungi are a unique group of soil fungi that can develop complex myclial traps to capture nematodes and play important roles in controlling nematode population density in diverse natural environments.1 A type of hybrid molecules with a polyketide (PK)-derived epoxycyclohexene nucleus and a terpenoid (TP)-derived sesquiterpenyl unit (Figure 1), including nematicidal flagranones (1), antibacterial oligosporons (2−3), and autoregulatory anthrobotrisins and arthrosporols (4−5), were obtained from nematode-trapping fungi Duddingtonia f lagrans and Arthrobotrys oligospora.2 They are among the well-known sesquiterpenyl epoxy-cyclohexenoids (SECs) including the potent antifungal yanuthones (6−11) from Aspergillus sp. and Penicillium sp. and notorious phytotoxic macrophorins (12−13) from the fungus Macrophoma causing fruit rot of apples.3 The characteristic epoxy-cyclohexene cores (EC) in SECs have long been regarded to be necessary for the biological activity of cyclohexane epoxides that have made these metabolites of considerable interest to biologists, pharmacologists, and synthetic chemists.3b The Yan gene cluster in the genome of As. niger containing the 6-methylsalicylic acid (6-MSA) PK synthase gene for biosynthesis of yanuthones (Figure 2) was the first of this class to be discovered4 since the model 6-MSA has been known to be the precursor to the notorious antibiotic and mycotoxin, patulin, produced by many Aspergillus, Penicillium, and Byssochlamys species via two important intermediates, cyclohexane epoxides, phyllostine (14) and isoepoxydon (15).5 The Yan gene cluster shared all the analogues (YanA−YanD and YanH) of the five biosynthetic genes for patulin biosynthesis © 2017 American Chemical Society
Received: June 16, 2017 Published: July 10, 2017 3923
DOI: 10.1021/acs.orglett.7b01846 Org. Lett. 2017, 19, 3923−3926
Letter
Organic Letters
Figure 3. HPLC profiles of the fermentations of WT (blue line), mutants Δ276 and Δ274 (red), arthrosporol A at around 11 min (green line), and targeted metabolites from the fermentation of the mutant Δ276 including toluquinol, C276−1 and C276−2 (cyan line).
toluquinol glycosides (C276−1 and C276−2). The sugar parts in C276−1 and C276−2 were elucidated as an α-6-deoxy-Lmannopyranosyl moiety by comparison of their NMR data (Table S2) with those of paganin B, a known metabolite from another carnivorous fungus Ar. entomopaga.9 The 1H−13C longrange correlations of the anomeric protons with C-1 and C-4 of the toluquinol moiety, respectively, in their HMBC spectra (Figures S15 and S22) finally determined the structures of C276−1 and C276−2 as shown in Figure 3. The results indicated that the suspectable peak in the HPLC profile was ascribed to C276−2 instead of arthrosporol A (5), and mutant Δ276 indeed abolished production of 5. Mutant Δ274 displayed so many weak peaks in the HPLC profile that the absence of 5 in mutant Δ274 seemed uncertain. Thus, mutant Δ274 strain was cultured on 100 L, and detailed investigation on the extract led to isolation of a number of intermediates C274−1−C274−11 (19−29). Their structures were fully determined by MS and NMR data (Tables S3−S6). Surprisingly, eight of them were new SECs, and three were new SEC derivatives including a 13-membered macrolide with a novel skeleton (Figure 4). The major metabolites 19 and 20 have quite similar structures to 3 from Ar. oligospora,3 and both lack the acetyl group in 3. Moreover, 20 also lacks one hydroxyl group at C-1′. The NOE correlations of H-4/H-6 in the ROESY spectrum of 19 and of H-4/H-1′ and H-6/H-1′ in that of 20 suggested that the hydroxyl groups at C-4 in 19−20 were both β configurations (Figures S30 and 36). Thus, 20 shares the similar features as 8 which has been assumed as the first hybrid precursor with the epoxy ring for yanuthones4 but has the ketone at C-1, while 8 has the ketone at C-4. Metabolite 21 has a similar EC to C280−1−C280−3 (16−18) from mutant Δ2808 and an identical side chain to 19. The ROESY correlations of H-1′/H-6 and H-1′/H-1 in 21 indicated that 1-OH was at the same side as the epoxy ring, consistent with the stereochemistry of 1-OH in oligosporol A and arthrobotrisin C.2 Metabolites 22−26 have similar structures to 19 but possess the ECs with OH at C-1 instead of a ketone as in 19. Comparison of the chemical shifts of H-1 and C-1 in 22−26 with those of two pairs of C-1 epimers, oligosporols B and A2a,b and arthrobotrisins B and C,2c together with lack of the key NOE correlations of H-1/H-1′ in their ROESY spectra suggested α configurations for 1-OH in 22−26. The OH at C4′ in 25 was elucidated as α-configuration on the basis of
Figure 1. Representative structures of SECs with four different oxygenated patterns in the epoxy-cyclohexene cores.
