The absolute configuration of isochamigrene: New insights into the cyclisation mechanism of trichodiene synthase

Article abstract of DOI:10.1039/c8cc01744a

Isochamigrene, a side product of the trichodiene synthase, which is a key enzyme from the biosynthesis of the trichothecene mycotoxins from Fusarium spp., was enantioselectively synthesised and compared to the natural product from Fusarium sporotrichioides. As a result, its absolute configuration was assigned to (S)-isochamigrene. Implications for the recently extensively discussed cyclisation mechanism towards trichodiene and its side products are discussed.

Full text of DOI:10.1039/c8cc01744a

ChemComm
View Article Online
View Journal
COMMUNICATION
The absolute configuration of isochamigrene: new
insights into the cyclisation mechanism of
trichodiene synthase†
Cite this:DOI: 10.1039/c8cc01744a
Received 2nd March 2018,
Accepted 19th March 2018
Immo Burkhardt and Jeroen S. Dickschat
*
DOI: 10.1039/c8cc01744a
rsc.li/chemcomm
Isochamigrene, a side product of the trichodiene synthase, which is
a key enzyme from the biosynthesis of the trichothecene mycotoxins
from Fusarium spp., was enantioselectively synthesised and compared
to the natural product from Fusarium sporotrichioides. As a result, its
absolute configuration was assigned to (S)-isochamigrene. Implica-
tions for the recently extensively discussed cyclisation mechanism
towards trichodiene and its side products are discussed.
With more than 50000 known compounds, terpenoids represent
the largest class of natural products. More than 120 carbon
skeletons for sesquiterpenes are known to date, which are all
generated from the universal precursor farnesyl pyrophosphate
1
(
FPP, 1). This chemical diversification is established by terpene
2
synthases that catalyse an initial carbocation formation. The
inherent reactivity together with the enzyme-dependent sub-
strate fold triggers a cationic cyclisation cascade towards a
3
specific product. QM/MM calculations based on enzyme crystal
structures or homology models have recently deepened our under-
Scheme 1 Cyclisation of FPP (1) to 4 and 5 and further conversion of 4 to
standing of this process and even allowed for structural predic- T-2 toxin (6).
4
tions of the product. One of the most prominent sesquiterpenes
is trichodiene (4, Scheme 1), the key intermediate in the bio-
5
synthesis of trichothecenes such as T-2 toxin (6, Scheme 1). These mechanism of TS was also in the focus of two recent QM/MM
mycotoxins pose a considerable health risk on humans and studies that have investigated the dynamics of the cyclisation
animals when ingested via consumption of Fusarium infested reaction, leading to fundamentally different and controversially
6
crops. The trichodiene synthase (TS) is one of the best investi- discussed results, inter alia with respect to the absolute configu-
7
11
gated terpene synthases for which the cyclisation mechanism,
ration of the intermediate 3. The cyclisation towards 4 involves
8
9
7
mutagenesis experiments, and the enzyme structure and a series of cationic intermediates from which at least 14 side
1
0
12
kinetics have been addressed. The generally accepted cyclisation products are formed, but for several of these products the
mode proceeds from 1 via the (R)-bisabolyl cation (2, Scheme 1) absolute configurations are unknown. Knowledge about their
and the cuprenyl cation (3) as key intermediates and yields 4 after absolute configurations is desirable, because it can clarify the
7
two final methyl migrations and deprotonation. The reaction controversy about the nature of the cationic pathway intermediates.
In a previous study in our laboratories the absolute configurations
Kekul ´e -Institut f u¨ r Organische Chemie und Biochemie, Rheinische
Friedrich-Wilhelms-Universit a¨ t Bonn, Gerhard-Domagk-Straße 1,
of (À)-a-cedrene, (+)-b-cedrene, (+)-a-funebrene, (+)-b-barbatene,
R)-cuparene and (R)-b-bisabolene from Fusarium sporotrichioides
were determined that can all arise from (R)-2 (the fungus was
(
5
3121 Bonn, Germany. E-mail: dickschat@uni-bonn.de
†
Electronic supplementary information (ESI) available: Synthetic procedures,
1
3
wrongly named Fusarium verticillioides in this work). A side
characterisation of (+)-isochamigrene and synthetic intermediates; culture conditions,
enantioselective GC and GC-MS analyses; NMR spectra of synthetic compounds.
