(+)-(10R)-Germacrene A synthase from goldenrod, Solidago canadensis; cDNA isolation, bacterial expression and functional analysis.

Article abstract of DOI:10.1016/S0031-9422(02)00165-6

Profiling of sesquiterpene hydrocarbons in extracts of goldenrod, Solidago canadensis, by GC-MS revealed the presence of both enantiomers of germacrene D and lesser amounts of germacrene A, alpha-humulene, and beta-caryophyllene. A similarity-based cloning strategy using degenerate oligonucleotide primers, based on conserved amino acid sequences in known plant sesquiterpene synthases and RT-PCR, resulted in the isolation of a full length sesquiterpene synthase cDNA. Functional expression of the cDNA in E. coli, as an N-terminal thioredoxin fusion protein using the pET32b vector yielded an enzyme that was readily purified by nickel-chelate affinity chromatography. Chiral GC-MS analysis of products from of (3)H- and (2)H-labelled farnesyl diphosphate identified the enzyme as (+)-(10R)-germacrene A synthase. Sequence analysis and molecular modelling was used to compare this enzyme with the mechanistically related epi-aristolochene synthase from tobacco.

Full text of DOI:10.1016/S0031-9422(02)00165-6

Phytochemistry 60 (2002) 691–702
www.elsevier.com/locate/phytochem
(+)-(10R)-Germacrene A synthase from goldenrod,
Solidago canadensis; cDNA isolation, bacterial
expression and functional analysis§
Ian Prossera, Andy L. Phillipsa, Simon Gittingsa, Mervyn J. Lewisa,
Antony M. Hooperb, John A. Pickettb, Michael H. Bealea,*
aLong Ashton Research Station, Department of Agricultural Sciences, University of Bristol, Long Ashton, Bristol, BS41 9AF, UK
bRothamsted Experimental Station, Harpenden, Hertfordshire AL5 2JQ, UK
Received 29 November 2001; received in revised form 24 April 2002
Abstract
Profiling of sesquiterpene hydrocarbons in extracts of goldenrod, Solidago canadensis, by GC–MS revealed the presence of both
enantiomers of germacrene D and lesser amounts of germacrene A, a-humulene, and b-caryophyllene. A similarity-based cloning
strategy using degenerate oligonucleotide primers, based on conserved amino acid sequences in known plant sesquiterpene syn-
thases and RT-PCR, resulted in the isolation of a full length sesquiterpene synthase cDNA. Functional expression of the cDNA in
E. coli, as an N-terminal thioredoxin fusion protein using the pET32b vector yielded an enzyme that was readily purified by nickel-
3
2
chelate affinity chromatography. Chiral GC–MS analysis of products from of H- and H-labelled farnesyl diphosphate identified
the enzyme as (+)-(10R)-germacrene A synthase. Sequence analysis and molecular modelling was used to compare this enzyme
with the mechanistically related epi-aristolochene synthase from tobacco. # 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Farnesyl diphosphate; Germacrene A; Goldenrod; Solidago canadensis; Sesquiterpene synthase
1. Introduction
known (Glasby, 1982) are derived metabolically from
some 300 distinct C15-hydrocarbon skeleta, which in
turn are produced from the single substrate farnesyl di-
phosphate (FPP) (see Fig. 1 for structures) by the action
of sesquiterpene synthases (cyclases). It is known that
these enzymic cyclisations proceed by ionisation of the
diphosphate group of FPP to form an enzyme-bound
allylic cation, which then undergoes isomerisation,
before cyclisation to yield new cations. Subsequent
rearrangements and final quenching of the carbocations
by elimination or by reaction with water yields the var-
ious C15 hydrocarbon and alcohol products that are
typical of a sesquiterpene synthase (Cane, 1985, 1990,
1999). Recently, our ability to study the enzymology of
these reactions has increased dramatically as a result of
the cloning and functional expression in E. coli of a
number of cDNAs encoding sesquiterpene synthases of
microbial (Hohn and Plattner, 1989; Proctor and Hohn,
1993; Cane et al., 1994) and plant origin (Facchini and
Chappell, 1992; Back and Chappell, 1995; Chen et al.,
1995, 1996; Crock et al., 1997; Bohlmann et al., 1998;
Colby et al., 1998; Steele et al., 1998; Back et al., 1998;
The sesquiterpenes are a large family of C15-iso-
prenoid natural products found in higher and lower
plants, microbes and some marine organisms. Many
have biological activity, including antimicrobial, anti-
tumour, and cytotoxic properties. In plants, they play
important ecological roles in interactions with insects
and microbes and act as attractants, deterrents, anti-
feedants and phytoalexins. Furthermore, sesquiterpenes
are key components of many essential oils, which are
important commercially for the flavour and fragrance
industries. The tens of thousands of sesquiterpenes
Abbreviations: FPP, farnesyl diphosphate; TEAS, tobacco epi-
aristolochene synthase; GC–MS, gas chromatography–mass spectro-
metry; PCR, polymerase chain reaction; SDS–PAGE, sodium dodecyl
sulfate–polyacrylamide gel electrophoresis.
§
The nucleotide sequence reported in this paper has been sub-
mitted to the EMBL database with accession number AJ304452.
* Corresponding author. Tel.: +44-1275-549289; fax: +44-1275-
394281.
E-mail address: mike.beale@bbsrc.ac.uk (M.H. Beale).
0031-9422/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0031-9422(02)00165-6

692
I. Prosser et al. / Phytochemistry 60 (2002) 691–702
Fig. 1. Structures of sesquiterpene hydrocarbons.
Hua and Matsuda, 1999; Mercke et al., 1999). Further-
more, crystal structures of three recombinant sesqui-
terpene synthases have given insight into the active site
components involved in the binding of FPP and cata-
lysis of its cyclisation (Lesburg et al., 1997; Starks et al.,
1997; Caruthers, 2000).
For those sesquiterpene synthases that convert FPP to
compounds of the eremophilane and eudesmane classes,
including epi-aristolochene (Facchini and Chappell,
1992), aristolochene (Proctor and Hohn, 1993), vetis-
piradiene (Back and Chappell, 1995) and dꢀselinene
(Steele et al., 1998), it is postulated that the 10-mem-
bered-ring germacrenes (isomers A, B, C or D) (Fig. 1)
are the initial products of the cyclisation and subsequent
elimination and rearrangement reactions. These neutral
compounds are apparently not released from the active
site of the sesquiterpene synthase but are re-protonated
by the enzyme to form new cations, which undergo fur-
ther cyclisations to yield the final products. Evidence for
cryptic germacrene A synthase activity within aris-
tolochene synthase has been provided by Cane and
Tsantrizos (1996) who observed the formation of dihy-
drogermacrene A from the substrate analogue dihydro-
FPP. More recently, site-directed mutagenesis of
tobacco epi-aristolochene synthase (TEAS) has given
rise to a mutant protein that produced germacrene A,
albeit at lower efficiency compared to the wild-type
production of epi-aristolochene (Rising et al., 2000).

