Confirmation of the stereochemistry of spiroviolene

  1. ‡,1 ,
  2. ‡,1 ,
  3. 1 ,
  4. 1 ORCID Logo ,
  5. 1 ,
  6. 1 ,
  7. 1 ,
  8. 2 ORCID Logo ,
  9. 3 ORCID Logo and
  10. 1 ORCID Logo
1State Key Laboratory of Natural and Biomimetic Drugs, School of Pharmaceutical Sciences, Peking University, Beijing 100191, China; Ningbo Institute of Marine Medicine, Peking University, Ningbo 315010, China
2State Key Laboratory of Microbial Resources, Institute of Microbiology, Chinese Academy of Sciences, Beijing 100101, China; University of Chinese Academy of Sciences, Beijing 100049, China
3School of Medicine, Yunnan University, 2 North Cui Hu Road, Kunming 650091, China
  1. Corresponding author email
  2. ‡ Equal contributors
Guest Editor: L. Barra
Beilstein J. Org. Chem. 2024, 20, 852–858.
Received 28 Dec 2023, Accepted 10 Apr 2024, Published 18 Apr 2024
A non-peer-reviewed version of this article has been posted as a preprint


We confirm the previously revised stereochemistry of spiroviolene by X-ray crystallographically characterizing a hydrazone derivative of 9-oxospiroviolane, which is synthesized by hydroboration/oxidation of spiroviolene followed by oxidation of the resultant hydroxy group. An unexpected thermal boron migration occurred during the hydroboration process of spiroviolene that resulted in the production of a mixture of 1α-hydroxyspiroviolane, 9α- and 9β-hydroxyspiroviolane after oxidation. The assertion of the cis-orientation of the 19- and 20-methyl groups provided further support for the revised cyclization mechanism of spiroviolene.


Terpenes represent one of the most fascinating families of natural products due to their structural complexity and diversity, as well as their indispensable biological functions that would be potentially applied as fragrances, pharmaceuticals etc. Until now, more than 80,000 terpenoid structures have been reported, which are found in all domains of life [1-3]. Despite their remarkable chemodiversity, the biosynthetic logic of terpenes is straightforward [4]. All terpenes are originated from two key C5 building blocks, namely isopentenyl pyrophosphate (IPP) and dimethylallyl pyrophosphate (DMAPP), which are biosynthesized via either the methylerythritol phosphate (MEP) pathway or the mevalonic acid (MVA) pathway by using the primary metabolites. Different numbers of IPP and DMAPP are assembled by prenyltransferases to afford oligoprenyl pyrophosphates, such as farnesyl pyrophosphate (FPP, 3 × C5) and geranylgeranyl pyrophosphate (GGPP, 4 × C5), with varied C5 units. The linear oligoprenyl pyrophosphates are typically converted by terpene synthases in a chemo- and stereoselective process to form complex terpene skeletons, normally with multiple stereocenters. In this context, the 3D-defined cyclization products retain the rich information of the complex cyclization process. Thus, assignment of the stereochemistry of the terpene skeleton with high confidence is crucial for proposing a reasonable cyclization mechanism [5].

Spiroviolene (1, Figure 1) was identified by Dickschat and co-workers as a nascent cyclization product of spiroviolene synthase (SvS), the coding gene of which was cloned from Streptomyces violens NRRL ISP-5597 [6]. Its unique spiro-fused linear triquinane to cyclopentane skeleton, as well as its stereochemistry, was originally elucidated as 1' as shown in Figure 1, on the basis of detailed analysis of 1D and 2D NMR spectroscopy. Spiroviolene was also found to be produced by several bacterial strains harboring SvS homologs [6,7], as well as putative ancestors of SvS generated by ancestral sequence reconstruction [8,9]. Related natural products with the same 5-5-5-5 tetracyclic ring system, including spirograterpene A (2) from Penicillium granulatum MCCC 3A00475 [10], and GJ1012A (3) from an engineered E. coli strain harboring FgGS (FgJ07623) cloned from Fusarium graminearum GJ1012 [11], have been reported almost at the same time. The discrepancy of the stereochemistry at C3 between 1' and 2 was first noticed by Snyder and co-workers [12]. The conserved stereochemistry of spiroviolene and 2 at C3 was later confirmed by the conversion of a synthetic intermediate of 2 to spiroviolene. By taking advantage of the DFT transition state analysis of the hydroboration reaction of a key intermediate, as well as NOE correlation analysis of the resultant product, Snyder and co-workers have reassigned the right structure of spiroviolene to 1. However, direct evidences such as single-crystal X-ray diffraction results were not reported in their study.


