Letter Cite This: Org. Lett. XXXX, XXX, XXX−XXX
pubs.acs.org/OrgLett
Targeted Gene Inactivations Expose Silent Cytochalasans in Magnaporthe grisea NI980 Chongqing Wang, Verena Hantke, Russell J. Cox, and Elizabeth Skellam* Institute for Organic Chemistry and Centre for Biomolecular Drug Research, Leibniz University Hannover, Schneiderberg 38, Hannover 30167, Germany
Org. Lett. Downloaded from pubs.acs.org by BOSTON COLG on 05/19/19. For personal use only.
S Supporting Information *
ABSTRACT: The biosynthetic gene cluster encoding the phytotoxin pyrichalasin H 5 was discovered in Magnaporthe grisea NI980, and the late-stage biosynthetic pathway of 5 was fully elucidated using targeted gene inactivations resulting in the isolation of 13 novel cytochalasans. This study reveals that the nonproteinogenic amino acid O-methyltyrosine is the true precursor of 5, and other cryptic cytochalasans and mutasynthesis experiments produce novel halogenated pyrichalasin H analogues.
C
ytochalasans are a class of structurally diverse fungal polyketide nonribosomal peptide secondary metabolites which exhibit a wide range of biological properties, acting as phytotoxins, virulence factors, antimicrobials, and cytotoxins.1 Their structural diversity arises from the different amino acids that can be incorporated and the range of polyketide-derived backbones that can be combined to generate the characteristic isoindole core fused to a macrocycle.2 Tailoring modifications such as oxygenation, acetylation, cyclization, and dimerization further diversifies the carbon skeleton.3−5 Several biosynthetic gene clusters (BGCs) encoding cytochalasan pathways have been reported and partially characterized including those for cytochalasin E 1, chaetoglobosin A 2, and two cryptic cytochalasan metabolites from Magnaporthe oryzae Guy11 3 and 4 (Figure 1).6−9 All cytochalasan BGCs contain genes encoding a polyketide synthase/nonribosomal peptide synthetase (PKS-NRPS), a trans-enoyl reductase (ER), an α,βhydrolase (HYD), and a putative Diels−Alderase (pDA).6,7,10,11 Additional genes encoding tailoring enzymes are also common.6,7,10,11 Molecular investigations of these pathways have revealed important information about key biosynthetic steps but have not yet been applied to the systematic engineering of novel cytochalasans in vivo. Pyrichalasin H 5, isolated from various strains of Magnaporthe grisea, inhibits the growth of rice seedlings, inhibits lymphocyte capping and actin polymerization, alters cell morphology, and is indicated as the responsible agent for the genus-specific pathogenicity of M. grisea toward Digitaria © XXXX American Chemical Society
Figure 1. Structures of cytochalasans associated with characterized biosynthetic gene clusters.
sp. (crabgrass).12−15 Due to the rarity of tyrosine-derived cytochalasans, the high titers of production of 5,15 and the observation that 5 has an O-acetyl moiety and not a carbonate functionality like many of the other tyrosine-derived cytochalasans (Figure S1), we reasoned that targeted gene disruption would enable us to access a number of novel cytochalasans. Production of 5 by M. grisea NI980 was investigated under various culture conditions. Careful examination of the culture extract also revealed the presence of a small amount of compound with high-resolution ESIMS (HRESIMS) m/z Received: April 17, 2019
A
DOI: 10.1021/acs.orglett.9b01344 Org. Lett. XXXX, XXX, XXX−XXX
Letter
Organic Letters 504.2729 [M + Na]+ (calcd for C29H39NO5Na+, 504.2726). 1H NMR confirmed this to be the deacetylated pyrichalasin H 6. We chemically prepared and fully characterized a standard of 6 from 5 by alkaline hydrolysis. The NMR data, LC-MS retention time, UV absorption profile, and mass fragmentation profiles are identical for the semisynthetic and extracted compounds. Genome sequencing of M. grisea NI980 was undertaken, and analysis using fungiSMASH16 identified 66 BGCs including seven PKS-NRPS clusters. Only one PKS-NRPS cluster, located on M. grisea Scaffold (Mg−Sc) 00012, contains all four core genes necessary for cytochalasan biosynthesis (Figure 2A, Table S1, GenBank accession number MK801691).