I. Prosser et al. / Phytochemistry 60 (2002) 691–702
693
There are many plant natural products that are derived
from oxidative metabolism of germacrene A, indicating
that sesquiterpene synthases that produce and release
this thermally unstable 10-membered ring hydrocarbon
are widespread. Evidence for germacrene A synthase
activity in plants has also been presented (de Kraker et
al., 1998). Germacrene A synthase is an important bio-
technological target involved in the biosynthesis of par-
thenolide, the active ingredient of feverfew, as well as in
the formation of the bitter sesquiterpenoids of lettuce
and chicory (Sessa et al., 2000). In this paper we
describe the cDNA cloning, functional expression in E.
coli, and characterisation of this gene — (+)-(10R)-
germacrene A synthase — from the goldenrod, Solidago
canadensis.
The compounds were identified by comparison of spec-
tra and Kovat’s retention indices with library and pub-
lished data (Adams, 1995). As expected from previous
literature reports (Niwa et al., 1980; Schmidt et al.,
1998; Bulow and Konig, 1997), germacrene D was the
major component (71% of total sesquiterpenes). Fur-
ther analysis by chiral GC-MS using columns and con-
ditions described by Schmidt et al. (1998) showed that
both enantiomers were present. The ratio of (+)-ger-
macrene D to (ꢀ)-germacrene D was 4:3, based on total
ion current. Other sesquiterpenes present in significant
amounts were bꢀcaryophyllene (3%), a-humulene
(1%), d-cadinene (7%) and germacrene A (7%). Minor
amounts of other compounds such as b-bourbonene, a-
ylangene, and b gurjurnene were also detected. There
was no significant difference between the profile of
previously wounded and that of unwounded tissue.
2. Results
2.2. Sesquiterpene synthase cDNA isolation and
expression
2.1. Solidago sesquiterpene hydrocarbon profile
The presence of both enantiomers of germacrene D
has been reported in extracts or hydrodistillates of var-
ious species of Solidago (S. altissima, S. canadensis, S.
gigantica and S. vigaurea) (Niwa et al., 1980; Schmidt et
al., 1998; Bulow and Konig, 1997; Kalemba, 1998).
Leaves of greenhouse-grown S. canadensis var. Gold
Baby were extracted with pentane and the sesquiterpene
hydrocarbon profile examined by GC–MS. The profile
is shown in Fig. 2 and structures are shown in Fig. 1.
A homology-based PCR cloning strategy was used.
Sequence alignments of published sesquiterpene synthase
genes were used to design degenerate oligonucleotide pri-
mers. PCR using the primer pair corresponding to the
amino acid sequences EDIL(D/E)(E/D)A and
WWKDLD indicated in Fig. 3 gave a sesquiterpene
synthase fragment, designated si6, as determined by
sequencing and analysis of homology with known ter-
pene synthases. cDNA fragment si6 was radio-labelled
Fig. 2. Sesquiterpene hydrocarbon profile of Solidago canadensis. 2A GC–MS (cold injection on BPX5) of pentane extract. Peaks—1 C12 hydro-
carbon standard; 2 aꢀylangene; 3 bꢀbourbonene; 4 C14 hydrocarbon standard; 5 bꢀcaryophyllene; 6 bꢀgurjunene; 7 aꢀhumulene; 8 germacrene
D; 9 germacrene A; 10 dꢀcadinene; 11 unknown sesquiterpene; 12–15 C16-C19 hydrocarbon standards. 2B Separation of germacrene D enantiomers
on Cyclosil-B chiral phase.

694
I. Prosser et al. / Phytochemistry 60 (2002) 691–702
Fig. 3. Amino acid sequence alignment of (+)-germacrene A synthase (Sc1) with related plant sesquiterpene synthases; germc=germacrene C
synthase (EMBL Accession AF035630) (Colby et al., 1998), cad=cadinene synthase (U27535) (Chen et al., 1996), eas=epi-aristolochene synthase
(L04680) (Facchini and Chappell, 1992) and vetis=vetispiradiene synthase (U20188) (Back and Chappell, 1995). Sequences corresponding to RT-
PCR primers TC12 and TC4 are shown. The characteristic DDXXD motif is underlined. Amino acid residues referred to in discussion of the enzyme
mechanism are highlighted with asterisks.