Figure 1: Structures of spiroviolene and related natural products.

The reassignment of the stereochemistry at C3 has resulted in the revision of the proposed cyclization mechanism [12-14]. The revised mechanism resembled the cyclization process for the formation of deoxyconidiogenol (4, Scheme 1A) by several terpene cyclases from fungus (PcCS, PchDS, PrDS) [15,16], which involves a 1,11-10,14 cyclization of GGPP, followed by 1,2-alkyl shift and a 2,10-cyclization, to give the key C3 cationic intermediate IM-1. A key 1,2-hydride shift from C2 to C3, which was observed in the isotope labeling experiments [6], followed by a 2,7-cyclization, afforded C6 cationic intermediate IM-3 with cyclopiane skeleton. Quench of the cation IM-3 with water would give 4, while upon two 1,2-alkyl shifts of IM-3, followed by deprotonation of cation IM-4, would give spiroviolene (1).


Scheme 1: Possible cyclization mechanisms for spiroviolene (1) and related natural products. A) Revised cyclization mechanism for 1 that is related to deoxycondiogenol (4). B) Originally proposed cyclization mechanism for the misassigned structure 1'. C) Possible cyclization mechanism for GJ1012A (3). D) Structures of biogenetically related diterpenes during the proposed cyclization process.

On the other hand, the originally proposed cyclization mechanism (Scheme 1B) involves a 3,6-cyclization of cation IM-1 through a conformation shown as IM-1A to generate cation IM-5, which was proposed to undergo a dyotropic rearrangement, followed by a 1,2-alkyl shift of cation IM-6 to yield the spirocyclic cation IM-7. A key 1,3-hydride shift of IM-7 from the β-face, followed by deprotonation of the formed C2-cation IM-8, would deliver the originally proposed structure 1' [6]. However, no related natural products that would be derived from the intermediates of this pathway have been found so far.

A third cyclization mechanism (Scheme 1C) leading to the same spirocyclic skeleton as spiroviolene with an altered stereochemistry at C7 found in GJ1012A (3) could be proposed [11,17,18]. A 3,6- or 3,7-cyclization of cation IM-1 through a conformation shown as IM-1B with β-oriented 20-methyl group, would generate either IM-9 or IM-11 cations. A direct dyotropic rearrangement, or two stepwise 1,2-alkyl migrations of IM-9, are possible pathways en route to cation IM-10. The presence of these intermediates IM-9, -10, -11 could be inferred by the identification of GJ1012B/D (5/6, Scheme 1D) [11], cattleyene (7) [19,20], and fusaterpenol (8, GJ1012E) [17]. A similar 1,2-alkyl shift of IM-10, followed by deprotonation of the formed spirocyclic cation IM-12, afforded 3. Although previous isotope labeling experiments did not support this pathway for spiroviolene cyclization, it should be noted that a subtle alteration of stereochemical assignment of spiroviolene would have consequences for a different mechanistic proposal. We herein report the production of spiroviolene (1) in a heterologous host by taking advantage of an artificial isopentenol utilization pathway [21-26], and confirm its stereochemistry by X-ray crystallography using a hydrazone derivative of 1.

