The amino-terminal segment in the β-domain of δ-cadinene synthase is essential for catalysis

Article abstract of DOI:10.1039/c6ob01398h

Despite its distance from the active site the flexible amino-terminal segment (NTS) in the β-domain of the plant sesquiterpene cyclase δ-cadinene synthase (DCS) is essential for active site closure and desolvation events during catalysis.

Full text of DOI:10.1039/c6ob01398h

Organic &
Biomolecular Chemistry
View Article Online
View Journal
COMMUNICATION
The amino-terminal segment in the β-domain of
δ-cadinene synthase is essential for catalysis†
Cite this: DOI: 10.1039/c6ob01398h
Received 28th June 2016,
Accepted 13th July 2016
Verónica González, Daniel J. Grundy, Juan A. Faraldos and Rudolf K. Allemann*
DOI: 10.1039/c6ob01398h
www.rsc.org/obc
Despite its distance from the active site the flexible amino-terminal lytic function of β-domains is unclear. The amino-terminal
segment (NTS) in the β-domain of the plant sesquiterpene cyclase segment (NTS) of β-domains appears to be disordered in un-
δ-cadinene synthase (DCS) is essential for active site closure and liganded open structures, i.e. in the absence of Mg2+ ions,
desolvation events during catalysis.
unreactive FDP analogues or mimics of reactive carbocationic
intermediates. However upon ligand binding to the α-domain,
the active site adopts a closed conformation and the NTS
becomes well ordered. Complexation of ligands leads to the
formation of stabilizing α–β interdomain hydrogen bond inter-
actions that anchor the flexible NTS segment to the dynamic
A–C and J–K loops in the closed conformation.9 The conserved
amino-terminal segments may therefore play a role in stabilis-
ing the fully closed Michaelis complex where the NTS caps the
active site to shield the reactive cationic intermediates from
water.
δ-Cadinene synthase (DCS) from Gossypium arboreum is a
typical α,β plant sesquiterpene synthase that catalyses the
metal-dependent conversion of FDP (1) to δ-cadinene (3), the
biological precursor of several cotton phytoalexins such as gos-
sypol.12 The NTS of DCS consists of the first 30 amino acids of
the β-domain and reaches across the α-domain to partially
cover the active site of the apo-enzyme (Fig. 1).13 Solution of
the co-crystal structure of DCS with three putative Mg2+-ions
and the substrate analogue 2-fluorofarnesyl diphosphate
(2F-FDP) (Fig. 1) yielded a structure that closely matched that
of unliganded DCS (Fig. 1) with an open and flexible active site
conformation; interestingly whereas three stabilizing inter-
domain interactions (P267/A25, P267/D26 and A269/Q28) appear
to fix the NTS to the ordered active site A–C loop segment in
the native structure of DCS, only one (A269/Q28) stabilizes the
fully liganded DCS complex (Fig. 1).13 Although weakly stabil-
ised closed conformations with more plastic active sites might
be expected for class I terpene enzymes that bind to struc-
turally different isoprenyl diphosphates,11 the unproductive
conformation adopted by 2F-FDP in DCS suggests that this
complexed structure is not a good representation of the
Michaelis complex. Furthermore the apparent lack of electro-
static interactions between the diphosphate group of 2F-FDP
and the putative trinuclear metal cluster is in stark contrast to
the closed structures observed for the Mg2+-bound 2F-FDP and
Sesquiterpene synthases catalyze the conversion of (E,E)-farne-
syl diphosphate (FDP, 1) to more than 300 distinct C15-iso-
prenoid hydrocarbons, which serve as precursors to the >7000
oxygenated sesquiterpenoids found in plants, fungi and bac-
teria.1 Catalysis by class I sesquiterpene synthases is depen-
dent on a shared α-helical fold (α domain),2 in which two
conserved aspartate-rich motifs (DDXXD and NSE/DTE3) bind
and activate the diphosphate group of the linear substrate 1
through coordination to a tri-nuclear Mg2+ cluster.4 After clea-
vage of the C1–O bond in FDP, the resulting farnesyl cation
undergoes a series of electrophilic cyclisations and rearrange-
ments to generate a final carbocation. Neutralization of the
positive charge either through loss of a proton or by nucleo-
philic water attack gives a unique sesquiterpene hydrocarbon
or alcohol.5,6
Inspection of the available X-ray crystal structures of single
domain microbial sesquiterpene synthases complexed with
unreactive analogues of FDP (1) and carbocationic intermedi-
ates5,7 as well as recent computational work8 suggest that com-
plexation of the essential metal ions4 induces structural
rearrangements of flexible loops leading to the formation of
the closed enzyme conformation that binds diphosphate 1 in
the reactive conformation. Plant mono- and sesquiterpene
synthases contain an additional N-terminal α-helical fold
known as the β-domain that was first observed in the X-ray
crystal structures of 5-epi-aristolochene synthase from Nicoti-
ana tabacum (TEAS)9 and subsequently in the structures of
bornyl diphosphate synthase from Salvia officinalis10 (BPPS)
and limonene synthase from Mentha spicata (LS).11 The cata-
School of Chemistry, Cardiff University, Main Building Park Place, Cardiff CF10 3AT,
UK. E-mail: allemannrk@cardiff.ac.uk
†Electronic supplementary information (ESI) available. See DOI:
10.1039/c6ob01398h
This journal is © The Royal Society of Chemistry 2016
Org. Biomol. Chem.

