Mechanistic investigations on multiproduct β-himachalene synthase from Cryptosporangium arvum

Article abstract of DOI:10.3762/bjoc.15.99

A bacterial terpene synthase from Cryptosporangium arvum was characterised as a multiproduct β-himachalene synthase. In vitro studies showed not only a high promiscuity with respect to its numerous sesquiterpene products, including the structurally demanding terpenes longicyclene, longifolene and α-longipinene, but also to its substrates, as additional activity was observed with geranyl- and geranylgeranyl diphosphate. In-depth mechanistic investigations using isotopically labelled precursors regarding the stereochemical course of both 1,11-cyclisation and 1,3-hydride shift furnished a detailed catalytic model suggesting the molecular basis of the observed low product selectivity. The enzyme's synthetic potential was also exploited in the preparation of sesquiterpene isotopomers, which provided insights into their EIMS fragmentation mechanisms.

Full text of DOI:10.3762/bjoc.15.99

Mechanistic investigations on multiproduct β-himachalene
synthase from Cryptosporangium arvum
Jan Rinkel and Jeroen S. Dickschat*
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2019, 15, 1008–1019.
Kekulé-Institute for Organic Chemistry and Biochemistry, University of
Bonn, Gerhard-Domagk-Str. 1, 53121 Bonn, Germany
doi:10.3762/bjoc.15.99
Received: 12 March 2019
Accepted: 26 April 2019
Published: 02 May 2019
Email:
Jeroen S. Dickschat* - dickschat@uni-bonn.de
*
Corresponding author
This article is part of the thematic issue "Terpenes".
Associate Editor: D. Spring
Keywords:
enzyme mechanisms; isotopes; mass spectrometry; promiscuity;
terpenes
© 2019 Rinkel and Dickschat; licensee Beilstein-Institut.
License and terms: see end of document.
Abstract
A bacterial terpene synthase from Cryptosporangium arvum was characterised as a multiproduct β-himachalene synthase. In vitro
studies showed not only a high promiscuity with respect to its numerous sesquiterpene products, including the structurally
demanding terpenes longicyclene, longifolene and α-longipinene, but also to its substrates, as additional activity was observed with
geranyl- and geranylgeranyl diphosphate. In-depth mechanistic investigations using isotopically labelled precursors regarding the
stereochemical course of both 1,11-cyclisation and 1,3-hydride shift furnished a detailed catalytic model suggesting the molecular
basis of the observed low product selectivity. The enzyme’s synthetic potential was also exploited in the preparation of sesquiter-
pene isotopomers, which provided insights into their EIMS fragmentation mechanisms.
Introduction
The organic chemist usually prefers to work with pure com- and verbenol, for which synergistic effects were observed com-
pounds which lead to high requirements for the selectivity of pared to the single compounds [2]. Also the sex pheromone of
reactions and often to tedious purification procedures, but en- the cranberry white grub Phyllophaga anxia was identified as a
countering a pure compound in nature is quite rare. This does compound mixture, consisting of L-valine methyl ester and
not result in reduced requirements for enzyme selectivity. The L-isoleucine methyl ester at a 3:1 ratio [3]. Moreover, if there is
very opposite is mostly true, because proteins working in a a single enzyme that can produce a beneficial mixture, the
compound mixture need to be precise [1]. However, in some advantage for the producing organism is even higher. Therefore,
cases, compound mixtures have proven to be superior to the selectivity is not in every case the highest goal for evolution. An
properties of the single compounds by evolution. Examples enzyme class, which is highly prone to a regulation of product
demonstrating this principle can be found in pheromone chem- selectivity for the production of either one or multiple com-
istry, like the bark beetle aggregation blend of ipsdienol, ipsenol pounds, are terpene synthases (TSs). These enzymes are able to
1008