I. Prosser et al. / Phytochemistry 60 (2002) 691–702
695
and used to probe a S. canadensis cDNA library, giving
rise to the full-length clone Sc1. The deduced amino
acid sequence of Sc1 is shown in Fig. 3, aligned with
selected published sesquiterpene synthase sequences.
The homology, length of the sequence and the lack of
obvious plastid targeting transit peptide indicated that
the cDNA encoded a sesquiterpene synthase rather than
a mono- or diterpene synthase.
by radio-counting and GC–MS. To confirm these
results with an isotope traceable by mass spectrometry,
the enzyme preparation was incubated with [1-2H]-FPP,
containing a small amount of [1-3H]-FPP to enable addi-
tional tracking by radio-counting. [2H1]-germacrene A
(98%) and [2H1]-a-humulene (2%) were the only products
detected in the hydrocarbon fraction (Fig. 5). Analysis of
the mass spectra indicated that there was total retention of
deuterium in both products, thus establishing the cloned
sesquiterpene synthase as germacrene A synthase.
To determine its function, Sc1 was expressed in E.
coli. After some experimentation, the pET32b vector
was selected. This vector, which gives rise to proteins
containing N-terminal thioredoxin fusions, gave a high
level of expression as determined by analysis by SDS–
PAGE. As shown in Fig. 4 a soluble protein was
induced at a molecular weight of circa 80 KDa. This is
consistent with the predicted molecular weight of the
sesquiterpene cyclase (ꢁ65 kDa) plus thioredoxin and
linkers. Incubation of soluble protein from E. coli cells
expressing Sc1 with [1-3H]-FPP produced pentane solu-
ble products, as determined by radiocounting. These
were identified by GC–MS as germacrene A (98%) and
a-humulene (2%) (Fig. 5). Control incubations of
[1-3H]-FPP with soluble proteins from cells transformed
with vector only did not produce pentane soluble meta-
bolites, as determined by radio-counting. Additionally,
no sesquiterpene hydrocarbons were detected by GC–
MS in these control experiments. Diethyl ether extracts
of the incubations, made subsequently to the pentane
extractions, of both control and expressed enzyme-con-
taining solutions contained small amounts of [1-3H]-far-
nesol and no other sesquiterpene alcohols, as determined
2.3. Chirality of enzyme-produced germacrene A
The chirality of the germacrene A produced by the
recombinant enzyme was determined by the chiral GC–
MS method of de Kraker et al. (1998). This method
relies on the thermal Cope rearrangement of germacrene
A to b-elemene in hot GC injectors (structures Fig. 1).
The two enantiomers of b-elemene are separable on
chiral GC phases based on heptakis (2,3-di-O-methyl-6-
O-t-butyl dimethylsilyl)-b-cyclodextrin (Konig et al.,
1994). Standards of (ꢀ)-b-elemene in Piper nigrum
(black pepper) essential oil and (+)-b-elemene from hot
injection of a pentane extraction of the liverwort Sca-
pania undulata were used to authenticate the separation
(Konig et al., 1994). As shown in Fig. 6, hot injection of
the germacrene A produced by the recombinant sesqui-
terpene synthase, Sc1, gave a single peak, coincident
with (ꢀ)-b-elemene standard. Cope rearrangement of
germacrene A occurs with retention of configuration
and (+)-germacrene A gives rise to (ꢀ)-b-elemene (de
Kraker et al., 1998). Thus, the cloned Solidago sesqui-
terpene synthase, Sc1, was identified as a (+)-germacrene
A synthase.
2.4. Enzyme purification and characterisation
As well as the N-terminal thioredoxin fusion, the
recombinant (+)-germacrene A synthase also contains
a polyhistidine tag. Single step purification of the
enzyme was achieved by affinity purification on a Ni-che-
late chromatography column, as shown by SDS–PAGE
analysis (Fig. 4). Enzyme purified in this way retained
activity after chromatography and after storage frozen at
ꢀ20 ꢂC in 50% glycerol. Proteolytic cleavage of the fusion
protein with thrombin yielded native enzyme that retained
germacrene A synthase activity. For convenience, most
characterisation was carried out on the affinity purified
fusion protein. The enzyme had a Km for FPP of 2.5 mM
at pH 7.0, as determined from a Hanes plot of [FPP]/
velocity versus [FPP] (data not shown). This value is
typical for higher plant sesquiterpene synthases (Steele
et al., 1998). The enzyme consistently produced (+)-
germacrene A and a-humulene in the ratio 98:2 from
feeds of unlabelled or [1-2H]-FPP. However, when
[12,13-2H6]-FPP was used as a substrate, the proportion
Fig. 4. SDS–PAGE analysis of proteins from E. coli expression of
(+)-germacrene A synthase. Cultures of E. coli strain AD494(DE3)-
pLysS containing the pET32b vector with or without the Sc1 insert
were induced with IPTG and protein extracts prepared and purified by
affinity chromatography as described in the experimental procedures.
Lane 1 Soluble protein from vector only control induction. Lane 2 solu-
ble protein from pET:Sc1 induction Lane 3 HisBind affinity purification-
column wash through Lane 4 purified (+)-germacrene A synthase eluted
from HisBind column. Lane M molecular weight markers.

696
I. Prosser et al. / Phytochemistry 60 (2002) 691–702
Fig. 5. GC–MS analysis (cold injection) of the products produced by Sc1-thioredoxin fusion from FPP and [2H1]-FPP substrates. A—total ion
current trace from FPP feed. B—mass spectra of peak 1 (aꢀhumulene) and peak 2 (germacrene A) from FPP incubation. C—mass spectra of peaks
1 and 2 from [2H1]-FPP incubation.
of [2H6]-a-humulene formed increased to 11%. Addi-
tionally [2H6]-germacrene D (9% of total products)
could be detected along with the expected [2H5]-germa-
crene A (Fig. 7). Analysis by chiral GC-MS revealed the
labelled [2H6]-germacrene D produced was exclusively
the (ꢀ) enantiomer.
2.5. Sequence analysis and structural model
The cloned cDNA encoded an open reading frame
giving rise to a protein of 549 amino acids, with a pre-
dicted molecular weight of 64.7 kDa. The deduced pro-
tein sequence is shown in Fig. 3, aligned with other