Results and Discussion

Our work commenced with the heterologous production of spiroviolene by E. coli using a recently developed isopentenol utilization pathway for the efficient supply of two C5 precursors for terpene biosynthesis (Figure 2) [21-26]. In this artificially generated pathway, DMAPP and IPP could be easily generated from prenol and isoprenol, respectively, by the effect of two kinases, such as hydroxyethylthiazole kinase from E. coli (EcThiM) and isopentenyl phosphate kinase from Methanocaldococcus jannaschii (MjIPK). Thus, we have cloned genes coding EcThiM, MjIPK, isopentenyl diphosphate isomerase of E. coli (IDI), farnesyl diphosphate synthase of E. coli (IspA) and geranylgeranyl diphosphate synthase of Pantoea agglomerans (CrtE) into the multiple cloning site-2 of pCDFDeut-1 to give pCDFDeut-TIIAE for GGPP production. Also, we have cloned the SvS-coding gene directly from Streptomyces violens CGMCC 4.1786 (= NRRL ISP-5597) into pET28a to give pET28a-svs. The resultant two plasmids were then co-transformed into commercially available E. coli BL21(DE3) for diterpene production. Spiroviolene could be produced by feeding prenol and/or isoprenol to the fermentation broth after E. coli being induced by IPTG, and fermented at 18 °C for 72 h. In our hand, feeding a mixture of prenol and isoprenol in a 1:1 ratio would give the best yield of spiroviolene. GC–MS analysis (Figure 2A) of the EtOAc extract of the fermentation broth gave a single peak, whose EIMS spectrum (Figure 2B) matches that of spiroviolene. We then carried out a large-scale fermentation using 2 L-shake flasks, and each flask contains 1.0 L of modified TB medium. We have isolated 201 mg (40 mg/L) of spiroviolene from 5.0 L of the fermentation broth. The physicochemical data of the isolated material are consistent with those reported for spiroviolene [6].


Figure 2: Heterologous production of spiroviolene using the isopentenol utilization pathway. A) Gas chromatogram of the EtOAc extract of the fermentation broth. B) EI mass spectrum of spiroviolene.

With sufficient amount of spiroviolene in hand, we next attempted to obtain a crystalline compound suitable for X-ray diffraction by introducing functional groups (e.g., a hydroxy group, or a keto group) for further derivatization. Spiroviolene was not transformed when subjected to conditions for allylic oxidation (SeO2) even at elevated temperature [27], and the starting material was fully recovered.

We have also tried hydroboration/oxidation conditions for transforming the double bond in a congested environment of spiroviolene (Scheme 2). Low conversion was observed when a tetrahydrofuran (THF) solution of 1 was treated with BH3·THF at ambient temperature. The hydroboration reaction could be driven to synthetically useful yield when 1 was directly dissolved in 1 M BH3·THF in THF, and heated at 60 °C for 3 days. After oxidative treatment of the resultant alkylborane products with NaOH/H2O2, we have obtained three derivatives 911 with one hydroxy group in 37%, 30% and 21% isolated yield, respectively, as well as recovery of 10% of the starting material. After detailed analysis of the NMR spectra, we have found that besides the normal hydroboration/oxidation product 1α-hydroxyspiroviolane (9), 9α- (10) and 9β-hydroxyspiroviolane (11) resulting from a formal boration at the homoallylic C9 position were also produced. The stereochemistry of the newly generated stereocenters was elucidated on the basis of their NOESY spectra. Thus, the key NOE correlations of H-1/H-6, H-1/H3-19, and H-2/H3-20 of 9 allowed to assign the 1-OH to be α-oriented, while correlations of H-9/H-14 of 10, and H-9/H-3 of 11, supported the assignment of 9α- and 9β-oriented hydroxy groups, respectively.


Scheme 2: Derivatization of spiroviolene for X-ray crystallography. A) Hydroboration/oxidation reaction of spiroviolene involving a borane migration process. B) Synthesis of the hydrazone derivative of spiroviolene for single-crystal X-ray diffraction.