Figure 2. Schematic view of the organization of the 6-MSA gene containing biosynthetic gene clusters.
cytochrome P450 yanH, and a previous study reported that the gene yanH was involved in hydroxylating the methyl at C-2 of 7-deacetoxyyanuthone A (8) to yield 22-deacetoxyyanuthone A (9).4 Thus, the biogenesis for conversions from SECs with ketone at C-1 to those with ketone at C-4 has remained ambiguous. The genes YanD and YanG have been assumed to be involved in the key steps, epoxidation and farnesylation, respectively, for biosynthesis of the hybrid EC in SECs.4 The genes 274 and 276 in Ar. oligospora (Figure 2) were annotated for a dehydrogenase (38% similarity to YanD) and a polyprenyltransferase (50% similarity to YanG). Thus, two mutants Δ274 and Δ276 were constructed by genetic manipulation.6−8 Mutant Δ276 yielded a predominant peak at around 9 min and, unexpectedly, a peak with retention time suspectable for arthrosporol A (5) in the HPLC profile (Figure 3). Targeted isolation of three metabolites was carried out from the extract of 10 L of cultural broth of Δ276. Their structures were fully resolved by MS and NMR data as toluquinol and two new 3924
DOI: 10.1021/acs.orglett.7b01846 Org. Lett. 2017, 19, 3923−3926
Letter
Organic Letters
Reanalysis of the gene cluster AOL_s00215g for SECs in Ar. oligospora revealed that the protein of the extra gene 278 in the upstream flanking region of the gene 276 formed a distinct cluster with CYP450 proteins from P. camemberti and Pc12g06350 from P. rubens (Figure S3). Analysis of the extract of mutant Δ278 showed that production of 5 was completely abolished, and new metabolites were generated (Figure S8). Further analysis of 20 L of cultural broth of mutant Δ278 revealed five targeted metabolites (30−34), and detailed 1D and 2D NMR analysis (Table S7) confirmed that the major metabolite 30 (Figure 5) was just the first hybrid precursor
Figure 4. Novel metabolites isolated from extracts of mutant Ar. oligospora Δ274, including a 13-membered macrolide (29).
Figure 5. Key precursors and novel metabolites isolated from extracts of Ar. oligospora mutant Δ278.