See DOI: 10.1039/c8cc01744a
product with still unknown absolute configuration is the spiro-
1
4
bicyclic compound isochamigrene (5). The formation of 5 by
This journal is ©The Royal Society of Chemistry 2018
Chem. Commun.

View Article Online
Communication
ChemComm
TS can be explained with an alternative ring expansion from 3
instead of the required methyl migration towards 4, and sub-
sequent deprotonation (Scheme 1). The absolute configuration
of 5 was initially tentatively assigned as R, because it was
assumed that the ring expansion starts from the same con-
formation of 3 – with attack of the cationic centre from the Si
1
4
face – as the methyl migration, with Re face attack. Based on
quantum chemical calculations in the gas phase Tantillo recently
noted that a 1801 rotation around the central bond in 3 has a
similar activation barrier than the ring expansion towards 5, and
15
thus both enantiomers of 5 could be reached. Here we report on
the first enantioselective synthesis of 5, determination of the
absolute configuration of the natural product from Fusarium,
and the consequences on the mechanism of TS.
The synthesis of 5 started from 2,2-dimethylcyclohexanone
1
6
17
(
7) that was alkylated with 4-iodo-2-methylbut-1-ene (12) to
yield trisubstituted cyclohexanone 8 (Scheme 2). Generation of
the sodium enolate and treatment with allyl chloroformate gave
carbonate 9 that was subjected to the Stoltz variation of the
1
8
Tsuji–Trost allylation employing chiral phosphine ligand 13,
giving rise to asymmetric ketone 10 with 82% ee (Fig. S1, ESI†).
Ring-closing metathesis with the Hoveyda–Grubbs catalyst of
the second generation gave access to 11. Methylenation of
Fig. 1 Enantiomeric composition of 5 by GC-MS on a homochiral stationary
phase. (a) Synthetic (rac)-5; (b) synthetic (+)-(R)-5; (c) F. sporotrichioides
1
9
this sterically hindered ketone was achieved using doubly headspace extract; (d) coinjection of (a) and (c); (e) coinjection of (b) and (c).
2
0
deprotonated dimethyldiphenylphosphonium iodide (14).
The peak marked by asterisk is (E)-b-farnesene.
Isochamigrene (R)-5 was obtained in a yield of 13% via five
steps and showed identical NMR data to the natural product
from F. sporotrichioides.
To demonstrate the separability of the enantiomers of 5 on
a homochiral stationary GC phase (Fig. 1) the racemate was
1
4
synthesised via the same route using PPh
in the Tsuji–Trost allylation. Enantioselectively synthesised
R)-5 was obtained with 82% ee and the natural product from
F. sporotrichioides that is present in headspace extracts
3
instead of ligand 13
(
1
3
prepared using a closed loop stripping apparatus (CLSA)
was unambiguously identified as (S)-5. Small amounts of
b-chamigrene (15) were also detected in the extracts that coeluted
with commercial (R)-(À)-15 on a chiral GC column (Fig. S2, ESI†).
Since no standard of (S)-15 was available, the absolute configu-
ration of the natural product was tentatively assigned as (R)-15.