View Article Online
Communication
Organic & Biomolecular Chemistry
likely lead to poor desolvation, reduced catalytic activity15 and
alteration of the product profile. To explore the effects of
partial or complete removal of the NTS, and those of single
amino acid substitutions in the NTS on the formation of the
Michaelis complex of DCS, a series of truncated (M8, M20 and
M30), hybrid (CH-DCS) and single substitution CH-DCS and
DCS variants (CH-DCS-S24W and DCS-S30W) were produced
and characterised (Table 1 and ESI†).
GC-MS analysis of the pentane extractable products gener-
ated from incubations of FDP (1) with the N-terminally trun-
cated proteins M8, M20 and M30, in which the first 8, 20 and
30 amino acid residues of DCS had been removed respectively,
revealed that while the catalytic efficiencies (kcat/KM) of M8 and
M20 were similar to that of the wild type enzyme, deletion of
30 residues (M30) severely impaired the catalytic function of
DCS (Table 1). Interestingly, the progressive deletion of NTS
segments of approximately 10-residues in length led to the pro-
duction of increasing amounts of a sesquiterpene alcohol,
reaching 57% in M20 and 94% in M30 (Table 1). Comparison
of the MS fragmentation patterns and the gas chromatograms
of this alcohol and the product (5) generated by the microbial
(−)-germacradiene-4-ol synthase16 established the chemical
identity of the M8–M30-produced sesquiterpenol as (−)-germa-
cradiene-4-ol (5) (ESI†). The production of 5 by M8, M20 and
M30 is most likely the result of premature quenching of
germacradiene-6-yl cation (B) by a water molecule (Scheme 1) due
to less efficient active site closure and desolvation in the trun-
cated enzymes.7,17 The absence of any detectable hedycaryol
(4)18 from incubations with M8, M20 and M30 demonstrates
that the formation of carbocation B occurs more rapidly than
the neutralisation of its precursor carbocation A by a water
molecule (Scheme 1). In principle, the competing hydrolytic
and alkylation pathways that convert B to 5 or 3 (Scheme 1)
could involve a bridged carbocation corresponding to the
common branching intermediate D, from which germacra-
diene-4-ol and δ-cadinene can be formed (Scheme 1). It has
been suggested previously that the reaction pathways to germa-
crene and humulenes might be connected through a similar
bridged carbocation intermediate that resembles bicyclo-
germacrene.19 Mechanistic studies with DCS20 have shown that
Fig. 1 Panel A. Cartoon representation of the X-ray crystal structure of
DCS complexed with 2F-FPP and three putative Mg2+-ions (3g4f.pdb)
illustrating the capping of the active site in the α-domain (blue) by the
amino-terminal segment (NTS) of the β-domain (green).13 The first
amino acid of the β-domain is Ala-25; segments M1 to K24 and K42 to
I44 (dotted line) were disordered. Panel B. Cartoon representation of the
NTS in apo-DCS (3g4d.pdb, sand) and DCS complexed with 2-fluoro-
FDP (3g4f.pdb, light blue). Interactions P267–A25, P267–D26 and
A269–Q28 fixing the NTS to the A–C loop of the α-domain are high-
lighted in orange (apo DCS) and blue (liganded DCS).