similar NOE correlations and biogenesis to the known SECs with 4′-OH from Ar. oligospora. Metabolites 27−29 did not show the characteristic signals for epoxy methine carbons at δC 58−60 ppm in their 13C NMR spectra. Instead, three distinct oxygenated carbons, including two methines and a quaternary carbon, were observed at above 60 ppm, suggesting that the epoxy rings in these metabolites might be opened. This was confirmed by the presence of two OH signals in their 1H NMR spectra. The only difference between 27 and 28 came from H-6 and C-6 (δH 4.21 and δC 60.6 in 27; δH 4.60 and δC 68.8 in 28). The ROESY correlations of H-4/H-1′ and H-6/H-1′ in 28 (Figure S84) indicated that the OH at C-4, C-5, and C-6 was in the same side as β configurations in 28. The lack of ROESY correlation of H-6/H-1′ in 27 suggested that its 6-OH was assignable as α configuration. Metabolite 29 has a trinorfarnesyl side moiety (C12) linked to the cyclohexenol part. The significant 1H−13C long-range correlation of H-6 with the carbonyl carbon at δC 173 in the HMBC spectrum of 29 indicated the occurrence of an oxygen linkage between C-10′ and C-6, thus forming a 13membered macrolide. The ROESY correlations of H-4/H-1′, H-6/H-1′, and 4-OH/5-OH in 29 indicated the β-orientations of the oxygen-bearing groups at C-4, C-5, and C-6 (Figure S90). Surprisingly, disruption of the dehydrogenase gene 274 assumed for epoxidation did not result in expected abolishment of epoxy-containing SECs in mutant Δ274 but led to a number of novel SEC products with diverse oxygenated degrees and patterns. Disruption of the polyprenyltransferase gene 276 led to huge accumulation of toluquinol and new glycosides in mutant Δ276, suggesting that toluquinol underwent farnesylation in the biosynthetic pathway for the SECs. The resulting product farnesyl hydroquinone (30) should be the first and key PK−TP hybrid precursor for SECs, instead of 8 as previously proposed.4
farnesyl hydroquinone.10 Among trace targeted metabolites, 31−33 were elucidated as three new derivatives of 30, and 34 was farnesyl quinone, an oxidative derivative of 30.10 The co-occurrence of 30 and 34 in mutant Δ278 allowed us to deduce that the conversion of 30 to SECs might be achieved via 34 to undergo epoxidation (Scheme 1), as epoxy-quinoneScheme 1. Proposed Biosynthetic Pathway for Natural SECs According to Metabolites from Mutants
3925
DOI: 10.1021/acs.orglett.7b01846 Org. Lett. 2017, 19, 3923−3926
Letter
Organic Letters containing SECs are widely distributed in fungi.2,3,10 There would be another possibility that 30 first undergoes epoxidation and then subsequent dehydrogenation to form farnesyl epoxyquinone. In any case, the gene 278 in Ar. oligospora might be involved in the epoxidation to yield the first and simplest SEC farnesyl epoxy-quinone. The occurrence of metabolites 19 and 21 with ketone at C-1 or C4, respectively, in the mutant Δ274 gave a hint of the presence of a key intermediate farnesyl epoxyquinone because they might be derived from farnesyl epoxyquinone through two regional ketone reductions either at C-1 or C-4 in the epoxy-quinone core. The plausible model for the biogenesis of SECs is proposed as shown in Scheme 1. The presence of major intermediates with ketone at C-1 in mutant Δ274 and the production of the SEC metabolites with ketone at C-4 in mutant Δ280 further suggested that the genes 274 and 280 were involved in specific regional ketone reductions at C-4 and C-1, respectively, of farnesyl epoxy-quinone. Thus, lack of the gene 280 led to first generation of the SECs with ketone at C-4, and disruption of the gene 274 resulted in extraordinary production of SECs with ketone at C-1. However, the homologue of gene 278 is just missing in the Yan pathway. How the same epoxidation can be achieved with a simpler gene set becomes interesting. Further study on the catalytic functions of 278, 280, and 274 with biochemical experiments is in progress. In addition, the escalation of the decorating levels in the intermediates from mutants Δ274 suggested that oxidative modifications of 20 to morphological regulatory arthrobotrin B should start from hydroxylation at C-1′ in the farnesyl chain and end with the oxygenation at 2-Me in the EC. The modification pathway to arthrobotrisin B involved 19, 22, and 25 as biosynthetic intermediates, while 21 could be the key intermediate for arthrobotrisin C. Metabolites 20 could be also the key precursor for 13-membered macrolide 29 via opening of the epoxy ring to form 27, conversion of α-OH at C-6 in 27 to β-OH in 28, an oxidative cleavage at Δ10′,11′, and an internal dehydration−condensation of the COOH with 6-OH (Scheme 1). In conclusion, we demonstrated that farnesyl hydroquinone (30) should be the first hybrid precursor for biosynthesis of SECs, and farnesyl quinone (34) might be the key precursor for the epoxy ring formation. A number of novel intermediates allowed us to dissect the modifying steps from key intermediates to morphological regulatory arthrobotrisins.
■
Author Contributions
The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was sponsored by NSFC U1502262 and 31470169, 2013CB127505, and Yunnan University Program for Excellent Young Talents awarded to X.N. (XT412003).