Our previous identification of (R)-cuparene (16), the oxidation
product of (R)-a-cuprenene, in Fusarium headspace extracts
together with the stereochemical course of the methyl group
migrations in the biosynthesis of 4 as determined in feeding
13
experiments are in agreement with a terpene cyclisation mecha-
nism via the (S)-cuprenyl cation. Interestingly, the three products 4,
(R)-15 and (S)-5 can be formed from the different conformers 3a, 3b
and 3c of the cuprenyl cation, with migration of one of the three
alkyl substituents at the quarternary carbon next to the cationic
centre (Scheme 3). For optimal orbital overlap the empty p orbital
at the cationic centre and the s orbital for the bond to the
migrating alkyl group should be oriented in parallel. The formation
of the three products 4, (R)-15 and (S)-5 requires only a minimal
conformational change of 3 with rotation around the central bond
by no more than 1201. QM/MM calculations by Major suggested
that formation of the chamigrenyl cation is sterically disfavoured in
Scheme 2 Synthesis of (+)-(R)-5. LDA = lithium diisopropylamide, THF =
tetrahydrofuran, HMPA = hexamethylphosphoramide, NaHMDS = sodium
11a
the enzyme, which is reflected by the only small amounts of 15
found in the F. sporotrichioides headspace extracts.
(bistrimethylsilyl)amide, HG-II = Hoveyda–Grubbs 2nd generation catalyst,
n-BuLi = n-butyllithium.
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2018

View Article Online
ChemComm
Communication
13
cuprenyl cation reported here and in our previous work can be
explained from the (S)-cuprenyl cation as a common intermediate.
1
1a
This intermediate is also supported by QM/MM studies by Major,
11b
while the (R)-cuprenyl cation suggested by Wu and coworkers can
explain the formation of only some products including 4, (R)-15 and
S)-5, but not (R)-cuparene (16). For the formation of (R)-15 and
1
3
(
(S)-5 from hypothetical ent-3 the ring expansion reactions shown in
Scheme 3 must proceed with alkyl group migration to the respective
opposite face of the cationic centre, while the biosynthesis of 4 from
ent-3 is difficult to understand with respect to the known course of
13
methyl group migrations. In summary, our experimental findings
for the absolute configurations of the trichodiene synthase side
products favour the (S)-cuprenyl cation (3) as a common
intermediate.
Scheme 3 Conformations of 3that explain the formation of 4,(R)-15 and (S)-5.
This work was funded by the DFG through grant DI1536/9-1.
Another trace compound in the F. sporotrichioides headspace Conflicts of interest
extracts was identified as (À)-thujopsene (20) by GC-MS with an
There are no conflicts to declare.
enantioselective stationary phase in this work (Fig. S3, ESI†).
This compound can be formed from the (S)-chamigrenyl cation
References
(17), the precursor of (R)-15, by ring expansion to 18, cyclisation
1
2
T. A. Klapschinski, P. Rabe and J. S. Dickschat, Angew. Chem., Int.
Ed., 2016, 55, 10141–10144.
J. S. Dickschat, Nat. Prod. Rep., 2011, 28, 1917–1936.
to 19 and deprotonation (Scheme 4). An alternative formation
of 20 is possible from the (S)-isochamigrenyl cation (21), the
precursor of (S)-5, by ring expansion to 22, methyl migration to
ent-18, cyclisation to ent-19 and deprotonation, but this mechanism
results in the wrong enantiomer of 20.
3 D. J. Tantillo, Angew. Chem., Int. Ed., 2017, 56, 10040–10045.
(a) D. T. Major, Y. Freud and M. Weitman, Curr. Opin. Chem. Biol.,
014, 21, 25–33; (b) P. Schrepfer, A. B u¨ ttner, C. G ¨o rner, M. Hertel,
4
2
J. van Rijn, F. Wallrapp, W. Eisenreich, V. Sieber, R. Kourist and
T. Br u¨ ck, Proc. Natl. Acad. Sci. U. S. A., 2016, 113, E958–E967;
In conclusion, the first enantioselective synthesis of (R)-(+)-
8a
(
c) J.-Y. Chow, B.-X. Tian, G. Ramamoorthy, B. S. Hillerich,
isochamigrene (5), the initially proposed structure for natural 5,
R. D. Seidel, S. C. Almo, M. P. Jacobson and C. D. Poulter, Proc.
Natl. Acad. Sci. U. S. A., 2015, 112, 5661–5666.