Beilstein J. Org. Chem. 2019, 15, 1008–1019.
guide complex cascade reactions from structurally simple oligo-
prenyl diphosphates to often complex, polycyclic products [4-6]
circumventing the low selectivity observed for carbocationic
reactions by a defined active-site architecture. Although these
enzymes are mostly highlighted for their great product selec-
tivity, TSs producing only one compound are by far not the
general case. Mostly, the main product is accompanied by
several side products. Prominent examples are the TS identified
from the plant Medicago truncatula with at least 27 products
[
[
7], γ-humulene synthase from Abies grandis with 52 products
8], and also the long known trichodiene synthase from
Fusarium sporotrichioides produces at least 15 sesquiterpenes
[
[
9]. Some TSs can even accept multiple chain length substrates
10], a concept which seems to occur frequently in plants [11].
Whether the reduced selectivity of TSs both for substrates and
for products can be attributed to imperfect catalysis, or if this
function is even beneficial for the producing organism, remains
elusive in most cases. Also the structural basis of promiscuous
catalysis by TSs is largely unknown [12]. In this study, we
present the characterisation of a bacterial TS with a reduced
selectivity both for substrates and for products together with the
challenging investigation of its cyclisation mechanism by
labelling experiments.
Results and Discussion
A bacterial β-himachalene synthase
produces numerous side products
Apart from the recently assigned (Z)-γ-bisabolene synthase
(
BbS) [13], the soil-dwelling actinomycete Cryptosporangium
arvum DSM 44712 also possesses a second TS gene (accession
no. WP_035852539). Its encoded amino acid sequence (Figure
S1, Supporting Information File 1) shares conserved motifs for
TSs, but is phylogenetically distant to BbS and does not pos-
sess a close characterised relative among other bacterial TSs
Figure 1: Total ion chromatograms of hexane extracts from the incu-
bations of HcS with A) FPP, B) GPP and C) GGPP. Peak numbering
refers to the compounds shown in Figure 2 and listed in Table 1.
(
Figure S2, Supporting Information File 1). Therefore, its gene
was cloned into the E. coli expression vector pYE-Express [14]
for functional characterisation (Table S1, Supporting Informa-
tion File 1). The purified recombinant protein (Figure S3, Sup- In a large scale incubation, β-himachalene (1) was isolated,
porting Information File 1) was incubated with the common TS accompanied by smaller amounts of the double oxidation prod-
substrates geranyl- (GPP, C10), farnesyl- (FPP, C15), geranyl- uct γ-dehydro-ar-himachalene (9). Since 9 was only observed
geranyl- (GGPP, C20) and geranylfarnesyl (GFPP, C25) diphos- after prolonged incubation times, an auto-oxidation mechanism
phate. Whereas the latter diphosphate did not lead to any involving oxygen is assumed. Both compounds were analysed
terpene product, the incubation with FPP showed a smooth by one- and two-dimensional NMR spectroscopy (Tables S2
conversion into several sesquiterpenes (Figure 1A) with com- and S3, Supporting Information File 1). The absolute configura-
pound 1 as the major peak after GC–MS analysis. However, tion of 1 was determined as the (+)-enantiomer, unanimously by
also the incubations with GPP (Figure 1B) and GGPP optical rotary power measurement and an isotopic labelling
(
Figure 1C) led to several less complex terpene products, strategy, which involved conversion of stereoselectively deuter-
demonstrating a broadened substrate range for this enzyme. The ated and at the same position 13C-labelled FPPs by the TS to
annotated peaks were correlated by mass spectral libraries and yield labelled 1 with incorporation of deuterium into diastereo-
retention indices (Table 1) to the known natural products 1–8 topic hydrogen positions. Together with the relative configura-
and 10–18 (Figure 2).
tion of the targeted methylene group as deduced by NOESY,
1009

Beilstein J. Org. Chem. 2019, 15, 1008–1019.
Table 1: HcS product identification by retention indices.
compound
Ia
I (lit.)
from GPP
α-pinene (13)
β-myrcene (14)
limonene (10)
α-terpinolene (12)
linalool (15)
934
934 [15]
992
992 [16]
1029
1089
1099
1191
1250
1031 [17]
1088 [18]
1098 [19]
1190 [19]
1253 [20]
α-terpineol (11)
geraniol (16)
from FPP
α-longipinene (6)
longicyclene (5)
longifolene (4)
1355
1376
1412
1461
1475
1489
1490
1507
1356 [21]
1377 [21]
1413 [21]
1461 [22]
1471 [23]
1487 [24]
1489 [22]
1503 [25]
α-himachalene (2)
9-epi-β-caryophyllene (7)
γ-humulene (8)
γ-himachalene (3)
β-himachalene (1)
from GGPP
β-springene (18)
cembrene A (17)
1921
1974
1918 [26]
1979 [27]
aRetention index I on a HP-5MS column.
isopentenyl diphosphate (IPP) by farnesyl diphosphate synthase
FPPS) from Streptomyces coelicolor [29] with a known stereo-
(
chemical course [30] (Figure S4, Supporting Information
File 1). This principle was also applied to C-4 and C-8 utilising
(
Z)- and (E)-(4-13C,4-2H)IPP [31] for the elongation of
dimethylallyl diphosphate (DMAPP) catalysed by FPPS in a
known course [32]. Since both hydrogens at C-4 possess the
same chemical shift, only C-8 could be used to solidify the
absolute configuration (Figure S5, Supporting Information
File 1).
GC analysis on a homochiral stationary phase was used to
assign the absolute configurations of the observed chiral
monoterpenes (R)-(+)-limonene (10), (R)-(+)-α-terpineol (11),
Figure 2: Structures of HcS products arising A) from FPP together
with related oxidation product 9, B) from GPP and C) from GGPP. The
carbon numberings of 1 and 9 refer to the carbon positions of FPP as
shown in Scheme 3, numberings of 10 and 17 are derived from that of
GPP and GGPP, respectively. Compounds known to also originate
from non-enzymatic hydrolysis are labelled with an asterisk. The enan-
tiomeric excess values were determined based on GC analysis on a
chiral phase.
(
+)-α-pinene (13) and (S)-(+)-linalool (15) as shown in Figure 2
by comparison with commercially available standards (Figures
S6–S9, Supporting Information File 1). The non-enzymatic deg-
radation of GPP as a background reaction to 15 resulted in a
substantial loss of stereoinformation for this compound (7% ee).
the stereochemical outcome of these experiments, which can be Also the cyclised products 10, 11 and 13 where not obtained in
easily monitored by sensitive HSQC, infers the absolute config- enantiomerically pure form (ee values were varying between
uration of 1. For 1, C-5 was targeted by (1R)- and (1S)-(1- 64% and 88%, as judged by integration), which may point to
1
3C,1-2H)GPP [28], which were enzymatically elongated with different possible binding and folding modes within the TS’s
1010

