Natural and directed biosynthesis of communesin alkaloids

Article abstract of DOI:10.1016/j.phytochem.2005.10.011

A role for tryptophan, acetate, mevalonate and methionine in the biosynthesis of communesins A and B, novel structurally-related and biologically-active Penicillium metabolites, has been established by isotopic labelling techniques. The incorporation of ¹⁴C-tryptamine has also been demonstrated. dl-2-¹³C-tryptophan specifically enriched two carbon atoms in the ¹³C NMR spectrum, thereby defining the intra-molecular arrangement of the two tryptophan-derived moieties. Feeding differentially labelled precursors during communesin production showed that tryptophan and methionine are involved early in the biosynthesis and that mevalonate provides an isoprene which is added later. A biosynthetic pathway involving an early precursor based on tryptophan is proposed. Indole-N-( ¹³C-methyl) tryptophan was not incorporated into communesins implying that N-methylation of tryptophan is not the first step of the communesin biosynthetic pathway. During deamination of indole-N-(¹³C-methyl) tryptophan to 1-¹³C-methylindole-3-carboxylic acid communesin biosynthesis was inhibited. Of several halogenated indoles tested for directed biosynthesis, only dl-6-fluoro-tryptophan and 6-fluoro-tryptamine caused accumulation of the corresponding monofluoro-analogues of communesins A and B.

Full text of DOI:10.1016/j.phytochem.2005.10.011

PHYTOCHEMISTRY
Phytochemistry 67 (2006) 561–569
www.elsevier.com/locate/phytochem
Natural and directed biosynthesis of communesin alkaloids
a
a,
b
Lucy J. Wigley , Peter G. Mantle *, David A. Perry
a
Department of Biochemistry, Imperial College of Science, Technology and Medicine, South Kensington, London SW7 2AZ, UK
b
Pfizer Global Research and Development, Groton Laboratories, Eastern Point Road, Groton, CT 06340, USA
Received 10 August 2005; received in revised form 6 October 2005
Available online 1 December 2005
Abstract
A role for tryptophan, acetate, mevalonate and methionine in the biosynthesis of communesins A and B, novel structurally-related
and biologically-active Penicillium metabolites, has been established by isotopic labelling techniques. The incorporation of 14C-trypt-
amine has also been demonstrated. DL-2-13C-tryptophan specifically enriched two carbon atoms in the 13C NMR spectrum, thereby
defining the intra-molecular arrangement of the two tryptophan-derived moieties. Feeding differentially labelled precursors during com-
munesin production showed that tryptophan and methionine are involved early in the biosynthesis and that mevalonate provides an iso-
prene which is added later. A biosynthetic pathway involving an early precursor based on tryptophan is proposed. Indole-N-(13C-methyl)
tryptophan was not incorporated into communesins implying that N-methylation of tryptophan is not the first step of the communesin
biosynthetic pathway. During deamination of indole-N-(13C-methyl) tryptophan to 1-13C-methylindole-3-carboxylic acid communesin
biosynthesis was inhibited. Of several halogenated indoles tested for directed biosynthesis, only DL-6-fluoro-tryptophan and 6-fluoro-
tryptamine caused accumulation of the corresponding monofluoro-analogues of communesins A and B.
Ó 2005 Elsevier Ltd. All rights reserved.
Keywords: Communesin; Alkaloids; Biosynthesis; Radiolabel precursor feeding; Directed biosynthesis; Fluoro-communesin
1. Introduction
between communesins and the plant-derived calycantha-
ceous alkaloids is implied by the presence of a 1,2-diphen-
ylethane group and two aminals. Studies on the biogenesis
of calycanthaceous alkaloids (Robinson and Teuber, 1954;
Henrickson et al., 1964; Kirby et al., 1969) may have rele-
vance to that of the communesins. The present biosynthetic
study was undertaken to seek any biosynthetic features in
common with the plant alkaloids and to provide evidence
for generating communesin analogues with potentially
increased or altered biological activity by directed biosyn-
thesis in microbial fermentation.
The structures of two fungal alkaloids communesin A
and B (1 and 2), isolated from a Penicillium. sp possessing
activity against cultured P-388 lymphocytic leukemia cells,
have been described (Numata et al., 1993). Previously in
the laboratories of Pfizer UK, structurally related com-
pounds were discovered in an antihelminthic screen exhibit-
ing activity against the free-living nematode Caenorhabditis
elegans. The name ‘‘commindolines’’ was used for these
metabolites partly on the basis of initial characterization
of the producing fungus as Penicillium commune and partly
due to the dominant indolic moiety. The structures of the
principal compounds, ‘‘commindolines’’ B and A, equated
to that of communesins A and B, respectively.
2. Results and discussion
2.1. Typical fungal growth and communesin production in
submerged fermentation
Communesins are of mixed biosynthetic origin, predict-
ably derived from tryptophan, mevalonate, acetate and a
methyl group from methionine. A biosynthetic relationship
Communesin accumulation commenced during biomass
accumulation (Fig. 1). However, decrease in biomass after
70 h is probably through respiration of assimilated reserve
*
Corresponding author. Tel.: +44 207 594 5245; fax: +44 207 225 0960.