Surprisingly, OME1 is part of the ACE1 cluster in M. grisea BR29 (MgBR29−Sc00037, Figure 2C) and not encoded within the SYN2 cluster (MgBR29−Sc00093, Figure 2C) as observed in M. oryzae Guy11. To confirm the putative pyi BGC in M. grisea NI980, we targeted pyiS using the bipartite gene knockout (KO) strategy.18 As expected, inactivation of the PKS-NRPS resulted in the complete abolition of 5 and 6 (Figure 3B). Next, pyiC,
Figure 2. Cytochalasan BGCs identified in Magnaporthe sp. (A) M. grisea NI980, (B) M. oryzae Guy11, and (C) M. grisea BR29. The pyi clusters are highlighted in yellow.
However, two truncated clusters were also identified at the end of two separate scaffolds containing either the α,βhydrolase or the pDA gene (Mg−Sc00220 and Mg−Sc00273, respectively; Figure 2A). An investigation of these clusters revealed that they have high sequence similarity to the ACE1 and SYN2 clusters in M. oryzae Guy11,11 with homologous genes appearing in the same order and orientation (Figure 2B, Tables S2 and S3). Pyrichalasin H 5 contains a para-methoxyphenyl group and an O-acetyl group at C-21 and, thus, must require O-methyltransferase (O-MeT) and O-acetyl-transferase (O-AT) catalysts. The BGC on Mg−Sc00012 (Figure 2A) is the most likely candidate as it contains genes putatively encoding these functions, while the other PKS-NRPS clusters do not. We designated this putative BGC as the pyi gene cluster and assigned trivial names to the genes in analogy to the names of genes from the cytochalasin E 1 cluster.6 Recently, and in parallel to our work, the genome sequence of M. grisea BR29 was published.17 We identified a putative BGC encoding 5 on MgBR29−Sc00009, with all homologous genes in the same order and orientation as the putative pyi BGC in M. grisea NI980 (Figure 2C). We also identified homologous gene clusters for ACE1 and SYN2 in M. grisea BR29 on two separate scaffolds (MgBR29−Sc00037 and MgBR29−Sc00098, respectively, Figure 2C) which are not physically clustered as in M. oryzae Guy11 (Figure 2B).
Figure 3. HPLC chromatograms of mycelial extracts derived from M. grisea NI980 knock out transformants. Chromatograms were monitored using a diode array detector scanning from 200 to 600 nm. Peaks labeled with * and + are unrelated unknown metabolites, and compounds eluting before 4 min are pyriculol-related metabolites (see the SI for details).
the gene encoding a trans-ER protein, was targeted; this was to confirm that there was no cross-talk between enzymes from the other cytochalasan pathways in M. grisea. Compounds 5 and 6 were completely abolished, and no new metabolites were detected (Figure 3C). In an attempt to generate novel cytochalasans we individually targeted all genes believed to encode tailoring modifications. Functional disruption of pyiB, encoding a putative acetyl-transferase, caused the loss of 5. As expected, 6 remained, but its titer was not significantly increased (Figure 3D, Scheme 1). Therefore, pyiB encodes the O-AT which participates in the biosynthesis of 5. B
DOI: 10.1021/acs.orglett.9b01344 Org. Lett. XXXX, XXX, XXX−XXX
Letter
Organic Letters
Scheme 1. Putative Late-Stage Biosynthetic Pathway of Pyrichalasin H (5) and Isolated Intermediates (Grey Compounds Inferred)
Disruption of pyiD, which encodes a P450 monooxygenase, resulted in the loss of 5 and production of four new compounds 10−13 (Figure 3F). All except 10, which was previously observed in a biotransformation experiment where cytochalasan L-696,474 was fed to Actinoplanes sp. ATCC,19 are novel, and all lack a hydroxyl group at C-18, confirming that pyiD encodes the C-18 hydroxylase. Compounds 10−12 most likely result from off-pathway oxidations at C-22, similar to the recently discovered oxichaetoglobosins.20 The targeted inactivation of pyiG, encoding a second P450 monooxygenase, abolished the production of 5, and five new compounds 14−18 were observed (Figure 3G). These were identified as novel cytochalasans using HRMS and NMR data. NMR analysis of 15 revealed that it is actually a mixture of two coeluting compounds, designated as compounds 15i and 15ii. These were, however, resolved by 2D NMR which allowed individual structure elucidation. As compounds 14−18 all lack a hydroxyl group at C-7, this confirms that pyiG encodes the C-7 hydroxylase.