Beilstein J. Org. Chem. 2019, 15, 1008–1019.
active site for GPP involving both enantiomers of linalyl incubation with GPP and 18 in the experiment with GGPP, this
diphosphate (LPP, Scheme 1) and the terpinyl cation (A). Other TS from C. arvum is characterised as a multiproduct (+)-β-
TSs producing an enantiomeric mixture of monoterpenes are himachalene synthase (HcS) possessing additional mono- and
also known, e g., from Pinus taeda [33]. However, the major diterpene cyclase activity.
enantiomer of each cyclised monoterpene product described
The structures of its minor products reveal
herein was found to be derived from (R)-A.
the cyclisation mechanism of HcS
Since 17 is a simple 1,14-cyclisation product, and all cyclised
monoterpenes are derived from the extensively studied terpinyl
cation [35,36], this work focusses on elucidating the more inter-
esting sesquiterpene cyclase mechanism of HcS. Most
sesquiterpene products 1–6 of HcS including the main product
1
, but also α-himachalene (2), γ-himachalene (3), longifolene
4), longicyclene (5), and α-longipinene (6) are traditionally
considered to be derived from the himachalyl cation (B) [8,37]
(
(
Scheme 2).
Scheme 1: Initial steps of the cyclisation of GPP towards monoter-
pene products [34]. Both pathways are likely co-occurring in the TS to
explain the formation of both enantiomers of 10, 11 and 13, with the
major enantiomer in all cases being derived from (R)-A.
Compound 17 was isolated from a large scale incubation of the
TS with GGPP and identified by NMR as cembrene A. Chiral
phase GC analysis showed also in this case a mixture of enan-
tiomers with the major one being (–)-cembrene A (61% ee), the
Scheme 2: Late stage cyclisations of the himachalyl cation B to HcS
products 1–6. Alternative mechanistic and reaction arrows belonging to
branching points are shown in red.
enantiomer of the product obtained from a cembrene A synthase Whereas 1–3 are simple deprotonation products of B, 4 and 5
CAS) from Allokutzneria albata [27], which was used for com- require a further 3,7-ring closure, leading to the non-classical
(
parison (Figure S10, Supporting Information File 1).
cation C, which is a derivative of the 2-norbornyl cation [38].
This system either collapses by deprotonation at the methyl
Taken together, the overall more sluggish conversion of GPP group to longifolene (4), or by deprotonation at C-4 with forma-
and GGPP by the TS leading to enantiomeric mixtures, the tion of a cyclopropane ring to longicyclene (5). Starting from B,
higher biosynthetic complexity of the obtained sesquiterpenes a 2,7-ring closure and deprotonation at the same carbon atom
and the absence of spontaneous hydrolysis products in the incu- gives α-longipinene (6). For the main product 1, the deproton-
bation with FPP compared to the appearance of 14 and 15 in the ation was followed by an incubation of HcS and FPPS with
1011

Beilstein J. Org. Chem. 2019, 15, 1008–1019.
(
2-2H)GPP [39] and IPP, which resulted in unlabelled 1 as ob- In case of a deprotonation at a methylene group, relevant
served by GC–MS (Figure 3).
for the formation of compounds 3, 5 and 6, the stereochemical
course of these final steps could be followed by stereoselective
deuterations. GC–MS analysis of the products obtained from
the incubations with HcS, FPPS, DMAPP and (Z)- or
(
(
E)-(4-13C,4-2H)IPP showed a specific loss of HZ in all cases
Figure 4).
Intriguingly, all deprotonation steps leading to 1, 3, 5 and 6
proceed from the same face of B. Giving access to most prod-
ucts, cation B can be considered as the central branching point
within the HcS catalysed cyclisation mechanism. To rationalise
the formation of B starting from FPP, two different pathways
were initially assumed (Scheme 3). Both start with a 1,3-syn-
allylic rearrangement of OPP to (R)-nerolidyl diphosphate
(
NPP). This step is usually proposed to generate a (Z)-config-
ured C-2,C-3 double bond after cyclisation [40]. Following the
first mechanism (path A), a 1,11-cyclisation can yield second-
ary cation D, which either stabilises by 2,10-ring closure to give
the caryophyllenyl cation E that can be deprotonated at the
methyl group to yield 9-epi-(E)-β-caryophyllene (7), or D
undergoes a 1,3-hydride shift to the allylic cation F. Deproton-
ation leads to γ-humulene (8), but a 1,6-ring closure gives
access to B.
Figure 3: EI mass spectrum of 1 arising from an incubation of
2-2H)GPP and IPP with FPPS and HcS showing a loss of label during
deprotonation.
(
Scheme 3: Proposed cyclisation mechanism towards cation B via an initial 1,11-cyclisation (path A) and an hypothetical alternative mechanism via an
initial 1,6-cyclisation (path B). Alternative mechanistic and reaction arrows belonging to branching points are shown in red.
1012