E-mail address: p.mantle@imperial.ac.uk (P.G. Mantle).
0031-9422/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.phytochem.2005.10.011

562
L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569
carbohydrate. Replicatory growth alone is probably com-
plete by 24 h, which is therefore the time point after which
it is advantageous to add radiolabelled putative precursors
of communesins that are also key intermediates in primary
metabolism. The phase during which communesin concen-
tration in broth continued to increase apparently lasted
until around 200 h for both 1 and 2. HPLC analysis of cell
and broth extracts suggested that 2 is more cell-associated
than 1, possibly attributable to the latterꢀs lipophilic sorbyl
side chain constraining its cellular release.
radiolabelled 1 and 2, as determined by scintillation count-
ing of HPLC-resolved 1 and 2, indicating that these com-
pounds are precursors of communesins (Table 1). The
similar incorporation values recorded for tryptophan,
methionine, mevalonate and tryptamine into 1 and 2 (mean
of 8.25%, 1.0%, 0.59% and 1.75%, respectively), is consis-
tent with a common structure in these communesins and
also provides confidence for the experimental procedure
employed. The incorporation of acetate into 2, 2.3%, was
approximately double that into 1 and equates closely to
the requirement for three acetates in the sorbyl side chain
of 2 as compared to the single acetate of 1, particularly
when the isoprene, common to both, is taken into account.
The high efficiency of the incorporation of radiolabel from
tryptophan contrasted with that of methionine. 2-14C-
Tryptophan theoretically only contributes twice as much
radiolabel to 1 and 2 as methyl-14C-methionine and there-
fore the sevenfold greater incorporation of tryptophan
inversely reflects the relative metabolic demands for trypto-
phan and methionine in primary metabolism. The results
suggest that the N-methyl group of methylated commune-
sins might arise via donation from S-adenosyl methionine.
Reduced competition from other pathways predictably
makes tryptamine an even more efficient experimental
biosynthetic precursor than tryptophan. However, the per-
centage incorporation of 14C-tryptamine into 1 and 2 was
much lower than that of 14C-tryptophan (3.5% and
16.5%, respectively), possibly reflecting vacuole sequestra-
tion (Songstad et al., 1990) or poor uptake of exogenous
tryptamine. Alternatively, if communesin biosynthesis is
catalysed by a multienzyme complex, the low incorporation
could be associated with problems of exogenous trypt-
amine gaining access to the site of catalysis, especially if
biosynthetic intermediates are at some stage enzyme-
bound.
2.2. Communesin biosynthesis: radiochemical experiments
Addition of 14C-tryptophan, 14C-acetate, 3H-mevalo-
nate, (methyl-14C)-methionine and 14C-tryptamine to sepa-
rate submerged fermentations consistently gave rise to
1.6
1.4
1.2
1
1.2
1
0.8
0.6
0.4
0.2
0
0.8
0.6
0.4
0.2
0
2.3. 13C NMR experiment: incorporation of DL[2-13C]-
tryptophan into 1 and 2
Natural abundance 13C NMR spectra of 1 and 2 in ace-
tone-d6showed well-separated signals so that the 13C NMR
spectrum of each communesins isolated from a fermenta-
tion treated with [2-13C]-tryptophan-showed specific
enrichment in the signals at 36.9 and 44.6 ppm, corre-
sponding to C-23 and C-17, respectively. This is consistent
with biosynthesis of communesins from two tryptophan-
derived moieties, the precise orientation of which was thus
confirmed. The intensity ratios of signals corresponding to
0
20 40 60 80 100 120 140 160 180 200 220
Time (h).
Biomass
1
2
Fig. 1. Changes in total communesin yields and biomass over the period
of feeding differentially labelled precursors.
Table 1
Comparison of percentage incorporation of 14C-tryptophan, [1-14C]acetic acid, RS-[2-3H]mevalonic acid, L-(methyl-14C)methionine and L-14C-tryptamine
into 1 and 2
14C-trp (1 lCi)
14C-acetate(1 lCi)
3H-mev (8.5 lCi)
14C-met (10 lCi)
14C-tryptamine (0.14 lCi)
1
2
8.10
8.40
16.50
1.20
2.30
3.50
1.10
0.90
2.00
0.58
0.60
1.18
2.00
1.50
3.50
Total

L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569
563
C-23 and C-17 in the natural abundance spectra of 1 (1.2:1)
and 2 (1.1:1) persist in the enriched spectra, suggesting that
the enrichment of each carbon atom occurred in an even-
handed manner. Using background signals C-3 and C-29
as a reference, with intensity similar to that of C-17 and
C-23 in the natural abundance 13C NMR spectrum, it
was estimated that enrichment in excess of 30-fold for 1
and approximately 18-fold for 2 had been achieved. Such
disparate enrichments could have resulted from different
rates of accumulation of 1 and 2 during the period of iso-
tope addition (up to ca. 100 h, Fig. 1).