The formation of all three products 9–11 can be explained as follows (Scheme 2A) [28-31]: Due to the favorable formation of cis-5,5-fused B/C ring system, the borane reagent is preferred to approach the double bond of 1 from the α-face, to give either a secondary 1-organoborane intermediate IM-13, or a tertiary 2-organoborane intermediate IM-14. Oxidation of IM-13 with H2O2/NaOH could stereoselectively furnish the normal hydroboration/oxidation product 9. On the other hand, the unstable tertiary 2-organoborane IM-14 would undergo two consecutive thermal 1,2-boron migrations to give the 9-organoborane intermediate IM-16 probably through borane–olefin complexes. The suprafacial nature of the boron migration allowed the boron to be α-oriented in intermediate IM-16, which would give 10 with retention of the configuration after NaOH/H2O2 oxidation. The formation of a significant amount of the C-9 epimerization product 11 might be explained by a competing radical oxidation of the organoborane intermediate IM-16 by oxygen when the oxidation step is opened to air [32,33]. On the other hand, an alternative mechanism for the formation of 11 involving the oxidation of the 9-epi-IM-16 organoborane intermediate, which was formed by elimination of BH3 from IM-16 followed by re-addition of BH3 from the opposite β-face to the proposed C8–C9 double bond intermediate, would also be possible.

To further advance the intermediate to crystalline hydrazone product (Scheme 2B), we have found that both 10 and 11 can be oxidized to the same 9-oxospiroviolane (12) in the same 92% isolated yield, hence confirming the structural assignment of 10 and 11. By reacting with 2,4-dinitrophenylhydrazine [34], ketone 12 was further converted to hydrazone derivative 13, which gave a brownish-yellow crystal suitable for X-ray diffraction [35]. The crystal structure of 13 clearly showed that the 19- and 20-methyl groups are cis-oriented in the D-ring which is consistent with that of spirograterpene A. This structural data reaffirms the revised structure of spiroviolene, and further support the unified cyclization process of fungi-derived deoxyconidiogenol and bacteria-derived spiroviolene by sharing the common C6-cation intermediate IM-3 with cyclopiane skeleton (Scheme 1A).


We have unambiguously confirmed the structural revision of spiroviolene with cis-oriented 19- and 20-methyl groups by obtaining a suitable crystal of the hydrazone derivative of 9-oxospiroviolane, which was synthesized by hydroboration/oxidation of spiroviolene followed by converting the resultant hydroxy group to a keto group, for single-crystal X-ray diffraction. The similar skeleton and conserved stereochemistry of spiroviolene and spirograterpene A therefore indicated that there must exist two terpene cyclases of different origins (Actinomycetes and Fungus) that were able to carry out similar chemical processes for the formation of the intriguing 5-5-5-5 tetracyclic ring system. Also, our study supports the proposed unified cyclization processes of spiroviolene and deoxyconidiogenol that bifurcate at the C6-cation intermediate IM-3 with a cyclopiane skeleton. Thus, further mutational studies of these related terpene cyclases would give us more insights into the complex cyclization processes.

Supporting Information

Supporting Information File 1: Materials, synthetic methods, and copies of NMR spectra for all compounds.
Format: PDF Size: 4.6 MB Download
Supporting Information File 2: Crystallographic information file of compound 13.
Format: CIF Size: 997.3 KB Download


We are grateful to the National Natural Science Foundation of China (No. 81973197, 81991525, 22107008), Beijing Natural Science Foundation (No. Z210007), and Peking University Medicine Fund for world’s leading discipline (BMU2022DJXK002) for financial support.

Data Availability Statement

All data that supports the findings of this study is available in the published article and/or the supporting information to this article.