I. Prosser et al. / Phytochemistry 60 (2002) 691–702
697
representative angiosperm sesquiterpene synthases. Two
of these, epi-aristolochene (TEAS) and vetispiradiene
synthases, are postulated to convert FPP to bicyclic
hydrocarbon products via germacrene A. Thus, it was
useful to compare these sequences with that of the Soli-
dago (+)-germacrene A synthase, particularly as the
crystal structure of TEAS has been determined. How-
ever, (+)-germacrene A synthase has slightly higher
similarity and identity with cadinene synthase (58%
similarity, 46% identity) and germacrene C synthase
(57% similarity, 44% identity) than it has with TEAS
(52% similarity, 40.5% identity) and vetispiradiene
synthase (54% similarity, 41.5% identity).
The amino-acid sequence of (+)-germacrene A syn-
thase contains the characteristic, DDXXD/E motif, that
is highly conserved amongst sesquiterpene synthases
(Fig. 3). Structural studies of TEAS (Starks et al., 1997)
and aristolochene synthase (Caruthers et al., 2000) have
indicated that elements of this motif co-ordinate to a
magnesium ion involved in the ionisation of the di-
phosphate group. Other residues involved in binding
and catalysis have been identified by Starks et al. (1997).
Some of the amino acid side-chains suggested to be
involved in the formation of intermediate germacrene A
by TEAS, are present in (+)-germacrene A synthase.
For example, Arg264 and Arg441 which take part in
magnesium binding and diphosphate ionisation in
TEAS are present in (+)-germacrene A synthase (resi-
dues 264 and 443), together with Asp525 (position 526 in
(+)-germacrene A synthase), a residue that is impli-
cated in proton elimination at C-13 of the germacrenyl
cation. However, Tyr527, a residue that was suggested to
play a key role in the stabilisation of the germacrenyl
cation in TEAS, is a leucine in (+)-germacrene A syn-
thase (position 528). This change may have some bearing
on the reaction mechanism. Further indicators were
sought by mapping of the (+)-germacrene A synthase
sequence onto the TEAS structure using the Swiss-Model
service (http://www.expasy.ch/swissmod/SWISS-MOD-
EL.html) to produce an energy minimised model.
Examination of overlays of the active site regions of the
two enzymes failed to reveal obvious differences in
architecture that may indicate why reaction in (+)-ger-
macrene A synthase terminates at germacrene A. From
the structure of TEAS, complexed with substrate ana-
logues, further cyclisation of germacrene A appears to
be initiated by protonation by the hydroxyl group of
Tyr520 which itself is polarised by hydrogen-bonding
from Asp444 and Asp525 (Starks et al., 1997). Indications
of the role of this tyrosine have also been obtained from
mutagenesis studies (Rising et al., 2000). In (+)-germa-
crene A synthase this tyrosine is conserved (residue 522)
along with the second aspartate (at position 526).
However the aspartate at position 444 in TEAS has
been changed to asparagine (position 446) in (+)-ger-
macrene A synthase. This change is also present in the
Fig. 6. Determination of the chirality of (+)-germacrene A produced
by Sc1-thioredoxin fusion protein by analysis of (ꢀ) -b-elemene
formed by hot injection GC–MS on Cyclosil-B chiral phase. Relevant
region of sesquiterpene profiles of A, black pepper oil; B Scapania
undulata extract; C germacrene A from FPP feed to Sc1 and D Co-
injection of germacrene A from Sc1 and Scapania undulata extract.

698
I. Prosser et al. / Phytochemistry 60 (2002) 691–702
Fig. 7. A GC–MS analysis of products from incubation of [12,13-2H6]-FPP with Sc1 protein. B Mechanism of formation of the observed products,
alternative pathways from the germacrenyl cation are indicated by arrows a and b.
equivalent position in the sequence of germacrene C
synthase, and thus could be correlated with the termi-
nation of the reaction at the neutral germacrene. How-
ever, aristolochene synthase, which also produces
germacrene A as an intermediate also contains this
aspartate to asparagine change.
been reported from this species (Schmidt et al., 1999a).
This may be due rapid metabolism of gurjunene to
cyclcolerenone or perhaps to varietal differences
between the cultivar used and wild populations used by
other workers. The second major sesquiterpene of the
profile was germacrene A, the product of the enzyme
encoded by the cloned Sc1 cDNA. It was perhaps sur-
prising that the homology cloning strategy yielded a
sesquiterpene synthase that produced predominately
germacrene A, given the large excess of germacrene D
over germacrene A in the plant extract. The profile of
sesquiterpene hydrocarbons in the plant may not, how-
ever, reflect the relative expression of sesquiterpene
synthase genes, as some products of synthase activity
could be rapidly metabolised further, while others
accumulate.
3. Discussion
The sesquiterpene profile observed for the cultivar of
Solidago canadensis agrees with previous reports on
extracts and hydrodistillates of wild collections in that
(+)/(ꢀ) germacrene D is by far the main constituent. It
is interesting that a- and g-gurjunenes were not present,
given that significant gurjunene synthase activity has