Disruption of the oxidoreductase encoded by pyiH resulted in the loss of 5 and production of three new compounds 7−9 (Figure 3E). These were identified as novel cytochalasans using HRMS and NMR (see the SI for details). Compound 8 contains the expected C-21 carbonyl group but lacks the expected C-19/C-20 olefin. Compound 7 contains an unexpected hydroxyl group at C-12, indicating that an offpathway enzyme can modify the intermediates, perhaps as a detoxification strategy. Compound 9 was isolated in a very low yield with HRESIMS m/z 464.2802 [M + H]+ (calcd for C29H38NO4+, 464.2801). Although the signals in the NMR spectra were weak, there was enough data to infer that 9 contains the expected C-19/C-20 olefin that 8 lacks. All other observable NMR signals were similar to 8, indicating that all other functionalities remained the same. Combined with the HRESIMS, where 9 m/z 488.2775 [M + Na]+ (calcd for C29H39NO4+, 488.2777) differs from 8 by 2 Da, a putative structure can be deduced. Thus, PyiH reduces the carbonyl at C-21 in preparation for the transfer of an acetyl group by PyiB. C
DOI: 10.1021/acs.orglett.9b01344 Org. Lett. XXXX, XXX, XXX−XXX
Letter
Organic Letters Finally, the O-MeT-encoding gene, pyiA, was disrupted. This resulted in the total loss of 5, and two new metabolites 19 and 20 were observed in low titers (Figure 3H). Compounds 19 and 20 were fully characterized and shown to be the tyrosine and phenylalanine analogues of 5, the novel 19 which we name magnachalasin H, and the already known cytochalasin H 20. The observation of 19 could be explained if the PKS-NRPS adenylation (A) domain is selective for tyrosine, and later in the biosynthesis, the O-MeT acts to methylate this residue. However, the very low titer of 19 and the observation of 20 are surprising, as 20 is not observed in any M. grisea NI980 wildtype (WT) extracts and thus is unlikely to be a shunt metabolite. A different interpretation of these results is that the NRPS A-domain preferentially selects O-methyltyrosine 21 but in its absence can weakly select and activate phenylalanine and tyrosine. During aspyridone biosynthesis, for example, the aspyridone synthetase ApdA from Aspergillus nidulans is able to accept tryptophan, tyrosine, phenylalanine, and fluorophenylalanine in vitro.21 21 has been reported as a substrate for fungal NRPS A-domains in 4-methoxycyclopeptin biosynthesis through examination of HPLC extracts and inferred in other pathways, e.g., 4′-methoxyviridicatin and penigequinolones.22,23 To test whether 21 is the true substrate for PyiS, we fed 21 to the pyiA KO strain and observed the restoration of 5 in titers comparable to the wild-type (Figure 3I). Compounds 19 and 20 could no longer be detected under these conditions clarifying that 21 is the preferred substrate of the NRPS A-domain. Thus it appears that the O-MeT PyiA methylates free tyrosine to generate O-methyltyrosine 21, which is selected by the PyiS A-domain (Scheme 1). Compound 22 is then released as an early enzyme-free intermediate. Due to the observation of 7 and 8 in the PyiH KO, it appears that PyiG is one of the first enzymes to act. The complexity of the pathway arises due to the promiscuity of the tailoring enzymes which are all able to act on more than one substrate, similar to the observations in chaetoglobosin A 2 biosynthesis.7 When PyiA is absent, or when the biosynthetic machinery is stressed, the PyiS Adomain is able to select phenylalanine and/or tyrosine, explaining the observation of 10, 11, 13, 15ii, 16, 19, and 20. In previous work we had attempted to identify the cryptic ACE1 avirulence molecule from M. oryzae Guy11 through heterologous gene expression experiments, resulting in very low yields of the linear tyrosine nonaketide 4.8 Considering that there is an O-MeT gene in the ACE1/SYN2 supercluster in M. oryzae Guy11, but this O-MeT homologue is clustered only with the ACE1 cluster in M. grisea BR29, we suspected that the ACE1 PKS-NRPS may also preferentially select 21. We therefore fed 21 to Aspergillus oryzae expressing ACE1 and RAP1 and observed accumulation of a new peak in the chromatogram (Figure 4B). This new compound was characterized and identified as the O-methyltyrosine nonaketide 23. The titer was 2 mg/L, an improvement on the original isolation of 4.8 As the NRPS A-domains of ACE1 and PyiS both demonstrate substrate tolerance, we next investigated precursor-directed biosynthesis in the A. oryzae ACE1 + RAP1 heterologous host and the M. grisea pyiA KO strain by feeding para-chlorophenylalanine 24. New metabolites 25 and 26 were observed for both strains, respectively (Figure 4C,E). Compound 25 could only be isolated in low titer, but HRESIMS revealed m/z 522.2759 [M + Na]+ (calcd for
Figure 4. LC-MS chromatograms of organic extracts from feeding experiments and the structures of the compounds isolated: (A) A. oryzae (ACE1 + RAP1) control; (B) A. oryzae (ACE1 + RAP1) + Omethyl-L -tyrosine; (C) A. oryzae (ACE1 + RAP1) + parachlorophenylalanine; (D) M. grisea NI980 ΔpyiA; (E) M. grisea NI980 ΔpyiA + para-chlorophenylalanine; (F) M. grisea NI980 ΔpyiA + para-fluorophenylalanine; and (G) M. grisea NI980 ΔpyiA + parabromophenylalanine. Chromatograms were monitored using a diode array detector scanning from 200 to 600 nm.