Table 1 Steady-state kinetic parameters and product profile for all
proteinsa
kcat/KM
Enzyme
3b
5b
kcat (s−1
)
KM (μM)
(μM−1 s−1
)
2-fluorolinalyl diphosphate (2F-LDP) complexes of TEAS and
LS.11,14 These observations may argue against an involvement
of the NTS of DCS in the open to closed conformational tran-
sition commonly assumed to be required for the stabilization
of the Michaelis complex.
The NTS in class I plant terpene synthases is on average
more than 11 Å away from their respective active sites. Never-
theless, due to the proposed capping function of the NTS,
inefficient anchoring of the NTS to the α-domain will most
DCS
M8
M20
M30
CH-DCS
CH-DCS-S24W
DCS-S30W
98
85
43
6
2
15
57
94
75
0.010
0.008
0.003
3.2
2.9
3.1
2.8
1.2
2.5
n.m.c
4.7
25
0.006
1.3
Inactive
Inactive
a Kinetic data are from 3 independent measurements with an error
smaller than 10%. b Percentage (%). c Not measurable in the kinetic
assay used here.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2016

View Article Online
Organic & Biomolecular Chemistry
Communication
Fig. 2 Amino-terminal sequence alignment of 6 plant sesquiterpene
synthases, including the hybrid CH-DCS, and showing the truncation
points to generate M8, M20 and M30. (GenBank accession numbers are
X96429 for DCS, AJ786642 for EBFS, AJ304452 for GAS, JQ319661 for
ADS, and JQ812050 for TEAS).
extending the catalytically impaired M30 at its N-terminus
with the first 24 amino acids (M1–S24) of (E)-β-farnesene
synthase (EBFS) from Mentha × piperita (Fig. 2).22
The catalytic efficiency (kcat/KM) of CH-DCS was similar to
that of M20 (Table 1) and the product profile resembled that of
M30. The increased production of germacradiene-4-ol (5) by
CH-DCS when compared to DCS and M20 further underlines
the catalytic relevance of the NTS motif and the existence of
catalytic and product specificity determinants within the con-
sensus sequence XRPXXXFXS. In agreement, CH-DCS-S24W,
which was produced as a cloning artifact, was inactive, thus
supporting the essential role of this serine residue in stabilis-
ing the closed conformation of CH-DCS. This rationale was
further corroborated by the observation that no activity could
be detected for DCS-S30W. Remarkably, this simple swapping
experiment illustrates how both the length and the compo-
sition of the NTS modulate the desolvation process that is
required for effective catalysis by DCS. Specific changes to the
NTS can also alter the product specificity and lead to novel
enzymes (CH-DCS, Table 1) that produce distinct product(s)
Scheme 1 Panel A. Conversion of FDP (1) to δ-cadinene (3) by DCS via
NDP (2) and carbocations A, B and C. Framed are the expected hydro-
lysis products arising from these cations. For more mechanistic details
see ref. 19. Compound 5 was assigned as (−)-germacradien-4-ol based
on chiral GC measurements and GC/MS comparisons with an authentic
sample (see ESI†). Panel B. Postulated enzymatic formation of 3 and 5
from a bridged carbocation D.
FDP (1) is first converted to nerolidyl diphosphate (NDP, 2) via with high efficiency.
an enzyme bound farnesyl cation/−OPP ion pair that most
In conclusion, despite minimal sequence identity and
likely prevents the intermolecular nucleophilic attack of a varying lengths, the N-terminal polypeptide segment of plant
water molecule, thus explaining the absence of farnesol or ner- sesquiterpene synthases is essential for their high catalytic
olidol when M8, M20 and M30 are incubated with 1.
activity and the product specificity. Indeed, NTSs contain
The low catalytic efficiency of M30 together with the near specific residues such as Ser30 and/or segments like
wild type efficiency of M8 and M20 suggest that the sequence M21RPKADF27QPS30 in DCS implicated in active site closure and
M21RPKADF27QPS30 in DCS, which is conserved as XRPXXXFXPS desolvation events upon formation of the Michaelis complex.