Beilstein J. Org. Chem. 2019, 15, 1008–1019.
Figure 4: Stereochemical course of the final deprotonation step towards 3, 5 and 6 investigated by GC–MS. EI mass spectra of labelled A) 3, B) 5
and C) 6 obtained from the incubation of HcS and FPPS with DMAPP and (Z)-(4-13C,4-2H)IPP (left) or (E)-(4-13C,4-2H)IPP (right) indicating stereo-
specific loss of one hydrogen atom. 13C-Labellings are indicated by red dots.
1013

Beilstein J. Org. Chem. 2019, 15, 1008–1019.
The second shown option, path B, assumes a 1,6-ring closure of
R)-NPP to the bisabolyl cation G. Proceeding with a 1,2-
(
hydride shift to H, the key step is a 1,6-proton shift to give the
tertiary cation I. This idea is derived from a very similar proton
transfer starting from the bisabolyl cation, which occurs in the
cyclisation mechanism to trichodiene [41]. A 1,11-cyclisation
yields tertiary cation J, which can undergo a 1,2-hydride shift to
B. While path B circumvents the secondary cation intermediate
D, the HcS products 7 and 8 are hard to explain from path B.
Together with the absence of any 1,6-cyclised bisabolene
derived molecules in the product mixture their occurrence
represent experimental evidence in favour of path A.
Incubation experiments enlighten the stereo-
chemical course of the 1,11-cyclisation and
the 1,3-hydride shift
Although the general idea of path A appears to be straightfor-
ward at first sight, the details proofed to be challenging as
deeper insights for the stereochemistry of each step were ob-
tained by incubation experiments. The question, whether
(
R)-NPP or its enantiomer (S)-NPP is involved in the HcS cata-
lysed cyclisation cascade, was addressed by incubation of both
enantiomerically pure substrates and (rac)-NPP [13] with HcS.
The resulting chromatograms (Figure 5) clearly demonstrate
(
R)-NPP as an intermediate, which is a substrate for the produc-
tion of 1–8 in approximately the same ratio as with FPP. In
contrast, (S)-NPP is selectively converted into (E)-β-farnesene
(
19, I = 1460 (HP-5MS), Lit: I = 1459 (HP-5MS) [42]). The
same outcome regarding the formation of (Z)-γ-bisabolene from
R)-NPP and FPP, but of 19 from (S)-NPP was recently also ob-
(
served for BbS [13].
Figure 5: Total ion chromatogram of hexane extracts from HcS incu-
bations with A) (R)-NPP, B) (S)-NPP and C) (rac)-NPP.
Targeting the stereochemical course of the 1,11-cyclisation of
(
R)-NPP to cation D, (12-13C)- [43] and enzymatically pre-
pared (13-13C)FPP from (9-13C)GPP [39] and IPP with FPPS allows to investigate the stereochemical course of the first cycli-
were incubated with HcS to follow the fate of the geminal sation step for this compound. First, the absolute configuration
methyl groups for 1 (Figure 6). Combined with the relative ori- of 7 was assigned as shown in Figure 2 from the incubation ex-
entation of each methyl group deduced by NOESY, these exper- periments with (E)- and (Z)-(4-13C,4-2H)IPP, DMAPP, FPPS
iments showed an 11Re attack preceding the formation of D. and HcS targeting the positions C-3 and C-7 (Figure S11, Sup-
The observed absolute configurations of the monoterpenes 10, porting Information File 1), using published NMR data for 7
1
1 and 13 and of the diterpene 17 support this finding, because [44]. The stereochemical fate for the hydrogens at C-1 was then
their formation requires involvement of the same face of the ter- targeted by the incubation of (1R)- and (1S)-(1-13C,1-2H)FPP
minal isoprenoid double bond (6Si from GPP and 14Si from [28] with HcS (Figure 7).
GGPP). Therefore, a similar binding conformation for the ter-
minal C5-unit is reasonable for the three substrates.
The selective incorporation of deuterium into the diastereotopic
positions of 7 is explainable by a 1Si,11Re-cyclisation of
To complete the mechanistic picture of the initial 1,11-cyclisa- (R)-NPP. Given the absolute configuration of NPP and its for-
tion, also the stereochemical course at C-1 was investigated. mation via a 1,3-syn-allylic rearrangement from FPP, this ring
Unfortunately, this position is disturbed by the follow-up 1,3- closure represents an example of a formal syn-SN2’ reaction.
hydride shift in 1 and most products. However, in the side prod- This is an intriguing observation, since for other TSs a NPP-
uct 7 C-1 remains untouched after 1,11-cyclisation, which cyclisation by anti-SN2’ is usually described [40,45-47]. This
1014