principle upon which experiments on temporal involve-
ment of precursors rely. Similar values for 1 and 2 follows
from their identical structure in the regions derived from
tryptophan and methionine. However, the skewed ratio
of label specific activities in communesins following 14C-
3
tryptophan and H-mevalonate administration at various
stages of the fermentation implies the existence of temporal
separation between the biogenic events involving these pre-
cursors (Fig. 2). Until 32 h into the fermentation there was
an overwhelming preference for tryptophan incorporation
into communesin biosynthesis, but later there was a swing
in favour of mevalonate, coincident with the first appear-
ance of end product (Fig. 1); the ratio of 14C: 3H decreased
between 0 and 48 h from 0.83 to 0.33 in 1 and from 0.75 to
0.33 in 2 (Fig. 2). Isolation of 14C-labelled 1 and 2 from cul-
tures fed with differentially labeled precursors before com-
munesins were detectable, i.e., before 32 h, signifies the
involvement of an early, tryptophan-based, biosynthetic
intermediate. These results suggest that during the early
part of the fermentation there is a build up of ꢁearly trypto-
phan-based intermediatesꢀ which later give rise to recognis-
able communesins via completion steps, e.g., prenylation.
Synthesis of enzymes catalysing later steps in the biogenesis
of these secondary metabolites might be in response to
accumulation of critical substrate levels (i.e., accumulation
of specific biosynthetic intermediates) and may account for
the sudden appearance of communesins, coinciding with
the preferential incorporation of mevalonate. Results of
administration of differentially-labelled precursors to Peni-
cillium sp. isolate N934-53 therefore indicates that trypto-
phan and methionine are involved early in communesin
biosynthesis and that mevalonate provides the isoprene
which is added later.
2.4. Dual label experiments, exploring biosynthetic sequence
3
via the mixtures 14C-tryptophan: H-methionine or 14C-
3
tryptophan: H-mevalonate
If two precursors are involved concurrently in biosyn-
thesis of any molecule throughout product accumulation,
a mixture of the two precursors, given at any stage, will
become incorporated into end product in an approximately
constant ratio (Mantle and Shipston, 1987). Therefore, to
gain information on the sequence of biosynthetic events,
the incorporation of two precursor permutations 14C-tryp-
3
3
tophan: H-methionine or 14C-tryptophan: H-mevalonate
into 1 and 2 at various stages in the fermentation was
investigated.
The approximately constant ratio, 1:1, of 14C-trypto-
phan- and H-methionine-derived label incorporation in 1
3
and 2 throughout the fermentation indicated that these pre-
cursors were involved in the biosynthesis at about the same
time (Fig. 2), and demonstrated validity in this case of the
2.5. Biotransformation of indole-N-(methyl-13C) tryptophan
Indole-N-(methyl-13C) tryptophan was not incorporated
into 1 or 2, which implied at first sight that methylation of
tryptophan was not the first biosynthetic step, consistent
with the dual-label experiments. However, exogenous
indole-N-(methyl-13C) tryptophan was transformed to a
compound identified as 1-methylindole-3-carboxylic acid
(Fig. 3), as determined by MS and NMR spectroscopy
(Wigley, 1995). Also communesins were not biosynthesised
in the presence of indole-N-(methyl-13C) tryptophan as
suggested by the absence of radiolabeled communesins in
Fig. 2. Changes in the 14C: 3H ratios in 1 and 2, attributed to radiolabelled
communesin precursors from standard 14C-tryptophan: 3H-methionine or
14C-tryptophan: 3H-mevalonate mixtures, measured at 186 h (solid lines)
and 192 h (dotted lines), respectively, following addition to individual
cultures at various stages through the fermentation.
Fig. 3. Biotransformation of indole-N-(13C-methyl)-tryptophan to indole-
N-(13C-methyl)-3-carboxylic acid.

564
L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569
cultures treated with indole-N-(methyl-13C) tryptophan
spiked with 14C-tryptophan.
methylation, analogous to the biosynthesis of the related
calycanthaceous alkaloids (Robinson and Teuber, 1954),
is more plausible because methylation does not appear to
be a crucial step for the biosynthesis of all communesins.
In the biosynthesis of the catharanthaceous alkaloids
two asymmetric centres (at the b-indolic positions) are
formed upon dimerisation and thus two diastereomers of
tetra-aminodialdehyde are possible; one is meso and the
other racemic or optically active, each giving rise to the five
isomers (a-e) with corresponding optical activity (Robinson
and Teuber, 1954). In the meso series however, only the
disymmetric e isomer has potential optical activity (Hen-
rickson et al., 1964). a, b and d are the only isomers that
have previously been isolated from natural sources, and
identified as (+)-calycanthine, (rac)-chimonanthine and
iso-calycanthine, respectively (Adjibade et al., 1992). Com-
munesins possibly represent the e isomer isolated, for the
first time to our knowledge in this form, from a natural
source.