  1. Christianson, D. W. Chem. Rev. 2017, 117, 11570–11648. doi:10.1021/acs.chemrev.7b00287
    Return to citation in text: [1]
  2. Minami, A.; Ozaki, T.; Liu, C.; Oikawa, H. Nat. Prod. Rep. 2018, 35, 1330–1346. doi:10.1039/c8np00026c
    Return to citation in text: [1]
  3. Rudolf, J. D.; Alsup, T. A.; Xu, B.; Li, Z. Nat. Prod. Rep. 2021, 38, 905–980. doi:10.1039/d0np00066c
    Return to citation in text: [1]
  4. Christianson, D. W. Curr. Opin. Chem. Biol. 2008, 12, 141–150. doi:10.1016/j.cbpa.2007.12.008
    Return to citation in text: [1]
  5. Nicolaou, K. C.; Snyder, S. A. Angew. Chem., Int. Ed. 2005, 44, 1012–1044. doi:10.1002/anie.200460864
    Return to citation in text: [1]
  6. Rabe, P.; Rinkel, J.; Dolja, E.; Schmitz, T.; Nubbemeyer, B.; Luu, T. H.; Dickschat, J. S. Angew. Chem., Int. Ed. 2017, 56, 2776–2779. doi:10.1002/anie.201612439
    Return to citation in text: [1] [2] [3] [4] [5]
  7. Lauterbach, L.; Rinkel, J.; Dickschat, J. S. Angew. Chem., Int. Ed. 2018, 57, 8280–8283. doi:10.1002/anie.201803800
    Return to citation in text: [1]
  8. Hendrikse, N. M.; Charpentier, G.; Nordling, E.; Syrén, P.-O. FEBS J. 2018, 285, 4660–4673. doi:10.1111/febs.14686
    Return to citation in text: [1]
  9. Schriever, K.; Saenz-Mendez, P.; Rudraraju, R. S.; Hendrikse, N. M.; Hudson, E. P.; Biundo, A.; Schnell, R.; Syrén, P.-O. J. Am. Chem. Soc. 2021, 143, 3794–3807. doi:10.1021/jacs.0c10214
    Return to citation in text: [1]
  10. Niu, S.; Fan, Z.-W.; Xie, C.-L.; Liu, Q.; Luo, Z.-H.; Liu, G.; Yang, X.-W. J. Nat. Prod. 2017, 80, 2174–2177. doi:10.1021/acs.jnatprod.7b00475
    Return to citation in text: [1]
  11. Bian, G.; Han, Y.; Hou, A.; Yuan, Y.; Liu, X.; Deng, Z.; Liu, T. Metab. Eng. 2017, 42, 1–8. doi:10.1016/j.ymben.2017.04.006
    Return to citation in text: [1] [2] [3]
  12. Chi, H. M.; Cole, C. J. F.; Hu, P.; Taylor, C. A.; Snyder, S. A. Chem. Sci. 2020, 11, 10939–10944. doi:10.1039/d0sc04686h
    Return to citation in text: [1] [2]
  13. Xu, H.; Dickschat, J. S. ChemBioChem 2021, 22, 850–854. doi:10.1002/cbic.202000682
    Return to citation in text: [1]
  14. Sato, H.; Takagi, T.; Miyamoto, K.; Uchiyama, M. Chem. Pharm. Bull. 2021, 69, 1034–1038. doi:10.1248/cpb.c21-00536
    Return to citation in text: [1]
  15. Mitsuhashi, T.; Kikuchi, T.; Hoshino, S.; Ozeki, M.; Awakawa, T.; Shi, S.-P.; Fujita, M.; Abe, I. Org. Lett. 2018, 20, 5606–5609. doi:10.1021/acs.orglett.8b02284
    Return to citation in text: [1]
  16. Shiina, T.; Nakagawa, K.; Fujisaki, Y.; Ozaki, T.; Liu, C.; Toyomasu, T.; Hashimoto, M.; Koshino, H.; Minami, A.; Kawaide, H.; Oikawa, H. Biosci., Biotechnol., Biochem. 2019, 83, 192–201. doi:10.1080/09168451.2018.1536518
    Return to citation in text: [1]
  17. Rinkel, J.; Steiner, S. T.; Bian, G.; Chen, R.; Liu, T.; Dickschat, J. S. ChemBioChem 2020, 21, 486–491. doi:10.1002/cbic.201900462
    Return to citation in text: [1] [2]
  18. Dickschat, J. S. Angew. Chem., Int. Ed. 2019, 58, 15964–15976. doi:10.1002/anie.201905312
    Return to citation in text: [1]