C30H42NO3ClNa+, 522.2751), corresponding to the incorporation of para-chlorophenylalanine 24. Compound 26 was purified and fully characterized by HRMS and NMR. paraFluorophenylalanine 27 and para-bromophenylalanine 28 were also fed to the M. grisea pyiA KO strain resulting in new compounds 29 and 30 (Figure 4F,G) which were fully characterized by HRMS and NMR. Precursor-directed biosynthesis has been reported in the chaetoglobosin A, B, and J producer Chaetomium globosum yielding novel compounds; however, feeding was performed on the wild-type fungus resulting in a mixture of products.24 To the best of our knowledge this is the first example of mutasynthesis applied to a cytochalasan pathway to engineer the production of new natural products. In conclusion, we have sequenced the genome of M. grisea NI980 and identified three putative cytochalasan BGCs through genome mining. Comparative genomics identified three homologous gene clusters in M. grisea BR29; one is a homologue of the pyi cluster in M. grisea NI980, and the other two are related to the ACE1/SYN2 gene cluster from M. oryzae Guy11. Targeted gene disruption of pyiS in M. grisea NI980 confirmed that the pyi BGC encodes the biosynthesis of 5. Targeted deletions of the tailoring genes resulted in the D
DOI: 10.1021/acs.orglett.9b01344 Org. Lett. XXXX, XXX, XXX−XXX
Letter
Organic Letters
(9) Nielsen, M. L.; Isbrandt, T.; Petersen, L. M.; Mortensen, U. H.; Andersen, M. R.; Hoof, J. B.; Larsen, T. O. PLoS One 2016, 11, No. e0161199. (10) Moore, G. G; Collemare, J.; Lebrun, M.-H; Bradshaw, R. E. Natural Products: Discourse, Diversity, and Design; John Wiley & Sons, Inc.: Hoboken, NJ, 2014. (11) Collemare, J.; Pianfetti, M.; Houlle, A.-E.; Morin, D.; Camborde, I.; Gagey, M.-J; Barbisan, C.; Fudal, I.; Lebrun, M.-H; Böhnert, H. U. New Phytol. 2008, 179, 196−208. (12) Klaubauf, S.; Tharreau, D.; Fournier, E.; Groenewald, J. Z.; Crous, P. W.; de Vries, R. P.; Lebrun, M.-H Stud. Mycol. 2014, 79, 85−120. (13) Nukina, M. Agric. Biol. Chem. 1987, 51, 2625−2628. (14) Tsurushima, T.; Don, L. D.; Kawashima, K.; Murakami, J.; Nakayashiki, H.; Tosa, Y.; Mayama, S. Mol. Plant Pathol. 2005, 6, 605−613. (15) Hirose, T.; Izawa, Y.; Koyama, K.; Natori, S.; Iida, K.; Yahara, I.; Shimaoka, S.; Maruyama, K. Chem. Pharm. Bull. 1990, 38, 971− 974. (16) Blin, K.; Wolf, T.; Chevrette, M. G.; Lu, X.; Schwalen, C. J.; Kautsar, S. A.; Suarez-Duran, H. G.; de Los Santos, E. L. C; Kim, H. U.; Nave, M.; Dickschat, J. S.; Mitchell, D. A.; Shelest, E.; Breitling, R.; Takano, E.; Lee, S. Y.; Weber, T.; Medema, M. H. Nucleic Acids Res. 2017, 45, W36−W41. (17) Chiapello, H.; Mallet, L.; Guerin, C.; Aguileta, G.; Amselem, J.; Kroj, T.; Ortega-Abboud, E.; Lebrun, M.-H; Henrissat, B.; Gendrault, A.; Rodolphe, F.; Tharreau, D.; Fournier, E. Genome Biol. Evol. 2015, 7, 2896−2912. (18) Nielsen, M.; Albertsen, L.; Lettier, G.; Nielsen, J.; Mortensen, U. Fungal Genet. Biol. 2006, 43, 54−64. (19) Chen, T. S.; Doss, G. A.; Hsu, A.; Hsu, A.; Lingham, R. B.; White, R. F.; Monaghan, R. L. J. Nat. Prod. 1993, 56, 755−761. (20) Wang, W.; Gong, J.; Liu, X.; Dai, C.; Wang, Y.; Li, X.-N; Wang, J.; Luo, Z.; Zhou, Y.; Xue, Y.; Zhu, H.; Chen, C.; Zhang, Y. J. Nat. Prod. 2018, 81, 1578−1587. (21) Xu, W.; Cai, X.; Jung, M. E.; Tang, Y. J. Am. Chem. Soc. 2010, 132, 13604−13607. (22) Ishikawa, N.; Tanaka, H.; Koyama, F.; Noguchi, H.; Wang, C. C. C; Hotta, K.; Watanabe, K. Angew. Chem., Int. Ed. 2014, 53, 12880−12884. (23) Zou, Y.; Zhan, Z.; Li, D.; Tang, M.; Cacho, R. A.; Watanabe, K.; Tang, Y. J. Am. Chem. Soc. 2015, 137, 4980−4983. (24) Ge, H. M.; Yan, W.; Guo, Z. K.; Luo, Q.; Feng, R.; Zang, L. Y.; Shen, Y.; Jiao, R. H.; Xu, Q.; Tan, R. X. Chem. Commun. 2011, 47, 2321−232.