I. Prosser et al. / Phytochemistry 60 (2002) 691–702
699
The sesquiterpene synthase encoded by Sc1 was
expressed to a high level in E. coli cells using the pET32
vector. This yielded a soluble thioredoxin fusion protein
that was readily purified by affinity chromatography.
Although (+)-germacrene A synthase activity was
readily detectable in total soluble proteins from the
bacterial cultures, most of the characterisation was done
on affinity-purified fusion protein. This avoided back-
ground phosphatase activity in the crude cell lysates and
gave cleaner samples for GC–MS analysis. The recom-
binant enzyme (fusion protein and native protein gen-
erated by thrombin cleavage) converted FPP to almost
exclusively (+)-germacrene A. a-Humulene was a by-
product produced consistently at the level of 2%.
Examples of recombinant sesquiterpene synthases that
produce single products have been reported, but often
the products are mixtures that are sometimes complex
reflecting the many possible reaction paths involving
carbocationic rearrangements (Steele et al., 1998). It is
not clear at this time whether this enzyme activity is the
sole source of the a-humulene present in the profile.
Definitive confirmation of the function of Sc1 protein
was obtained using deuterium-labelled substrate and
analysis of the products by GC-MS. Deuterium labels
from C-1 of FPP were retained in both (+)-germacrene
A and aꢀhumulene. However, when [12,13-2H6]-FPP
was used as substrate for Sc1, (ꢀ)-[2H6]- germacrene D
(9%) was also observed as a product, along with an
increased level of [2H6]- a-humulene (11%), in addition
to the expected labelled (+)-[2H5]- germacrene A. The
increased proportion of aꢀhumulene in these experi-
ments, appears to be due to secondary isotope effects on
the relative stability of the germacrenyl and humulyl
cations (Fig. 7). Hyperconjugative stabilisation of
[12,13- 2 H6] germacrenyl ion is less than that in the fully
protio- species and this may favour increased formation
of the humulyl ion in the hexadeuterio analogue. The
appearance of (ꢀ)-[2H6]-germacrene D can be attributed
to the introduction of a primary isotope effect into the
reaction pathway leading to (+)-germacrene A. The
outcome is a slowing of the elimination step resulting in
the formation of the 11,12 double bond, thereby allow-
ing the alternative pathway leading to (ꢀ)-germacrene
D to compete more effectively. This type of isotopically
sensitive branching is not uncommon in terpene bio-
synthesis and has been observed before for mono-
terpene (Croteau et al., 1987; Wagschal et al., 1991,
1994) and diterpene synthases Williams et al. (1970). It
is known, from experiments with [10-2H]-FPP and par-
tially purified germacrene D synthases from Solidago,
that the formation of (ꢀ)-germacrene D from the inter-
mediate germacrenyl cation is accomplished via a 1,3-H
shift from C-1 to C-11 (Schmidt et al., 1999b). The
chirality of the products obtained from germacrene A
synthase is consistent with the intermediacy of the
(10R)-germacrenyl cation as shown in Fig. 7. This gives
rise to (+)-(10R)-germacrene A by loss of a C-12
hydrogen (deuterium), or to (ꢀ)-germacrene D by the
1,3-H shift and subsequent elimination. The absolute
stereochemistry of (+)-(10R)-germacrene A follows
from that of the (ꢀ)-(10S)-enantiomer determined by
Weinheimer et al. (1970).
Sequence analysis and molecular modelling of (+)-
germacrene A synthase was used to try to gain insight
into the reaction mechanism, with particular reference
to that of TEAS whose structure has been determined
(Starks et al., 1997). Comparison of total sequence
homologies between known sesquiterpene synthases,
particularly across species, is not that informative at
present. Homologies amongst plant examples are nor-
mally of the order of 50% regardless of similarities in
the reaction pathway. Nevertheless, there is conserva-
tion of certain key residues that have been implicated as
being involved in catalysis. Homologies between plant
and microbe sequences are even lower. For example,
epi-aristolochene and aristolochene synthase show only
16% identity despite much similarity between the reac-
tions catalysed (Caruthers et al., 2000). When consider-
ing reaction paths, more informed speculation of the
role of certain amino acid side-chains comes from
examination of structures predicted from modelling of
sequences onto known sesquiterpene synthase structures
and consideration of putative active site architecture.
The (+)-germacrene A synthase model generated was
examined with particular emphasis on identifying dif-
ferences with respect to the TEAS structure. The overall
shape of the active site in germacrene A synthase was
very similar to that of TEAS. The analysis failed to
reveal obvious differences that could be attributed to
termination of the reaction at (+)-germacrene A in the
new synthase. The presence of a leucine in place of tyr-
osine (position 527 in TEAS) could be significant in
removing stabilisation of the germacrenyl cation. This
may have the effect of accelerating the elimination
reaction to (+)-germacrene A, but cannot explain the
failure of this molecule to undergo further reaction. The
tyrosine at position 520 in TEAS has been implicated in
catalysing further reaction of germacrene A by proto-
nation. Mutagenesis of Tyr520 gives a mutant TEAS
that makes germacrene A (Rising et al., 2000). This,
however, does not confirm that wild-type TEAS pro-
duces germacrene A as an intermediate. Tyr520 is pre-
served in (+)-germacrene A synthase but only one of
two aspartate residues that assist in orientation and
polarisation of the hydroxyl group of this tyrosine is
present. It is possible that this change may form the
molecular basis for the differences in reaction pathway
between the two enzymes. However, the aspartate to
asparagine change in question is also present in aris-
tolochene synthase, an enzyme that produces (ꢀ)-ger-
macrene A as an intermediate. Thus the significance of
this change is unclear at present. More definitive data