Beilstein J. Org. Chem. 2019, 15, 1008–1019.
Figure 6: The origin of the two diastereotopic methyl groups in 1.
Partial 13C NMR spectrum of A) unlabelled 1, B) an extract obtained
from the incubation of HcS with (12-13C)FPP and C) an extract ob-
tained from the incubation of HcS with (9-13C)GPP, which is enzymati-
cally converted to (13-13C)FPP. Red dots indicate 13C labelled carbon
atoms.
cyclisation mechanism is thought to be the predominant case,
giving rise to a more energetically favoured transition state, but
occasionally also the syn-stereochemistry was observed [48].
The rather unexpected stereochemical course of the HcS-cata-
lysed cyclisation of NPP found herein therefore shows, that this
Figure 7: Stereochemical course of the 1,11-cyclisation at C-1 for 7.
Partial HSQC spectra of HcS incubation with A) (1R)-(1-13C,1-2H)FPP
and B) (1S)-(1-13C,1-2H)FPP. The reference chemical shifts for 7 are
taken from ref. [44]. Red dots represent 13C-labelled carbon atoms.
step has to be investigated for anti- versus syn-attack experi- deuterium is found at C-1 (Figure S16, Supporting Information
mentally for every single case, especially for a conformation- File 1). With the opposite stereochemical course than for 15, 14
ally flexible situation like a 1,11-cyclisation. Intriguingly, the is likely derived from the minor enantiomer (S)-LPP in analogy
stereochemical course of the initial cyclisation step can even be to 19 observed from (S)-NPP. For the diterpene 17 (Table S9,
substrate dependent. The 1,6-cyclisation towards the monoter- Supporting Information File 1), similar investigations using
penes 10, 11 and 13 as investigated by the incubation of (1S)- (1S)- and (1R)-(1-13C,1-2H)GGPP [49] with HcS resulted in the
and (1R)-(1-13C,1-2H)GPP with HcS and comparison to the expected outcome for a direct 1,14-cyclisation of GGPP (Figure
NMR data of the commercial available products (Table S4–S8, S17, Supporting Information File 1) in line with the results ob-
Supporting Information File 1) clearly obeys the anti-SN2’ case tained with CAS from A. albata for ent-17 [27]. Assuming simi-
(
Figures S12–S14, Supporting Information File 1). The obser- lar chemical shifts at C-1 for 14 and 18, the analogous signals
vation that 15 was obtained as a nearly racemic mixture for C-1 of 18 gave comparable results with the same stereo-
contrasts the far more selective incorporation of deuterium into chemical course as observed for 14, although with lower preser-
the olefinic positions at C-1 of 15 (Figure S15, Supporting vation of stereoinformation (Figure S18, Supporting Informa-
Information File 1). This result supports (R)-LPP as an interme- tion File 1).
diate, formed by a 1,3-syn-allylic rearrangement to determine
the observed stereochemical course at C-1, while the tertiary To shed light on the stereochemical course of the 1,3-hydride
diphosphate might then undergo a non-enzymatic degradation shift connecting cations D and F, a series of labelling experi-
to explain the high loss of stereoinformation in 15. Also for the ments were conducted to determine the origin of the shifting
achiral β-myrcene (14), an imbalanced incorporation of hydrogen (C-1) and its destination (C-10) for 1 (Figure 8). A
1015

Beilstein J. Org. Chem. 2019, 15, 1008–1019.
Figure 8: Investigation of the 1,3-hydride shift in the cyclisation towards 1. Partial 13C NMR spectra of A) unlabelled 1, and incubations with HcS and
B) (1R)-(1-13C,1-2H)FPP or C) (1S)-(1-13C,1-2H)FPP compared to incubations of HcS with FPPS, (2-13C)DMAPP and D) (1R)-(1-2H)IPP or E) (1S)-
(1-2H)IPP showing a movement of HR from C-1 to C-10. The singlets for B) and E) are slightly shifted to higher field (designated as Δδ) because of
the deuterium location nearby. 13C-Labelled carbon atoms are indicated by red dots. Dashed grey lines show the chemical shifts of the carbon atoms
for unlabelled 1.
comparison of the 13C NMR spectra from the incubations of FPPS and HcS resulting in the expected opposite outcome than
HcS with (1R)- and (1S)-(1-13C,1-2H)FPP, resulting in a singlet stated above, namely a triplet in the (R)-case and a singlet for
for the (R)-case and a triplet in the (S)-case indicating a direct the (S)-sample.
C–D bond, clearly demonstrated the stereospecific migration of
HR from C-1. To complete the observations also for C-10, HSQC analysis of the material obtained from the incubation of
(
2-13C)DMAPP was synthesised from (2-13C)-3-methylbut-2- (1R)-(1-2H)IPP and (2-13C)DMAPP with FPPS and HcS also
en-1-ol [43] and incubated with (1R)- or (1S)-(1-2H)IPP [50], allowed for the assignment of the newly introduced diastereo-
1016