Plant hormones, indolacetic acid (IAA), 1-naphthalene-
acetic acid (NAA) and 2,4-dichlorophenoxyacetic acid
(2,4-D) down-regulate transcription of the tryptophan
decarboxylase (TDC) gene in cultured Catharanthus roseus
cells. Thus negative regulation by IAA has been suggested
as a mechanism controlling terpenoid indole alkaloid bio-
synthesis in plants (Goddijn et al., 1992). It is therefore
possible that the structurally related 1-methylindole-3-car-
boxylic acid could have a similar affect in fungi and may
explain why communesins were not produced in the cul-
ture fed with indole-N-(methyl-13C) tryptophan. The cellu-
lar uptake, fungal transformation capacity and effect upon
communesin production of this foreign compound was
however, demonstrated.
The accumulated evidence concerning communesin bio-
synthesis suggests that the decarboxylation of tryptophan
to yield tryptamine is the first biogenic step (rather than
methylation of tryptophan), complementary to reports by
various authors that tryptophan decarboxylase forms a
regulatory link between primary and secondary metabo-
lism (Aerts et al., 1992 and Noe et al., 1984). The subse-
quently formed tryptamine is then methylated either
before or after bb0-oxidative dimerisation with a second
tryptamine moiety. Prenylation and acetylation of the
putative di-tryptamine intermediate takes place at later
stages (Fig. 4). Formation of a di-tryptamine moiety (by
bb0-oxidative dimerisation of N-methyl tryptamine, hydro-
lysis of which gives rise to tetra-aminodialdehyde) prior to
2.6. Suppression by ethionine of the methylation step in
communesin biosynthesis to reveal an unmethylated
communesin
Ethionine, as an exogenous competitive inhibitor of
methylation, was used to alter flow of the communesin
pathway to 1 and 2 and thus potentially to select for a dem-
ethyl-communesin. Preliminary experiments showed incor-
poration of 14C from 14C-methionine into 1 and 2, and
other methionine-derived metabolites (e.g., ergosterol) in
Fig. 4. Partial biosynthetic pathway for communesin alkaloids proposed on the basis of the present experimental findings and known biosynthesis of
Catharanthaceous alkaloids.

L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569
565
scale with these halogenated indoles unlabelled. Also tested
similarly were 6-fluoro-tryptamine hydrochloride, 5-fluoro-
tryptamine and 5-methyl tryptophan.
The mass spectra of 1 and 2 that had been generated in
the presence of 6-fluoro-tryptamine hydrochloride con-
tained, in addition to ions characteristic of native 1 and
2, ions corresponding to the communesin molecular ion
plus 18 amu (m/z 474 and 526 for 1 and 2, respectively).
Similar data was obtained for 2 from culture fed DL-6-flu-
orotryptophan. The elemental composition of these new
ions, deduced from accurate mass measurement, was
C28H32N4O2F (m/z 474) and C32H36N4O2F (m/z 526),
consistent with the incorporation of only one fluori-
nated-indole moiety, thus demonstrating the feasibility
of generating an analogue of the various communesins
by substrate feeding. Absence of ions 36 amu greater than
the native communesin molecular ion, implies that di-fluo-
rinated communesins were not formed. Presence of frag-
ment ions in the EI mass spectra corresponding to
fluorination of specific native communesin fragment ions,
but not of others, is in accordance with the fluoro-group
being located exclusively on the ‘‘non-prenylated’’ side of
the molecule (Fig. 6). It would seem that the 6-fluoro-
tryptophan and -tryptamine derivatives are unacceptable
substrates for insertion into the ultimately-prenylated
moiety of communesins. However, insertion specificity
could be a useful tool in generating new communesin ana-
logues biosynthetically.
The position of fluorine in the ꢁtryptophan-ringꢀ moiety
of fluoro-communesins derived from fluoro-tryptophan
was assumed to be the same as in the parent precursor.
The absence of a fluorine atom on the prenylated indole
moiety suggests enzyme specificity at some stage in the
biosynthesis, e.g., specificity of the dimerase, methylase
or enzymes involved in prenylation whereby DL-6-fluoro-
tryptophan or 6-fluoro-tryptamine cannot be accepted as
sole indolic substrates for communesin biosynthesis. Since
many secondary metabolite biosynthetic pathways are
catalysed by multienzyme complexes (Gutierrez et al.,
1991; Kleinkauf and von Dohren, 1990; Beck et al.,
1990), it is conceivable that a metabolite precursor
remains bound to the enzyme complex during a number
of sequential biosynthetic steps. For example, tryptophan,
implicated early in the biosynthesis, could be the starter
unit, enzyme-bound to a very specific active site via inter-
actions involving the aromatic ring. It seems rational
therefore that tryptophan analogues can not be tolerated
at this stage due to enzyme specificity, but this does not
eliminate the possible acceptance of a second, modified
tryptophan. It is probable that the presence of a highly
electronegative fluorine atom would alter some chemical
properties of communesins. However, low enrichment of
native communesins with co-chromatographing 3-fluoro-
communesins generated by directed biosynthesis made
accurate biological assessment impossible, a problem also
encountered in the generation of some squalestatin ana-
logues via directed biosynthesis (Cannell et al., 1993).