700
I. Prosser et al. / Phytochemistry 60 (2002) 691–702
will come from a crystal structure of (+)-germacrene A
synthase. This aspect of the work is currently being
pursued, along with mutagenesis of this enzyme and the
cloning of further S. canadensis sesquiterpene synthases
that must be responsible for the biosynthesis of the
other sesquiterpenes of the pentane soluble profile.
compared with the Wiley library and published data
(Adams, 1995).
Chiral GC–MS was carried out on a 0.25 mm i.d.ꢃ30
m fused silica column with a layer (0.25 mm) of Cyclo-
Sil-B [30% heptakis (2,3-di-O-methyl-6-O-t-butyl dime-
thylsilyl)-b-cyclodextrin in DB-1701] supplied by J and
W Scientific. The column was fitted with a 2 m RGK-1
retention gap (SGE). Injection was either cool on-col-
ꢂ
4. Experimental
umn as above or at 250 C. The temperature program
was as above.
4.1. Plant materials, substrates and standards
4.4. mRNA isolation, cDNA synthesis, PCR-
amplification, probe generation and library preparation
and screening
Seeds of Solidago canadensis var. Gold Baby (Aster-
aceae=Compositae) were purchased from Chiltern
Seeds (Cumbria, UK) and germinated in Levington M2
peat-based compost. Plants were grown in 15 cm pots of
John Innes No 2 compost with a 16 h photoperiod.
[1-3H]-farnesyl diphosphate (FPP) (15–30 Ci/mmol) was
purchased from NEN-Dupont and used as supplied.
Unlabelled FPP was purchased from Sigma, UK.
[1-2H]- and [12,13-2H6]-FPP were synthesised from the
corresponding labelled farnesols by the method of
Davisson et al. (1985). [12,13-2H6]-farnesol was pre-
pared by a published method with minor modifications
(DeRopp and Troy, 1984). Sesquiterpene standards
were obtained as follows:– (+)/(ꢀ)-germacrene D from
hexane extraction and silica chromatography of S.
canadensis aerial tissue; a-humulene was from our
reference collection; black pepper oil containing (ꢀ)-b-
elemene was purchased from Amphora Aromatics,
Bristol, UK and (ꢀ)-germacrene A was a constituent of
the pentane extract of the liverwort Scapania undulata,
which was kindly collected by David Rycroft, University
of Glasgow.
RNA was extracted from S. canadensis leaf tissue by
the method of Verwoerd et al. (1989). mRNA was iso-
lated by affinity chromatography on oligo-dT-cellulose
(Aviv and Leder, 1972). cDNA was prepared using the
Ready To Go kit (Pharmacia) according to the manu-
facturer’s instructions. PCR was performed with the
degenerate forward and reverse primers 50-GAGGA-
TAT(ATC)(TC)T(GATC)GA(GATC)GA(GATC)GC -
30 and 50-TC(GATC)A(GA)(GA)TC(TC)TTCCACCA-
30 in a 50ml reaction containing 50 mM KCl, 10 mM
Tris-HCl pH 8.3, 1.5 mM MgCl2 and 0.001% (w/v)
gelatin and 2 units Taq DNA polymerase (AmpliTaq
Gold, Perkin Elmer). Cycling conditions were initial
ꢂ
denaturation at 94 C for 1min; 35 cycles consisting of
ꢂ
ꢂ
denaturation 1 min, 94 C, annealing 1min, 48 C and
ꢂ
extension 3 min at 72 C; followed byꢂfinal extension at
ꢂ
72 C for 30 min before cooling to 4 C. The resultant
PCR product of approx 290 base-pairs was purified by
electrophoresis on a 1.5% agarose gel, excised, and elec-
troeluted. The cDNA fragment was ligated into the
PCRII vector and cloned using the TA-cloning kit
according to the manufacturer’s instructions (Invitrogen).
Selected colonies containing inserts were grown up in
overnight suspension cultures and plasmid DNA was iso-
lated from the cultures using the Wizard Plus SV Mini-
prep system (Promega). Clones confirmed as containing
inserts of the expected size, by restriction analysis, were
sequenced. A PCR product designated Si6 was identified
as a sesquiterpene synthase by sequence alignment.
The insert of Si6 was labelled with 32P-dCTP using
Ready-To-Go DNA Labelling Beads (Amersham Phar-
macia Biotech) and used to screen a Solidago canadensis
leaf cDNA library constructed in l-ZAPII using the
ZAP-cDNA Synthesis Kit (Stratagene). Hybridisation
and washing conditions were as described by Phillips et
al. (1995). Positive plaques hybridising to Si6 were re-
screened through two further cycles until plaque-pure
and rescued into phagemids as described by the manu-
facturer. Representative clones were analysed for insert
size by PCR and restriction endonuclease analysis and
sequenced.
4.2. Extraction of plant tissue
Leaves of S. canadensis or whole plants of Scapania
undulata were frozen in liquid nitrogen and ground to a
powder at this temperature. After warming to circa 0ꢂC,
the pulverised material was extracted with pentane. The
resultant extract was filtered through a short column of
silica gel (BDH, 40–63 mm)—magnesium sulfate (10:1),
washing through with pentane. The pentane solution was
analysed either directly or after concentration of the
cooled solution with a stream of nitrogen.
4.3. GC–MS analysis
GC–MS was carried out on Kratos MS80RFA
instrument, using cool (ꢁ20 ꢂC) on-column injection to
a BPX5 (SGE) capillary column (0.32 i.dꢃ25 m). The
temperature programme was 45 ꢂC for 1 min, then 5 ꢂC
per min to 210 ꢂC which was held for 30 min. A
straight-chain alkane standard mix was co-injected with
the samples. Spectra and Kovat’s retention indices were