Beilstein J. Org. Chem. 2019, 15, 1008–1019.
topic position at C-10 (Figure 9). Together with the assignment wards” pointing hydrogen atoms from different positions which
of the hydrogens by NOESY in 1, these data show a stereose- reflects the observation of the regio- and stereochemistry of the
lective incorporation of HR-1 into the Hα-position at C-10 by a deprotonations.
vanished crosspeak.
HcS provides access to labelled sesquiter-
penes for EIMS fragmentation studies
Since HcS produces a mixture of structurally interesting
sesquiterpenes, its synthetic abilities were also exploited to
study EIMS fragmentation mechanisms. Therefore, all fifteen
singly-13C labelled FPP isotopomers, either obtained by synthe-
sis or enzymatically [39,43,51], were converted with HcS to
result in mixtures of specifically labelled 1–8. The incorpora-
tion of label into 1 was checked by 13C NMR (Figure S20, Sup-
porting Information File 1) and all samples were analysed by
GC–MS. This allowed for the assignment of carbon positions to
specific EI-fragments of the corresponding mass spectrum by
observing an increase of +1 Da, if the labelled position is part of
the fragment (position specific mass shift analysis, PMA
[
41,52,53]). Although for many fragments multiple overlaying
fragmentation pathways were observed, some of them showed
clear position dependent results, which are summarised in
Figure 10. The EI mass spectra for each position and molecule
laying the basis for the presented three fragments for 1 together
with one fragment each for 4–8 are depicted in Figures
S21–S26 (Supporting Information File 1). Possible EI-fragmen-
tation mechanisms connected to them are discussed in Schemes
S1–S3 (Supporting Information File 1).
Figure 9: Stereochemical course of the 1,3-hydride shift at C-10 in 1.
Partial HSQC spectra of A) unlabelled 1 and B) labelled 1 arising from
the incubation of (1R)-(1-2H)IPP and (2-13C)DMAPP with FPPS and
HcS. Red dots indicate 13C-labelled carbon atoms.
Combining the information deduced from the extensive incuba-
tion experiments stated above, a structural model for the reac-
tive conformation of cation D is proposed (Figure S19, Support-
ing Information File 1). This intermediate, or structurally
related transition states for the corresponding concerted reac-
tions to avoid its secondary nature, are of central importance in
understanding the initial HcS catalysed cyclisation towards
cation B. The discussed conformation is imprinted by the struc-
ture of 7 with its relative conformation at the four-membered
ring system allowing for a 2Si,10Si-cyclisation to E without
major rotational changes and also reflects the short distance be-
tween HR and C-10 for the 1,3-hydride shift towards the 10Si
face leading to F. Intriguingly, the unusual syn-SN2’ ring
closure from (R)-NPP leads to the diphosphate moiety (OPP−)
being located close to the “backside” of the cyclising molecule,
which may give rise to an explanation of the multiproduct
nature of HcS. At this location, OPP− can easily abstract “back-
Figure 10: Position specific mass shift analysis for selected EIMS ions
of HcS products. Black dots represent an increased mass of the ion
(m/z = +1) in case of a 13C-labelling in this position. Proposed frag-
mentation mechanisms for these ions are presented in Schemes
S1–S3 (Supporting Information File 1).
1017