production of 13 previously unreported cytochalasans and established a likely biosynthetic network for the production of 5. Disruption of pyiA revealed that the precursor for 5 is the nonproteinogenic amino acid O-methyltyrosine 21 and that this is also true for the cryptic ACE1 metabolite from M. oryzae Guy11. Due to the tolerance of the A-domain of PyiS, we were able to engineer novel halogenated cytochalasans 26, 29, and 30 at yields close to the WT production of 5.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.orglett.9b01344. Experimental details and additional data and figures including structures, chromatograms, UV spectra, HRMS results, 1H NMR, 13C NMR, H−H COSY spectra, HMBC spectra, and HSQC spectra (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Russell J. Cox: 0000-0002-1844-0157 Elizabeth Skellam: 0000-0003-4087-0708 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the Chinese Scholarship Council [C.W. (201608310143)] and the German Research Foundation (DFG) [Grants CO 1328/2-1, INST 187/621-1, INST 187/686-1)]. The authors would also like to thank Professor Marc-Henri Lebrun (UMR BIOGER INRA AgroParisTech) and Dr. Didier Tharreau (CIRAD UMR BGPI) for the gift of the M. grisea NI980 strain and useful discussions. In addition we would like to thank Dr. Daniel Wibberg and Professor Jörn Kalinowski at CeBiTec, Bielefeld, for sequencing and annotation of M. grisea NI980. Leibniz University Hannover students Hao Xuan Zeng and Yan Cui are also thanked for assistance with extracting fungal cultures and preparing gDNA.
■
REFERENCES
(1) Scherlach, K.; Boettger, D.; Remme, N.; Hertweck, C. Nat. Prod. Rep. 2010, 27, 869−886. (2) Skellam, E. Nat. Prod. Rep. 2017, 34, 1252−1263. (3) Hu, Y.; Dietrich, D.; Xu, W.; Patel, A.; Thuss, J. A. J; Wang, J.; Yin, W.-B; Qiao, K.; Houk, W.-B; Vederas, J. C.; Tang, Y. Nat. Chem. Biol. 2014, 10, 552−554. (4) Chen, C.; Wang, J.; Liu, J.; Zhu, H.; Sun, B.; Wang, J.; Zhang, J.; Luo, Z.; Yao, G.; Xue, Y.; Zhang, Y. J. Nat. Prod. 2015, 78, 1193− 1201. (5) Zhu, H.; Chen, C.; Xue, Y.; Tong, Q.; Li, X. N.; Chen, X.; Wang, J.; Yao, G.; Luo, Z.; Zhang, Y. Angew. Chem., Int. Ed. 2015, 54, 13374−13378. (6) Qiao, K.; Chooi, Y.-H; Tang, Y. Metab. Eng. 2011, 13, 723−732. (7) Ishiuchi, K.; Nakazawa, T.; Yagishita, F.; Mino, T.; Noguchi, H.; Hotta, K.; Watanabe, K. J. Am. Chem. Soc. 2013, 135, 7371−7377. (8) Song, Z.; Bakeer, W.; Marshall, J. W.; Yakasai, A. A.; Khalid, R. M.; Collemare, J.; Skellam, E.; Tharreau, D.; Lebrun, M.-H; Lazarus, C. M.; Bailey, A. M.; Simpson, T. J.; Cox, R. J. Chem. Sci. 2015, 6, 4837−4845. E
DOI: 10.1021/acs.orglett.9b01344 Org. Lett. XXXX, XXX, XXX−XXX