I. Prosser et al. / Phytochemistry 60 (2002) 691–702
701
4.5. E. coli expression
subjected to three cycles of freeze /thaw to lyse the cells.
The expressed protein was purified using a HisꢄBind kit
following the supplied protocol (Novagen). Fractions
were analysed by SDS–PAGE and those containing the
purified protein were pooled and passed through a
NAP-10 column (Pharmacia) equilibrated with SSAB.
Enzyme was stored at ꢀ20 ꢂC in SSAB containing 50%
glycerol.
Oligonucleotide primers were used to amplify the
coding region of the Sc1 cDNA and to introduce clon-
ing sites for subsequent subcloning into pET32b
(Novagen). The amplified product was digested with
EcoRIand XhoIand cloned into pET32b in frame with
thioredoxin. The sequence of the insert was verified
before expression was attempted. For expression, the
recombinant plasmid, designated pET32b:Sc1, was
transformed into chemically-competent E. coli strain
AD494(DE3)pLysS. An overnight starter culture (2 ml)
of pET32b:Sc1 transformed cells was used to inoculate
50 ml LB medium containing ampicillin (100 mg/ml).
4.8. DNA sequencing
The PCR product Si6, the full length cDNA clone Sc1
and the modified cDNA cloned in pET32b were
sequenced on an ABI377 Prism using a BigDye termi-
nator cycle sequencing system (PE Biosystems). Sequence
analysis and alignments used the GCG software.
ꢂ
The culture was grown for 3 h at 37 C (OD600=0.5)
before adding IPTG to 1 mM. The incubation was
continued for a further 4 h, but at 30 C. The cells were
ꢂ
harvested by centrifugation and resuspended in 2 ml of
sesquiterpene synthase assay buffer (SSAB, 100 mM
NaPhosphate pH 7.0, 10 mM MgCl2, 1 mM DTT, 10%
v/v glycerol). The cells were lysed by three cycles of
freeze/thaw and the debris was removed by centrifuga-
tion at 13,000ꢃg. Soluble and insoluble proteins were
analysed by SDS–PAGE.
Acknowledgements
We thank David Rycroft, University of Glasgow, for
the collection and identification of Scapania undulata,
Steve Thomas (LARS) for discussions on cloning tech-
niques and Luke Vincent (LARS) for exploratory
expression studies. The Institute of Arable Crops
Research receives grant-aided support from the Bio-
technology and Biological Sciences Research Council of
the United Kingdom.
4.6. Enzyme assay
Small-scale assays to test for activity were performed
as follows. Up to 100 ml extract was added to 10 ml
(10mg) FPP and 1ml [1-3H] FPP (0.5 mCi/ml), made up to
1 ml with SSAB (see above) and incubated for 2 h at
References
ꢂ
25 C. The reaction mix was extracted with 1 ml pen-
tane, which was then eluted through a short column of
silica gel/MgSO4 (10:1) in a pasteur pipette, washing
through with 1 ml pentane and then 2 ml of diethyl
ether. Aliquots of the pentane and ether fractions were
analysed by liquid scintillation counting (HiSafe scintil-
lant, LKB 1215 Rackbeta). For larger scale assays to
enable identification of the products, either crude lysate
(500 ml) or purified (+)-germacrene A synthase (30 ml)
was added to 50 mg unlabelled or deuterated FPP, 1 ml
[1-3H] FPP, made up to 2 ml with SSAB and incubated
Adams, R. P., 1995. Identification of Essential Oil Components by
Gas Chromatography–Mass Spectroscopy. Allured Publ. Corp.,
Carol Stream, IL.
Aviv, H., Leder, P., 1972. Purification of biologically active globin
messenger RNA by chromatography on oligothymidylic acid-cellu-
lose. Proc. Natl. Acad. Sci. USA 69, 1408–1412.
Back, K., Chappell, J., 1995. Cloning and bacterial expression of a
sesquiterpene cyclase from Hyoscyamus muticus and its molecular
comparison to related terpene cyclases. J. Biol. Chem. 270, 7375–
7381.
Back, K.W., He, S.L., Kim, K.U., Shin, D.H., 1998. Cloning and
bacterial expression of sesquiterpene cyclase, a key branch point
enzyme for the synthesis of sesquiterpenoid phytoalexin capsidiol in
UV-challenged leaves of Capsicum annuum. Plant Cell Physiol. 39,
899–904.
ꢂ
for 2 h at 25 C. The reaction mix was extracted three
times with 0.5 ml pentane/ether (9:1 v/v) and the
organic phases were pooled. The combined pentane/
ether extracts were passed through the silica column and
washed with 0.5 ml pentane/ether. The pentane/ether
fraction was concentrated as described above and was
analysed by GC–MS.
Bohlmann, J., Crock, J., Jetter, R., Croteau, R., 1998. Terpenoid-
based defenses in conifers: cDNA cloning, characterization, and
functional expression of wound-inducible (E)-a-bisabolene synthase
from grand fir (Abies grandis). Proc. Natl. Acad. Sci. USA 95, 6756–
6761.
Bulow, N., Konig, W.A., 1997. Germacrene D—a source of rare ses-
quiterpene hydrocarbons. In: Franz, C.M.A., Buchbauer, G., (Eds).
Essential Oils: Basic and Applied Research; Proc. 27th International
Symposium on Essential Oils, Vienna. Allured Publ. Corp., Carol
Stream, IL, pp. 209–213.
4.7. Purification of recombinant (+)-germacrene A
synthase
A 50 ml culture of E. coli harbouring pET32b:Sc1 was
grown and induced as described above. The harvested
cells were resuspended in 4 ml binding buffer and then
Cane, D.E., 1985. Isoprenoid biosynthesis—stereochemistry of the
cyclization of allylic pyrophosphates. Accounts of Chem. Res. 18,
220–226.