Fig. 5. Autoradiograph showing effect of ethionine (right), added to a
fermentation at 48 h, on incorporation of 14C-tryptophan into metabolites
at 72 h, when labelled tryptophan was added 30 min after the ethionine
and to a control fermentation (left).
cell extract, was markedly reduced after treating fermenta-
tions with ethionine (50 mg per 100 mlꢀ1) in early idiophase
and adding 14C-methionine 30 min after the ethionine (Wig-
ley, 1995). A similar experiment, substituting 14C-trypto-
phan for 14C-methionine so as to detect previously
unrecognised indole alkaloids according to their biosynthe-
sis from a tryptophan precursor, clearly showed reduced
yields of 1 and 2 in broth extract (Fig. 5). Similar reduction
in the trace amount of an unidentified compound (b) would
be consistent with it being a closely related communesin.
However, another minor compound (a), that was not evi-
dent on TLC chromatograms under UV254nm light and
had only been recognised on account of the radioactivity
in the control extract, was much increased in response to
ethionine, implying that it may be a de-methyl communesin.
2.7. Directed biosynthesis of communesin analogues
Ability of Penicillium sp. N934-53 to accept biosynthetic
precursor analogues that could direct biosynthesis towards
communesin analogues was explored first using 14C
labelled tryptophan analogues as qualitative probes. Since
such were not available commercially, sufficient of several
probes (14C-bromotryptophan substituted in the 4 or 5
position, 14C-fluorotryptophan substituted in the 4 or 6
position, 14C-5-nitrotryptophan) was prepared by an
E. coli expressing high tryptophan synthetase activity (see
Section 3.2.8). Radiolabelled 4-fluorotryptophan, 4-brom-
otryptophan and 5-nitrotryptophan were not incorporated
into 1 and 2, showing that the biosynthetic enzymes were
too specific for these substrates. However, for the haloge-
nated analogues, 8% and 11%, respectively, of the 14C
given was found in fungal protein. Additional amounts
were in non-communesin metabolites, indicating uptake
and acceptance of these tryptophan analogues into other
metabolic pathways. In contrast, 6-fluorotryptophan and
5-bromotryptophan were incorporated into analogues of
1 and 2 and this finding prompted experiments on a larger

566
L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569
Fig. 6. Interpretation of EI mass spectrum fragmentation patterns of mono-fluorinated 1 and 2.
The incorporation of 6-fluoro-tryptamine into commune-
sins provides further evidence that tryptamine is a biosyn-
thetic precursor.
Mass spectra of communesins biosynthesised in the pres-
ence of DL-5-bromo-tryptophan, 5-chloro-tryptamine and
DL-5-methyl-tryptophan only contained signals associated

L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569
567
with native communesins (Wigley, 1995), suggesting that
these putative precursor analogues can not be incorporated
into communesins. The preferential incorporation of fluo-
rine reported in the directed biosynthesis of squalestatin
analogues, was attributed to the almost isosteric nature of
hydrogen and fluorine atoms (Cannell et al., 1993) and pos-
sibly accounts for the same phenomenon observed during
communesin analogue generation.
calis (Sigma). Dry cells 1 mg, were added to DL-
tryptophan, 0.8 mg plus 14C-L tryptophan, 5 lCi, in
50 mM Tris–HCl, containing 0.1 mM pyridoxal phosphate,
1 ml (Taylor and Wightman, 1987), at pH 5.5, incubated at
37 °C for 48 h. Cells were removed by centrifugation and
tryptamine extracted from the supernatant, basidified to
pH > 12 with NaOH, by partition into diethyl ether and
purified by thin layer chromatography (as in 3.2.X), mobile
phase: methyl acetate: propan-2-ol: 25% ammonium
hydroxide, 9:3:1. 14C-tryptamine: specific activity
(7.2 lCi mmolꢀ1), 0.36 lCi total, was isolated.
3. Experimental
3.1. Fungal isolate, media and fermentation conditions
3.2.3. Addition of differentially labelled precursors at various
stages in the submerged culture fermentation
Sporulating communesin-producing cultures of Penicil-
lium sp. isolate N934-53 (from soil, Miyazaki, Japan; Pfizer
culture collection) were stored at ꢀ20 °C following growth
at 25 °C for 7–10 days on Bacto potato dextrose agar
(PDA) slopes. Spores were transferred, as an opaque green
suspension in a 0.01% Tween 80 solution, 2 ml, into baffled
Erlenmeyer flasks (500 ml) containing seed-stage medium,
(g/l: Corn starch 20, Pharmamedia 15, yeast extract 5,
CaCO3 2, pH 7.0) 100 ml. Flasks were incubated at 27 °C
on a rotary shaker (200 rpm, 10 cm eccentric throw) for
approximately 36 h, giving a dense mycelial suspension,
when 4% (v/v) transfer was made to production media,
(ARM; g/l: Soluble starch 7.5, Trusoy flour 7.5, sodium
chloride 2, yeast extract 0.05, Pharmamedia 2, pH 6.5),
100 ml in 500 ml un-baffled Erlenmeyer flasks, and incu-
bated as above for between 96 and 144 h.