in plant sesquiterpene cyclases, is vital for efficient catalysis Despite its distance from the active site, the NTS seems to help
and product fidelity. For DCS, this deca-peptide is the minimal shape the active site contour that dictates substrate folding
amino-terminal segment required to stabilise the Michaelis and ultimately product specificity of plant sesquiterpene
complex. Despite the unusual, open X-ray crystal structure of synthases. As a consequence of its late involvement in cataly-
DCS in its ligand bound form,13 the production of increasing sis, changes to the NTS by single mutation, truncation and
amounts of germacradien-4-ol (5) with increasing extent of domain swap only compromise the turnover number kcat
N-terminal truncation clearly implicates the NTS in the shield- while the Michaelis constants KM remain largely unchanged
ing of the active site from bulk solvent. (Table 1). Furthermore, our work reveals the indirect contri-
,
Sequence alignments of DCS with other class I plant sesqui- bution of NTS in defining the function of dual sesquiterpene/-
terpene synthases (Fig. 2 and ESI†) indicate that there is essen- sesquiterpenol synthases. Comparisons between CH-DCS, M20
tially no N-terminal sequence similarity outside the consensus and M30 reveal that the sesquiterpene/sesquiterpenol ratio
motif XRPXXXFXPS, i.e. M21RPKADF27QPS30 in DCS. To assess is modulated by precisely balancing the kcat values with
the function of the XRPXXXFXPS sequence as a determinant of the degree of solvation/desolvation. Key modulators are the
product specificity, a domain-swapping strategy21 was adopted conserved Ser30 in DCS at the C-terminal end of the consensus
for DCS, in which the hybrid CH-DCS was constructed by motif XRPXXXFXS and the length of the NTS. Although pro-
This journal is © The Royal Society of Chemistry 2016
Org. Biomol. Chem.

View Article Online
Communication
Organic & Biomolecular Chemistry
gressive shortening of the NTS leads to enzymes with a lower
hydrocarbon/alcohol ratio (M8, M20 and M30), the increase in
the alcohol fraction is paralleled by a severe reduction of the
catalytic activity. NTSs can be moved between different plant
sesquiterpene synthases as shown here by importing the NTS
from EBFS into DCS in CH-DCS. The increasing amount of
(−)-germacradien-4-ol (5) produced by progressive N-terminal
truncation of DCS strongly suggests that the NTS contributes
to the shielding of the active site from water. The absence of
hedycaryol (4) and cadinol (6) (Scheme 1) shows that carbo-
Chem., Int. Ed., 2014, 53, 7652; (b) M. Chen, W. K. W. Chou,
N. Al-Lami, J. A. Faraldos, R. K. Allemann, D. E. Cane and
D. W. Christianson, Biochemistry, 2016, 55, 2864;
(c) M. Chen, N. Al-Lami, M. Janvier, E. L. D’Antonio,
J. A. Faraldos, D. E. Cane, R. K. Allemann and
D. W. Christianson, Biochemistry, 2013, 52, 5441.
8 M. W. van der Kamp, J. Sirirak, J. Zurek, R. K. Allemann
and A. J. Mulholland, Biochemistry, 2013, 52, 8094.
9 C. M. Starks, K. Back, J. Chappell and J. P. Noel, Science,
1997, 277, 1815.
cations A and C are not quenched by water, either due to the 10 D. A. Whittington, M. L. Wise, M. Urbansky, R. M. Coates,
formation of bridged carbocation (D) or because they are pro-
tected as diphosphate ion pairs. The sustainable production of
R. B. Croteau and D. W. Christianson, Proc. Natl. Acad.
Sci. U. S. A., 2002, 99, 15375.
high value terpenes and terpene alcohols using metabolically 11 D. C. Hyatt, B. Youn, Y. Zhao, B. Santhamma, R. M. Coates,
engineered microbial hosts23 with catalytically optimized and
altered versions of terpene synthases24 profits from a more
R. B. Croteau and C. Kan, Proc. Natl. Acad. Sci. U. S. A.,
2007, 104, 5360.
detailed understanding of the molecular mechanisms that 12 K. J. Abraham, M. L. Pierce and M. Essenberg, Phytochemis-
control product specificities and reaction rates in terpene
synthase chemistry.