Beilstein J. Org. Chem. 2019, 15, 1008–1019.
Conclusion
In summary, a new terpene synthase from C. arvum was
ORCID® iDs
Jeroen S. Dickschat - https://orcid.org/0000-0002-0102-0631
characterised as a multiproduct (+)-β-himachalene synthase.
Accepting GPP, FPP and GGPP, HcS is a promiscuous enzyme, References
whose catalysis suffers from poor selectivity. Nevertheless, the
formation of multiple sesquiterpene products demands for a
challenging mechanistic model, which was refined by exten-
sive labelling experiments. Several interesting details were
disclosed including the stereochemical course of a 1,3-hydride
migration from C-1 to C-10 and the 1,11-cyclisation featuring
the unusual syn-SN2’ attack. Combining various aspects of the
initial cyclisation, the proposed conformer of cation D may also
rationalise the reduced selectivity of HcS by its positioning of
OPP−. Providing access to labelled isotopomers of its products,
including structurally demanding polycyclic terpenes, HcS also
served as a platform for investigating selected aspects of their
EIMS fragmentation mechanisms. The labelling experiments
performed with HcS described in this study therefore represent
an encouragement to experimentally explore and elucidate
every stereochemical detail of a terpene cyclisation mechanism
for a comprehensive picture of the complex reactions, these
amazing enzymes are able to catalyse.
1. Hedstrom, L. Enzyme Specificity and Selectivity; eLS; John Wiley &
Sons, Ltd.: Chichester, UK, 2010.
doi:10.1002/9780470015902.a0000716.pub2
Silverstein, R. M.; Rodin, J. O.; Wood, D. L. Science 1966, 154,
2
.
.
509–510.
3
Zhang, A.; Robbins, P. S.; Leal, W. S.; Linn, C. E., Jr.; Villani, M. G.;
Roelofs, W. L. J. Chem. Ecol. 1997, 23, 231–245.
doi:10.1023/b:joec.0000006356.47959.ed
4
.
.
Christianson, D. W. Chem. Rev. 2017, 117, 11570–11648.
doi:10.1021/acs.chemrev.7b00287
5
Dickschat, J. S. Nat. Prod. Rep. 2016, 33, 87–110.
doi:10.1039/c5np00102a
6. Tantillo, D. J. Angew. Chem., Int. Ed. 2017, 56, 10040–10045.
doi:10.1002/anie.201702363
7.
8.
9.
Garms, S.; Köllner, T. G.; Boland, W. J. Org. Chem. 2010, 75,
590–5600. doi:10.1021/jo100917c
Steele, C. L.; Crock, J.; Bohlmann, J.; Croteau, R. J. Biol. Chem. 1998,
73, 2078–2089. doi:10.1074/jbc.273.4.2078
5
2
Vedula, L. S.; Jiang, J.; Zakharian, T.; Cane, D. E.; Christianson, D. W.
Arch. Biochem. Biophys. 2008, 469, 184–194.
doi:10.1016/j.abb.2007.10.015
1
1
1
1
1
0.Rinkel, J.; Lauterbach, L.; Dickschat, J. S. Angew. Chem., Int. Ed.
2017, 56, 16385–16389. doi:10.1002/anie.201711142
1.Pazouki, L.; Niinemets, Ü. Front. Plant Sci. 2016, 7, No. 1019.
doi:10.3389/fpls.2016.01019
Supporting Information
2.Dixit, M.; Weitman, M.; Gao, J.; Major, D. T. ACS Catal. 2017, 7,
Experimental details for gene cloning, gene expression,
protein purification, incubation experiments with
812–818. doi:10.1021/acscatal.6b02584
3.Rinkel, J.; Dickschat, J. S. Beilstein J. Org. Chem. 2019, 15, 789–794.
doi:10.3762/bjoc.15.75
isotopically labelled precursors, preparative scale
incubation and synthesis of (2-13C)DMAPP. The amino
acid sequence of HcS, a phylogenetic tree of bacterial
terpene synthases, SDS-PAGE analysis of the recombinant
protein, listed NMR data for 1 and 9, labelling experiments
for the determination of the absolute configurations of 1
and 7, chiral phase GC analysis of 10, 11, 13, 15 and 17,
labelling experiments for the stereochemical course at C-1
of the monoterpenes and diterpenes, a graphical model for
cation D, NMR spectra for the incubations of singly
labelled FPPs with HcS, EIMS data for compounds 1 and
4.Dickschat, J. S.; Pahirulzaman, K. A. K.; Rabe, P.; Klapschinski, T. A.
ChemBioChem 2014, 15, 810–814. doi:10.1002/cbic.201300763
15.Saroglou, V.; Dorizas, N.; Kypriotakis, Z.; Skaltsa, H. D.
J. Chromatogr. A 2006, 1104, 313–322.
doi:10.1016/j.chroma.2005.11.087
1
6.Hazzit, M.; Baaliouamer, A.; Faleiro, M. L.; Miguel, M. G.
J. Agric. Food Chem. 2006, 54, 6314–6321. doi:10.1021/jf0606104
7.Karioti, A.; Hadjipavlou-Litina, D.; Mensah, M. L. K.; Fleischer, T. C.;
Skaltsa, H. J. Agric. Food Chem. 2004, 52, 8094–8098.
doi:10.1021/jf040150j
1
1
1
2
8.Yáñez, X.; Pinzón, M. L.; Solano, F.; Sánchez, L. R. Molecules 2002,
7, 712–716. doi:10.3390/70900712
4
-8 arising from these incubations and discussion of
9.Benkaci–Ali, F.; Baaliouamer, A.; Meklati, B. Y.; Chemat, F.
Flavour Fragrance J. 2007, 22, 148–153. doi:10.1002/ffj.1773
0.Yu, Y.; Huang, T.; Yang, B.; Liu, X.; Duan, G. J. Pharm. Biomed. Anal.
fragmentation mechanisms for selected ions.
Supporting Information File 1
2007, 43, 24–31. doi:10.1016/j.jpba.2006.06.037
Additional material.
21.Zeng, Y.-X.; Zhao, C.-X.; Liang, Y.-Z.; Yang, H.; Fang, H.-Z.; Yi, L.-Z.;
Zeng, Z.-D. Anal. Chim. Acta 2007, 595, 328–339.
doi:10.1016/j.aca.2006.12.022
[https://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-15-99-S1.pdf]
2
2.Andriamaharavo, N. R., Rentention Data. NIST Mass Spectrometry
Data Center, 2014.
2
3.Mesa-Arango, A. C.; Betancur-Galvis, L.; Montiel, J.; Bueno, J. G.;
Baena, A.; Durán, D. C.; Martínez, J. R.; Stashenko, E. E.
J. Essent. Oil Res. 2010, 22, 568–574.
Acknowledgements
This work was funded by the Deutsche Forschungsgemein-
doi:10.1080/10412905.2010.9700402
schaft (DI1536/7-1).
1018