  19. Rinkel, J.; Steiner, S. T.; Dickschat, J. S. Angew. Chem., Int. Ed. 2019, 58, 9230–9233. doi:10.1002/anie.201902950
    Return to citation in text: [1]
  20. Xing, B.; Xu, H.; Li, A.; Lou, T.; Xu, M.; Wang, K.; Xu, Z.; Dickschat, J. S.; Yang, D.; Ma, M. Angew. Chem., Int. Ed. 2022, 61, e202209785. doi:10.1002/anie.202209785
    Return to citation in text: [1]
  21. Chatzivasileiou, A. O.; Ward, V.; Edgar, S. M.; Stephanopoulos, G. Proc. Natl. Acad. Sci. U. S. A. 2019, 116, 506–511. doi:10.1073/pnas.1812935116
    Return to citation in text: [1] [2]
  22. Clomburg, J. M.; Qian, S.; Tan, Z.; Cheong, S.; Gonzalez, R. Proc. Natl. Acad. Sci. U. S. A. 2019, 116, 12810–12815. doi:10.1073/pnas.1821004116
    Return to citation in text: [1] [2]
  23. Lund, S.; Hall, R.; Williams, G. J. ACS Synth. Biol. 2019, 8, 232–238. doi:10.1021/acssynbio.8b00383
    Return to citation in text: [1] [2]
  24. Couillaud, J.; Rico, J.; Rubini, A.; Hamrouni, T.; Courvoisier-Dezord, E.; Petit, J.-L.; Mariage, A.; Darii, E.; Duquesne, K.; de Berardinis, V.; Iacazio, G. ACS Omega 2019, 4, 7838–7849. doi:10.1021/acsomega.9b00561
    Return to citation in text: [1] [2]
  25. Johnson, L. A.; Dunbabin, A.; Benton, J. C. R.; Mart, R. J.; Allemann, R. K. Angew. Chem., Int. Ed. 2020, 59, 8486–8490. doi:10.1002/anie.202001744
    Return to citation in text: [1] [2]
  26. Couillaud, J.; Leydet, L.; Duquesne, K.; Iacazio, G. Genes 2021, 12, 1974. doi:10.3390/genes12121974
    Return to citation in text: [1] [2]
  27. Warpehoski, M. A.; Chabaud, B.; Sharpless, K. B. J. Org. Chem. 1982, 47, 2897–2900. doi:10.1021/jo00136a017
    Return to citation in text: [1]
  28. Brown, H. C.; Zweifel, G. J. Am. Chem. Soc. 1960, 82, 1504–1505. doi:10.1021/ja01491a058
    Return to citation in text: [1]
  29. Lhermitte, F.; Knochel, P. Angew. Chem., Int. Ed. 1998, 37, 2459–2461. doi:10.1002/(sici)1521-3773(19981002)37:18<2459::aid-anie2459>;2-o
    Return to citation in text: [1]
  30. Laaziri, H.; Bromm, L. O.; Lhermitte, F.; Gschwind, R. M.; Knochel, P. J. Am. Chem. Soc. 1999, 121, 6940–6941. doi:10.1021/ja9843977
    Return to citation in text: [1]
  31. Hupe, E.; Denisenko, D.; Knochel, P. Tetrahedron 2003, 59, 9187–9198. doi:10.1016/j.tet.2003.09.043
    Return to citation in text: [1]
  32. Pelter, A.; Smith, K. Oxidation of carbon-boron bonds. In Comprehensive Organic Synthesis; Trost, B. M., Ed.; Pergamon Press: Oxford, UK, 1991; Vol. 7, pp 593–611. doi:10.1016/b978-0-08-052349-1.00203-1
    Return to citation in text: [1]
  33. Brown, H. C.; Midland, M. M.; Kabalka, G. W. J. Am. Chem. Soc. 1971, 93, 1024–1025. doi:10.1021/ja00733a040
    Return to citation in text: [1]
  34. Hearn, M. J.; Lebold, S. A.; Sinha, A.; Sy, K. J. Org. Chem. 1989, 54, 4188–4193. doi:10.1021/jo00278a036
    Return to citation in text: [1]
  35. Deposition number CCDC 2274944 for 13 contain the supplementary crystallographic data for this paper. These data are provided free of charge by the Cambridge Crystallographic Data Centre.
    Return to citation in text: [1]
Other Beilstein-Institut Open Science Activities