We thank Dr Xiao-Ya Chen (Institute of Plant Physiology
and Ecology, and Institute for Biological Sciences, Shanghai,
try, 1999, 52, 829.
13 H. A. Gennadios, V. Gonzalez, L. Di Costanzo, A. Li, F. Yu,
D. J. Miller, R. K. Allemann and D. W. Christianson, Bio-
chemistry, 2009, 48, 6175.
China) for plasmid pXC1 harboring the cDNA of (+)-δ-cadi- 14 J. P. Noel, N. Dellas, J. A. Faraldos, M. Zhao, B. A. Hess,
nene, Dr Linda Field (Rothamsted Research, Harpenden, UK)
L. Smentek, R. M. Coates and P. E. O’Maille, ACS Chem.
for a cDNA coding for (E)-β-farnesene synthase, Dr Y. Ohnishi
Biol., 2010, 5, 377.
(University of Tokyo) for the germacren-4-ol synthase 15 M. Y. Kim, Y. J. Chang, M. H. Bang, N. I. Baek, J. Jin,
expression vector (pET16b-SC1) and Dr David Miller (Cardiff C. H. Lee and S. U. Kim, J. Plant Biol., 2005, 48, 178.
University) for helpful comments on the manuscript. This 16 C. Nakano, F. Kudo, T. Eguchi and Y. Ohnishi, ChemBio-
work was supported by the UK’s Biotechnology and Biological Chem, 2011, 12, 2271.
Sciences Research Council (BBSRC) through grant BB/H01683X/1, 17 D. J. Grundy, M. Chen, V. Gonzalez, S. Leoni, D. J. Miller,
the Engineering and Physical Sciences Research Council
(EPSRC) through grants EP/D069580/1 and EP/K301635/1 and
Cardiff University.
D. W. Christianson and R. K. Allemann, Biochemistry, 2016,
55, 2112.
18 P. Baer, P. Rabe, C. A. Citron, C. C. O. Mann, N. Kaufmann,
M. Groll and J. S. Dickschat, ChemBioChem, 2014, 15,
213.
19 V. Gonzalez, S. Touchet, D. J. Grundy, J. A. Faraldos and
R. K. Allemann, J. Am. Chem. Soc., 2014, 136, 14505.
20 J. A. Faraldos, D. J. Miller, V. González, Z. Yoosuf-Aly,
O. Cascón, A. Li and R. K. Allemann, J. Am. Chem. Soc.,
2012, 134, 5900.
Notes and references
1 D. E. Cane, Acc. Chem. Res., 1985, 18, 220.
2 L. C. Tarshis, M. Yan, C. D. Poulter and J. C. Sacchettini,
Biochemistry, 1994, 33, 10871.
3 D. W. Christianson, Chem. Rev., 2006, 106, 3412.
4 J. A. Aaron and D. W. Christianson, Pure Appl. Chem., 2011,
82, 1585.
5 E. Y. Shishova, F. Yu, D. J. Miller, J. A. Faraldos, Y. Zhao,
R. M. Coates, R. K. Allemann, D. E. Cane and
D. W. Christianson, J. Biol. Chem., 2008, 283, 15431.
6 J. A. Faraldos, S. Wu, J. Chappell and R. M. Coates, J. Am.
Chem. Soc., 2010, 132, 2998.
21 Y. Back and J. Chappell, Proc. Natl. Acad. Sci. U. S. A., 1996,
93, 6841.
22 J. A. Faraldos, V. Gonzalez, A. Li, F. Yu, M. Koksal,
D. W. Christianson and R. K. Allemann, J. Am. Chem. Soc.,
2012, 134, 20844.
23 J. C. Paddon and J. D. Keasling, Nat. Rev. Microbiol., 2014,
12, 355.
24 J. X. Li, X. Fang, Q. Zhao, J. X. Ruan, C. Q. Yang, L. J. Wang,
D. J. Miller, J. A. Faraldos, R. K. Allemann, X. Y. Chen and
P. Zhang, Biochem. J., 2013, 451, 417.
7 (a) P. Baer, P. Rabe, K. Fischer, C. A. Citron,
T. A. Klapschinski, M. Groll and J. S. Dickschat, Angew.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2016