Beilstein J. Org. Chem. 2019, 15, 1008–1019.
2
4.Bendiabdellah, A.; El Amine Dib, M.; Djabou, N.; Allali, H.; Tabti, B.;
Muselli, A.; Costa, J. Chem. Cent. J. 2012, 6, No. 431.
doi:10.1186/1752-153x-6-48
48.Benedict, C. R.; Lu, J. L.; Pettigrew, D. W.; Liu, J.; Stipanovic, R. D.;
Williams, H. J. Plant Physiol. 2001, 125, 1754–1765.
doi:10.1104/pp.125.4.1754
2
2
2
5.Quijano, C. E.; Salamanca, G.; Pino, J. A. Flavour Fragrance J. 2007,
49.Rinkel, J.; Rabe, P.; Chen, X.; Köllner, T. G.; Chen, F.; Dickschat, J. S.
Chem. – Eur. J. 2017, 23, 10501–10505. doi:10.1002/chem.201702704
50.Rinkel, J.; Lauterbach, L.; Dickschat, J. S. Angew. Chem., Int. Ed.
2019, 58, 452–455. doi:10.1002/anie.201812216
22, 401–406. doi:10.1002/ffj.1812
6.Heinrich, G.; Pfeifhofer, H. W.; Stabentheiner, E.; Sawidis, T.
Ann. Bot. (Oxford, U. K.) 2002, 89, 459–469. doi:10.1093/aob/mcf062
7.Rinkel, J.; Lauterbach, L.; Rabe, P.; Dickschat, J. S.
Angew. Chem., Int. Ed. 2018, 57, 3238–3241.
51.Mitsuhashi, T.; Rinkel, J.; Okada, M.; Abe, I.; Dickschat, J. S.
Chem. – Eur. J. 2017, 23, 10053–10057. doi:10.1002/chem.201702766
52.Rabe, P.; Klapschinski, T. A.; Dickschat, J. S. ChemBioChem 2016,
17, 1333–1337. doi:10.1002/cbic.201600237
doi:10.1002/anie.201800385
2
8.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
53.Rinkel, J.; Rabe, P.; Dickschat, J. S. Eur. J. Org. Chem. 2019,
351–359. doi:10.1002/ejoc.201800217
2
9.Rabe, P.; Rinkel, J.; Nubbemeyer, B.; Köllner, T. G.; Chen, F.;
Dickschat, J. S. Angew. Chem., Int. Ed. 2016, 55, 15420–15423.
doi:10.1002/anie.201608971
License and Terms
3
3
3
3
0.Thulasiram, H. V.; Poulter, C. D. J. Am. Chem. Soc. 2006, 128,
1
5819–15823. doi:10.1021/ja065573b
1.Lauterbach, L.; Rinkel, J.; Dickschat, J. S. Angew. Chem., Int. Ed.
018, 57, 8280–8283. doi:10.1002/anie.201803800
This is an Open Access article under the terms of the
Creative Commons Attribution License
2
(http://creativecommons.org/licenses/by/4.0). Please note
2.Cornforth, J. W.; Cornforth, R. H.; Popják, G.; Yengoyan, L.
J. Biol. Chem. 1966, 241, 3970–3987.
that the reuse, redistribution and reproduction in particular
requires that the authors and source are credited.
3.Phillips, M. A.; Wildung, M. R.; Williams, D. C.; Hyatt, D. C.;
Croteau, R. Arch. Biochem. Biophys. 2003, 411, 267–276.
doi:10.1016/s0003-9861(02)00746-4
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
3
4.Dickschat, J. S. Nat. Prod. Rep. 2011, 28, 1917–1936.
doi:10.1039/c1np00063b
(https://www.beilstein-journals.org/bjoc)
3
5.Srividya, N.; Davis, E. M.; Croteau, R. B.; Lange, B. M.
Proc. Natl. Acad. Sci. U. S. A. 2015, 112, 3332–3337.
doi:10.1073/pnas.1501203112
The definitive version of this article is the electronic one
which can be found at:
3
3
3
3
6.Hong, Y. J.; Tantillo, D. J. Org. Biomol. Chem. 2010, 8, 4589–4600.
doi:10.1039/c0ob00167h
doi:10.3762/bjoc.15.99
7.Isegawa, M.; Maeda, S.; Tantillo, D. J.; Morokuma, K. Chem. Sci.
2014, 5, 1555–1560. doi:10.1039/c3sc53293c
8.Moss, R. A. J. Phys. Org. Chem. 2014, 27, 374–379.
doi:10.1002/poc.3290
9.Bian, G.; Rinkel, J.; Wang, Z.; Lauterbach, L.; Hou, A.; Yuan, Y.;
Deng, Z.; Liu, T.; Dickschat, J. S. Angew. Chem., Int. Ed. 2018, 57,
15887–15890. doi:10.1002/anie.201809954
4
4
4
0.Arigoni, D. Pure Appl. Chem. 1975, 41, 219–245.
doi:10.1351/pac197541010219
1.Hong, Y. J.; Tantillo, D. J. Org. Lett. 2006, 8, 4601–4604.
doi:10.1021/ol061884f
2.Asuming, W. A.; Beauchamp, P. S.; Descalzo, J. T.; Dev, B. C.;
Dev, V.; Frost, S.; Ma, C. W. Biochem. Syst. Ecol. 2005, 33, 17–26.
doi:10.1016/j.bse.2004.06.005
4
3.Rabe, P.; Barra, L.; Rinkel, J.; Riclea, R.; Citron, C. A.;
Klapschinski, T. A.; Janusko, A.; Dickschat, J. S.
Angew. Chem., Int. Ed. 2015, 54, 13448–13451.
doi:10.1002/anie.201507615
4
4
4
4
4.Hinkley, S. F. R.; Perry, N. B.; Weavers, R. T. Phytochemistry 1994,
35, 1489–1494. doi:10.1016/s0031-9422(00)86882-x
5.Cane, D. E.; Tandon, M. J. Am. Chem. Soc. 1995, 117, 5602–5603.
doi:10.1021/ja00125a029
6.Hu, Y.; Chou, W. K. W.; Hopson, R.; Cane, D. E. Chem. Biol. 2011, 18,
32–37. doi:10.1016/j.chembiol.2010.11.008
7.Rinkel, J.; Rabe, P.; Garbeva, P.; Dickschat, J. S.
Angew. Chem., Int. Ed. 2016, 55, 13593–13596.
doi:10.1002/anie.201608042
1019