(a) Delivery of a mixture of 14C-tryptophan and 3H-
methionineA solution of 14C-tryptophan and L-
[methyl-3H] methionine (aqueous solution containing
0.2% 2-mercaptoethanol, sterilised): specific activity,
70–85 Ci mmolꢀ1 (Amersham International); in the
ratio of 1:8 counts per minute (cpm), was delivered
to production stage cultures at either time 0, 8, 24,
32, 48, 72 or 96 h. All cultures were harvested at
168 h.
(b) Delivery of a mixture of 14C-tryptophan and RS-
(2-3H)mevalonic acidA solution of 14C-tryptophan
and DL-[2-3H]mevalonic acid lactone (toluene solu-
tion): specific activity 1 Ci mmolꢀ1, in the ratio of
1:30 (cpm), was delivered to production stage cultures
at either time 0, 12, 24, 48, 72, 96, 120 or 144 h. All
cultures were harvested at 192 h. The ratio (cpm) of
14C: 3H in 1 and 2 (measured in the total of each com-
pound extracted) was determined by scintillation
counting (see the following section).
3.2. Isotopically-labelled-precursor feeding experiments
3.2.1. Radiolabelled putative precursor addition
L-[side chain (methylene)-3-14C] tryptophan (aqueous
solution containing 2% ethanol): specific activity,
53.80 mCi mmolꢀ1 (New England Nuclear), 1 lCi; [1-14C]
acetic acid, sodium salt (aqueous solution): specific activity,
60 mCi mmolꢀ1 (Amersham International), 1 lCi; RS-
[2-14C] mevalonic acid lactone (benzene solution): specific
activity 50–60 mCi mmolꢀ1 (Amersham International),
DL-[2-3H] mevalonic acid lactone (toluene solution): specific
activity 1 Ci mmolꢀ1, 8.5 lCi; L-[methyl-14C]met (freeze-
dried, under nitrogen): specific activity, 50–60 mCi mmolꢀ1
(Amersham International), 10 lCi; L-14C tryptamine,
0.14 lCi (see below for preparation): specific activity, ca.
7.2 lCi mmolꢀ1. All compounds in sterile distilled water,
1 ml, were delivered to separate cultures of Penicillium sp.
N934-53 in four equal sub-doses at 48, 64, 68 and 72 h.
The tryptamine and tryptophan-fed cultures were har-
vested at 80 and 144 h respectively; all others were har-
vested at 96 h.
3.2.4. Metabolite extraction and analysis
Fungal mycelium, separated under vacuum from broth
by filtration (Whatman 541, 9 cm diameter) was extracted
twice with methanol (25 ml per 100 ml of culture) for
12 h. Metabolites in broth (filtrate) that are soluble in
organic solvent were isolated by partition extraction
twice with an equal volume of ethyl acetate. Combined
methanol and ethyl acetate extracts were taken to dry-
ness under reduced pressure prior to analysis by HPLC
and/or TLC. TLC employed silica gel, 0.25 mm thick
on plastic backed plates, (Polygram SIL G/UV254, Cam-
lab). In the mobile phase chloroform:acetone (6:1), 1 and
2 had Rf values 0.3 and 0.48, respectively. HPLC was
performed on a Waters Z-module C-18 reverse phase
‘‘Novapak’’ (0.8 cm by 10 cm) HPLC column controlled
by an Apple II GS computer using the Gilson HPLC
system manager with detection at 268 nm. Communesins
were purified and analysed using methanol:water 80:20,
flow 0.75 ml minꢀ1, Rt for 1 and 2 10 and 14 min,
respectively.
3.2.2. Preparation of 14C-tryptamine by decarboxylation of
14C L-tryptophan by L-phenylalanine decarboxylase
Decarboxylation of trp was achieved using L-phenylala-
nine decarboxylase (EC 4.1.1.53) from Streptococcus fae-
Radiolabel incorporation was determined quantitatively
by scintillation counting (Kontron Intertechnique) of the

568
L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569
HPLC-purified communesins [sample:scintillant (Eco-
120 h and organic solvent extracts were analysed by HPLC
and TLC.
3
lumee ICN) ca. 8:1]. Counting efficiency for 14C and H
was 90–93% and ca. 40%, respectively. Semi-quantitative/
qualitative analysis was by autoradiography of TLC-
resolved extracts (Fuji X-ray film NIF RX developed auto-
matically in a Fuji FPM 2100 processor).