■
(1) (a) Meerupati, T.; Andersson, K. M.; Friman, E.; Kumar, D.; Tunlid, A.; Ahren, D. PLoS Genet. 2013, 9, e1003909. (b) Li, J.; Zou, C.; Xu, J.; Ji, X.; Niu, X.; Yang, J.; Zhang, K. Annu. Rev. Phytopathol. 2015, 53, 67−95. (2) (a) Anderson, M. G.; Rickards, R. W.; Lacey, E. J. Antibiot. 1999, 52, 1023−1028. (b) Stadler, M.; Sterner, O.; Anke, H. Z. Naturforsch. C 1993, 48, 843−850. (c) Anderson, M. G.; Jarman, T. B.; Rickards, R. W. J. Antibiot. 1995, 48, 391−398. (d) Wei, L.; Zhang, H.; Tan, J.; Chu, Y.; Li, N.; Xue, H.; Wang, Y.; Niu, X.; Zhang, Y.; Zhang, K. J. Nat. Prod. 2011, 74, 1526−1530. (e) Zhang, H.; Tan, J.; Wei, L.; Wang, Y.; Zhang, C.; Wu, D.; Zhu, C.; Zhang, Y.; Zhang, K.; Niu, X. J. Nat. Prod. 2012, 75, 1419−1423. (3) (a) Matsuda, Y.; Abe, I. Nat. Prod. Rep. 2016, 33, 26−53. (b) Marco-Contelles, J.; Molina, M. T.; Anjum, S. Chem. Rev. 2004, 104, 2857−2899. (c) Shan, W.-G.; Ying, Y.-M.; Ma, L.-F.; Zhan, Z.-J., Chapter 6-Drimane-Related Merosesquiterpenoids, a Promising Library of Metabolites for Drug Development. In Studies in Natural Products Chemistry; Atta-Ur-Rahman, Ed.; Elsevier: Amsterdam, 2015; Vol. 45, pp 147−215. (4) Holm, D. K.; Petersen, L. M.; Klitgaard, A.; Knudsen, P. B.; Jarczynska, Z. D.; Nielsen, K. F.; Gotfredsen, C. H.; Larsen, T. O.; Mortensen, U. H. Chem. Biol. 2014, 21, 519−529. (5) (a) Puel, O.; Galtier, P.; Oswald, I. P. Toxins 2010, 2, 613−631. (b) Huffman, J.; Gerber, R.; Du, L. Biopolymers 2010, 93, 764−776. (c) Tannous, J.; El Khoury, R.; Snini, S. P.; Lippi, Y.; El Khoury, A.; Atoui, A.; Lteif, R.; Oswald, I. P.; Puel, O. Int. J. Food Microbiol. 2014, 189, 51−60. (6) Xu, Z.; Wang, B.; Sun, H.; Yan, N.; Zeng, Z.; Zhang, K.; Niu, X. J. Agric. Food Chem. 2015, 63, 9076−9082. (7) Xu, Z.; Chen, Y.; Song, T.; Zeng, Z.; Yan, N.; Zhang, K.; Niu, X. J. Agric. Food Chem. 2016, 64, 7949−7956. (8) Song, T.; Xu, Z.; Chen, Y.; Ding, Q.; Sun, Y.; Miao, Y.; Zhang, K.; Niu, X. J. Agric. Food Chem. 2017, 65, 4111−4120. (9) Wu, D.; Zhang, C.; Zhu, C.; Wang, Y.; Guo, L.; Zhang, K.; Niu, X. J. Agric. Food Chem. 2013, 61, 4108−4113. (10) (a) Son, B. W.; Kim, J. C.; Choi, H. D.; Kang, J. S. Arch. Pharmacal Res. 2002, 25, 77−79. (b) Mothana, R. A. A.; Jansen, R.; Julich, W. D.; Lindequist, U. J. Nat. Prod. 2000, 63, 416−418.
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.orglett.7b01846. Complete description of materials and methods and NMR and MS data of 19 metabolites (PDF)
■
REFERENCES
AUTHOR INFORMATION
Corresponding Authors
*E-mail:
[email protected]. *E-mail:
[email protected]. *E-mail:
[email protected]. ORCID
Xue-Mei Niu: 0000-0002-8977-8956 3926
DOI: 10.1021/acs.orglett.7b01846 Org. Lett. 2017, 19, 3923−3926