702
I. Prosser et al. / Phytochemistry 60 (2002) 691–702
Cane, D.E., 1990. Enzymatic formation of sesquiterpenes. Chem. Rev.
90, 1089–1103.
Crystal structure of pentalenene synthase: mechanistic insights on
terpenoid cyclization reactions in biology. Science 277, 1820–1824.
Mercke, P., Crock, J., Croteau, R., Brodelius, P.E., 1999. Cloning,
expression, and characterization of epi-cedrol synthase, a sesqui-
terpene cyclase from Artemisia annua L. Arch. Biochem. Biophys.
369, 213–222.
Cane, D.E., Sohng, J.K., Lamberson, C.R., Rudnicki, S.M., Wu, Z.,
Lloyd, M.D., Oliver, J.S., Hubbard, B.R., 1994. Pentalenene syn-
thase—purification, molecular cloning, sequencing, and high-level
expression in Escherichia coli of a terpenoid cyclase from Strepto-
myces Uc5319. Biochemistry 33, 5846–5857.
Niwa, M., Iguchi, M., Yamamura, S., 1980. Co-occurence of (ꢀ) and
(+) germacrene D in Solidago altimissa L. Determination of the
optical rotation of optically pure germacrene D. Chem. Pharm.
Bull. 28, 997–999.
Cane, D.E., Tsantrizos, Y.S., 1996. Aristolochene synthase. Elucida-
tion of the cryptic germacrene A synthase activity using the anom-
alous substrate dihydrofarnesyl diphosphate. J. Am. Chem. Soc.
118, 10037–10040.
Phillips, A.L., Ward, D.A., Uknes, S.J., Appleford, N.E.J., Lange, T.,
Huttly, A.K., Gaskin, P., Graebe, J.E., Hedden, P., 1995. Isolation
and expression of three gibberellin 20-oxidase cDNA clones from
Arabidopsis. Plant Physiol. 108, 1049–1057.
Cane, D.E., 1999. Sesquiterpene biosynthesis: cyclisation mechanisms.
In: Cane, D.E. (Ed.), Isoprenoids Including Carotenoids and Steroids,
Vol. 2. Elsevier Science, Oxford, pp. 155–200.
Caruthers, J.M., Kang, I., Rynkiewicz, M.J., Cane, D.E., Chris-
tianson, D.W., 2000. Crystal structure determination of aris-
tolochene synthase from the blue cheese mold, Penicillium roqeforti.
J. Biol. Chem. 275, 25533–25539.
Proctor, R.H., Hohn, T.M., 1993. Aristolochene synthase — isolation,
characterization, and bacterial expression of a sesquiterpenoid bio-
synthetic gene (Ari1) from Penicillium roqueforti. J. Biol. Chem. 268,
4543–4548.
Chen, X.Y., Chen, Y., Heinstein, P., Davisson, V.J., 1995. Cloning,
expression, and characterization of (+)-d cadinene synthase: a cat-
alyst for cotton phytoalexin biosynthesis. Arch. Biochem. Biophys.
324, 255–266.
Rising, K.A., Starks, C.M., Noel, J.P., Chappell, J., 2000. Demon-
stration of germacrene A as an intermediate in 5-epi-aristolochene
synthase catalysis. J. Am. Chem. Soc. 122, 1861–1866.
Schmidt, C.O., Bouwmeester, H.J., deKraker, J.W., Konig, W.A.,
1998. Biosynthesis of (+)- and (ꢀ)-germacrene D in Solidago cana-
densis: isolation and characterization of two enantioselective ger-
macrene D synthases. Angew. Chem., Int. Ed. 37, 1400–1402.
Schmidt, C.O., Bouwmeester, H.J., Bulow, N., Konig, W.A., 1999a.
Isolation, characterization, and mechanistic studies of (ꢀ)-a- gurju-
nene synthase from Solidago canadensis. Arch. Biochem. Biophys
364, 167–177.
Chen, X.Y., Wang, M.S., Chen, Y., Davisson, V.J., Heinstein, P.,
1996. Cloning and heterologous expression of a second (+)-d-cadi-
nene synthase from Gossypium arboreum. J. Nat. Prod. 59, 944–951.
Colby, S.M., Crock, J., DowdleRizzo, B., Lemaux, P.G., Croteau, R.,
1998. Germacrene C synthase from Lycopersicon esculentum cv.
VFNT Cherry tomato: cDNA isolation, characterization, and bac-
terial expression of the multiple product sesquiterpene cyclase. Proc.
Natl. Acad. Sci. USA 95, 2216–2221.
Schmidt, C.O., Bouwmeester, H.J., Franke, S., Konig, W.A., 1999b.
Mechanisms of the biosynthesis of sesquiterpene enantiomers (+)-
and (ꢀ)-germacrene D in Solidago canadensis. Chirality 11, 353–
362.
Crock, J., Wildung, M., Croteau, R., 1997. Isolation and bacterial
expression of a sesquiterpene synthase cDNA clone from peppermint
(Menthaꢃpiperita L.) that produces the aphid alarm pheromone (E)-b-
farnesene. Proc. Natl. Acad. Sci. USA 94, 12833–12838.
Croteau, R.B., Wheeler, C.J., Cane, D.E., Ebert, R., Ha, H.J., 1987.
Isotopically sensitive branching in the formation of cyclic mono-
terpenes—proof that (ꢀ)-a-pinene and (ꢀ)-b-pinene are synthesized
by the same monoterpene cyclase via deprotonation of a common
intermediate. Biochemistry 26, 5383–5389.
Sessa, R.A., Bennett, M.H., Lewis, M.J., Mansfield, J.W., Beale,
M.H., 2000. Metabolite Profiling of sesquiterpene lactones from
Lactuca species. Major latex components are novel oxalate and sul-
fate conjugates of lactucin and its derivatives. J. Biol. Chem. 275,
26877–26884.
Starks, C.M., Back, K.W., Chappell, J., Noel, J.P., 1997. Structural
basis for cyclic terpene biosynthesis by tobacco 5-epi-aristolochene
synthase. Science 277, 1815–1820.
Davisson, V.J., Woodside, A.B., Poulter, C.D., 1985. Synthesis of
allylic and homoallylic isoprenoid pyrophosphates. Meth. Enzymol.
110, 130–144.
DeRopp, J.S., Troy, F.A., 1984. Chemical synthesis and 2H NMR
investigations of polyisoprenols — dynamics in model membranes.
Biochemistry 23, 2691–2695.
Steele, C.L., Crock, J., Bohlmann, J., Croteau, R., 1998. Sesquiterpene
synthases from grand fir (Abies grandis)—comparison of con-
stitutive and wound-induced activities, and cDNA isolation, char-
acterization and bacterial expression of d-selinene synthase and g-
humulene synthase. J. Biol. Chem. 273, 2078–2089.
Facchini, P.J., Chappell, J., 1992. Gene family for an elicitor-induced
sesquiterpene cyclase in tobacco. Proc. Natl. Acad. Sci. USA 89,
11088–11092.
Verwoerd, T.C., Dekker, B.M.M., Hoekema, A., 1989. A small-scale
procedure for the rapid isolation of plant RNAs. Nucl. Acids Res.
17, 2362.
Glasby, J.S., 1982. Encyclopedia of Terpenoids. Wiley, Chichester.
Hua, L., Matsuda, S.P.T., 1999. The molecular cloning of 8-epicedrol
synthase from Artemisia annua. Arch. Biochem. Biophys. 369, 208–
212.
Wagschal, K., Savage, T.J., Croteau, R., 1991. Isotopically sensitive
branching as a tool for evaluating multiple product formation by
monoterpene cyclases. Tetrahedron 47, 5933–5944.
Hohn, T.M., Plattner, R.D., 1989. Expression of the trichodiene syn-
thase gene of Fusarium sporotrichioides in Escherichia coli results in
sesquiterpene production. Arch. Biochem. Biophys. 275, 92–97.
Kalemba, D., 1998. Constituents of the essential oil of Solidago vir-
gaurea L. Flavour and Fragrance J. 13, 373–376.
Wagschal, K.C., Pyun, H.J., Coates, R.M., Croteau, R., 1994.
Monoterpene biosynthesis—isotope effects associated with bicyclic
olefin formation catalyzed by pinene synthases from sage (Salvia
officinalis). Arch. Biochem. Biophys. 308, 477–487.
Weinheimer, A.J., Youngblood, W.W., Washecheck, P.H., Karns,
T.K.B., Ciereszko, L.S., 1970. Isolation of the elusive (ꢀ)-germa-
crene A from the gorgonian, Eunicea mammosa. Chemistry of coe-
lenterates. XVIII. Tetrahedron Lett. 497–500.
Konig, W.A., Rieck, A., Hardt, I., Gehrcke, B., Kubeczka, K.-H.,
Muhle, H., 1994. Enantiomeric composition of the chiral con-
tituents of essential oils. Part 2. Sesquiterpene hydrocarbons. J.
High Res. Chrom. 17, 315–320.
Williams, D.C., Wildung, M.R., Jin, A.Q., Dalal, D, Oliver, J.S.,
Coates, R.M., Croteau, R., 2000. Heterologous expression and
characterisation of a ‘‘pseudomature’’ form of taxadiene synthase
involved in paclitaxel (taxol) biosynthesis and evaluation of a
potential intermediate and inhibitors of the multistep diterpene
cyclisation reaction. Arch. Biochem. Biophys. 379, 137–146.
deKraker, J.W., Franssen, M.C.R., deGroot, A., Konig, W.A.,
Bouwmeester, H.J., 1998. (+)-Germacrene A biosynthesis—the
committed step in the biosynthesis of bitter sesquiterpene lactones in
chicory. Plant Physiol. 117, 1381–1392.
Lesburg, C.A., Zhai, G.Z., Cane, D.E., Christianson, D.W., 1997.