3.2.8. Biosynthesis of tryptophan analogues
Escherichia coli, W 3110 trpAEdel2, containing the plas-
mid pHP3 encoding tryptophan synthetase, can convert L-
serine and indole to L-tryptophan (Matthews et al., 1992).
Provision also of plasmid encoded marker genes for ampi-
cillin and tetracycline resistance facilitates selection and
maintenance of bacteria containing plasmids (Enger-Valk
et al., 1980). The following protocol for bacterial growth
and the incubation procedure for tryptophan synthesis
was modified from Matthews et al. (1992).
Mass spectrometry (MS) used either a VG Micromass
7070E instrument, operated at 70 eV for electron impact
(EI) or a VG Auto Spec instrument at 70 or 100 eV for
EI and chemical ionisation (CI), respectively. Ammonia
was the reagent gas for CI.
1H and 13C NMR spectroscopy used a Brucker AM
500 MHz instrument. Chemical shift values were all rela-
tive to tetramethylsilane.
The E. coli strain was grown and maintained in medium
2TY (Maniatis et al., 1989); Bacto tryptone 16 g/l, Bacto
yeast extract 10 g/l, NaCl 5 g/l. pH was adjusted to 7.0 with
5 M NaOH. Filter-sterilised ampicillin (Acrodisc, 0.2 lm),
50 lg/ml, was added to the sterilised medium. For solid
medium Bacto agar, 15 g/l, was added. Liquid cultures
were grown for 15 h at 37 °C. Sterile glycerol was added
(10%) and 1 ml aliquots dispensed into sterile eppendorff
tubes. Tubes were cooled slowly to ꢀ20 °C before storage
in liquid nitrogen. For fermentations, the contents of one
thawed tube was used to inoculate 200 ml of 2TY/ampicil-
lin medium and incubated on a rotary shaker for 15 h at
37 °C. Cells were harvested by centrifugation and the pellet
re-suspended in 100 mM potassium phosphate buffer, pH
7.8, supplemented with pyridoxal phosphate, 10 mg/l,
(NH4)2SO4, 15.6 g/l, Na2SO3 Æ 7H2O, 1.25 g/l. To cells
from ca. 50 ml culture, suspended in the supplemented buf-
fer, serine (151 mg) and 4-bromo-indole (36 mg) were
added. Since the bromo-indole is only slightly soluble in
water it was ꢁtrickle-fedꢀ in 18 drops of DMSO during the
first 3 h of incubation. Incubation was for 6 h at 37 °C
and, after centrifugation, the supernatant was treated with
5 volumes of ethanol (96%) to precipitate salts. Evapora-
tion of the supernatant left a residue rich in tryptophan
which was purified by PLC on 1 mm silica gel plates (Poly-
gram SIL G/UV254, Camlab) in n-butanol:acetic acid:water
(12:3:5, Rf values for tryptophan and analogues in the
range 0.7–0.8): yield 66%. 4-Bromotryptophan structure
was confirmed (NMR, UV and EIMS) by Prof C. Moody
(Loughborough University, UK).
3.2.5. Dry cell weight
Dry cell weight as a measure of fungal growth was deter-
mined by lyophilisation of filtered mycelium in pre-weighed
flasks to constant weight.
3.2.6. Administration of DL-2-13C-tryptophan to submerged
fermentations
Preliminary experiments, measuring incorporation effi-
ciency of 14C-tryptophan into communesins in the presence
of significant additions of tryptophan (Wigley, 1995),
showed less than a 3-fold reduction when 10 mg tryptophan
was added to a 100 ml fermentation. With 20 mg the effi-
ciency was correspondingly halved, and thus there was no
advantage in adding more than 10 mg 13C-tryptophan for
13C NMR enrichment. Further, maximum incorporation
into communesins of a single probe-dose of tryptophan,
added to a fermentation at 96 h was only reached after a
further 24 h. Therefore, near-optimal administration of a
13C-tryptophan probe was designed. DL-2-13C-tryptophan
(99 atom% 13C, C/D/N Isotopes INC), 10 mg, was dis-
solved in distilled water, 25 ml, and filter sterilised (Acro-
discÒ, 0.2 lm). Aliquots were added to a 100 ml culture
on 5 evenly-spaced occasions during the 3rd and 4th day
of fermentation. The probe was therefore added between
46.5 and 75.5 h, and there were four replicate 100 ml cul-
tures. The cultures were harvested at 78.5 h to prevent
undue dilution of 13C-communesins by new unlabelled
compound. 13C-tryptophan-enriched 1 and 2, 9.9 and
7.3 mg, respectively, were purified, dissolved in acetone-d6,
analysed by 13C NMR and the spectra compared to the nat-
ural abundance 13C NMR spectra of 1 and 2.