is presented in this work. Comparison to naturally occurring
isochamigrene showed that F. sporotrichioides exclusively produces
5 L. O. Zamir, M. J. Gauthier, K. A. Devor, Y. Nadeau and F. Sauriol,
J. Chem. Soc., Chem. Commun., 1989, 598–600.
(
(
S)-5. Two other trace compounds were identified as (R)-chamigrene
15) and (À)-thujopsene (20). The biosynthesis of 4 and all side
products with known absolute configurations formed via the
6
7
V. I. Shifrin and P. Anderson, J. Biol. Chem., 1999, 274, 13985–13992.
(a) Y. J. Hong and D. J. Tantillo, Org. Lett., 2006, 8, 4601–4604;
(b) D. E. Cane, J. L. Pawlak and R. M. Horak, Biochemistry, 1990, 29,
5
476–5490.
8
9
(a) D. E. CaneandQ. Xue, J. Am. Chem. Soc., 1996, 118, 1563–1564; (b) D. E.
Cane, Q. Xue and B. C. Fitzsimons, Biochemistry, 1996, 35, 12369–12376.
(a) M. J. Rynkiewicz, D. E. Cane and D. W. Christianson, Proc. Natl.
Acad. Sci. U. S. A., 2001, 98, 13543–13548; (b) L. S. Vedula,
M. J. Rynkiewicz, H.-J. Pyun, R. M. Coates, D. E. Cane and
D. W. Christianson, Biochemistry, 2005, 44, 6153–6163.
1
0 D. E. Cane, H.-T. Chiu, P.-H. Liang and K. S. Anderson, Biochemistry,
997, 36, 8332–8339.
1 (a) M. Dixit, M. Weitman, J. Gao and D. T. Major, ACS Catal., 2017, 7,
12–818; (b) Y.-H. Wang, H. Xie, J. Zhou, F. Zhang and R. Wu, ACS
1
1
8
Catal., 2017, 7, 5841–5846; (c) M. Dixit, M. Weitman, J. Gao and
D. T. Major, ACS Catal., 2018, 8, 1371–1375; (d) Y.-H. Wang,
F. Zhang, J. Zhou, H. Xie and R. Wu, ACS Catal., 2018, 8, 1363–1370.
2 L. S. Vedula, J. Jiang, T. Zakharian, D. E. Cane and D. W. Christianson,
Arch. Biochem. Biophys., 2008, 469, 184–194.
1
1
1
3 J. S. Dickschat, N. L. Brock, C. A. Citron and B. Tudzynski, ChemBioChem,
2
011, 12, 2088–2095.
4 D. E. Cane, Q. Xue and J. E. Van Epp, J. Am. Chem. Soc., 1996, 118,
499–8500.
8
1
1
5 Y. Hong and D. J. Tantillo, J. Am. Chem. Soc., 2014, 136, 2450–2463.
6 S.-I. Nakano, Y. Todoroki, N. Hirai and H. Ohigashi, Biosci., Biotech-
nol., Biochem., 1995, 59, 1699–1706.
1
7 H. Helmboldt, D. K ¨o hler and M. Hiersemann, Org. Lett., 2006, 8,
1
573–1576.
1
1
8 D. C. Behenna and B. M. Stoltz, J. Am. Chem. Soc., 2004, 126, 15044–15045.
9 D. E. White, I. C. Stewart, R. H. Grubbs and B. M. Stoltz, J. Am. Chem.
Soc., 2008, 130, 810–811.
Scheme 4 Cyclisation towards thujopsene (20) via (a) the (S)-chamigrenyl
cation (17) or (b) the (S)-isochamigrenyl cation (21) resulting in either (À)-20 20 M. G. Banwell, M. D. McLeod and A. G. Riches, Aust. J. Chem., 2004,
or (+)-20.
57, 53–66.
This journal is ©The Royal Society of Chemistry 2018
Chem. Commun.