Similar protocol on a smaller scale was used to make
radiolabelled tryptophan analogues from 4-bromo-indole,
4-fluoro-indole, 5-bromo-indole, 6-bromo-indole and
appropriate L-serine by adding 5–10 lCi 14C-serine. Prod-
ucts, purified by TLC, were added (total 0.1–0.4 lCi accord-
ing to availability) to Penicillium fermentations in aliquots
at 52.5, 53, 69.5, 70.25, and 72 h and the cultures harvested
at 96 h. Communesins were isolated, separated by TLC and
the chromatograms autoradiographed (Fuji NIF RX).
3.2.7. Administration of indole-N-(methyl-13C) tryptophan
to submerged fermentations
Indole-N-(methyl-13C)tryptophan, 50 mg per 100 ml
culture (pH 5.0) (prepared by G. Goodwin, Pfizer, Sand-
wich by the method of Yamada et al., 1965) was adminis-
tered to ten production stage cultures, 100 ml, in eight 1 ml
aqueous aliquots between 27.5 and 48 h of the fermenta-
tion. One culture flask also received 1 lCi 14C-tryptophan.
Control cultures were fed distilled water (pH 5.0) or were
not treated in any way. All cultures were harvested at
3.2.9. Preparation of halogenated and methylated
communesins by directed biosynthesis
DL-6-Fluorotryptophan (Aldrich) 11.9 mg, DL-5-brom-
otryptophan (Sigma) 11.9 mg, 6-fluorotryptamine (Sigma)

L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569
569
Cannell, R.J.P., Dawson, M.J., Hale, R.S., Hall, R.M., Nobel, D., Lynn,
S., Taylor, N.L., 1993. The squalestatins, novel inhibitors of squaline
10 mg, 6-chlorotryptamine (Sigma) 10.6 mg and DL-5-
methyltryptophan (Aldrich) 10 mg, were ꢁtrickle-fedꢀ, each
in 25 ml sterile distilled water, to separate culture flasks
of production medium, 100 ml, frequently between 40
and 69 h of the fermentation, which continued until 96 h.
The ethyl acetate broth extracts and methanol cell extracts
were purified by TLC, and compounds co-chromatograph-
ing with 1, 2 and other selected bands were eluted from the
silica and investigated by CIMS. Accurate mass measure-
ments were obtained for any important ions in the spectra
that did not correspond to either 1 or 2.
synthase produced by
a species of Phoma 1V. Preparation of
fluorinated squalestatins by directed biosynthesis. Journal of Antibi-
otics 46, 1381–1389.
Enger-Valk, B.E., Heyneker, H.L., Oosterbaan, R.A., Pouwells, P.H.,
1980. Construction of new cloning vehicles with genes of the
tryptophan operon of Escherichia coli as genetic markers. Gene 9,
69–85.
Goddijn, O.J.M., De Kam, R.J., Zanetti, A., Schilperoort, R.A., Hoge,
J.H.C., 1992. Auxin rapidly down-regulates transcription of the
tryptophan decarboxylase gene from Cantharanthus roseus. Plant
Molecular Biology 18, 1113–1120.
Gutierrez, S., Diez, B., Montenegro, E., Martin, J.F., 1991. Character-
isation of the Cephalosporium acremonium pcbAB gene encoding a-
aminoadipyl-cysteinyl-valine synthetase, a large multidomain peptide
synthetase: Linkage to the pcbC gene as a cluster of early cephalo-
sporin biosynthetic genes and evidence of multiple functional domains.
Journal of Bacteriology 173, 2354–2365.
Henrickson, J.B., Goschke, R., Rees, R., 1964. Total synthesis of the
calycanthaceous alkaloids. Tetrahedron 20, 565–579.
Kirby, G.W., Shah, S.W., Herbert, E.J., 1969. Biosynthesis of chimonan-
thine from [2-3H] tryptophan and [2-3H] tryptamine. Journal of the
Chemical Society (C), 1916–1919.
3.2.10. Probing for un-methylated communesins with 14C-
tryptophan following ethionine addition to fermentations
DL-Ethionine (Sigma), 50 mg in 10 ml sterile distilled
water, was added to a 100 ml fermentation at 38 h. 10 ml
water was added to a control fermentation. 30 min later
14C-tryptophan (1 lCi) in 1 ml sterile water was added to
each culture. Fermentations were harvested at 72 h and
the broth extracted with ethyl acetate. Extract was chro-
matographed and the TLC plate autoradiographed for
one month.
Kleinkauf, H., von Dohren, H., 1990. Nonribosomal biosynthesis of
peptide antibiotics. European Journal of Biochemistry 192, 1–15.
Maniatis, T., Fritsch, E.L., Sambrook, J., 1989. Molecular cloning, a
laboratory manual. Cold Spring Harbor, New York.
Mantle, P.G., Shipston, N.F., 1987. Temporal separation of steps in the
biosynthesis of verruculogen. Biochemistry International 14, 1115–
1120.
Acknowledgements
Matthews, S., Jandu, S.K., Leatherbarrow, R.J., 1992. 13C NMR study of
the effects of mutation on the tryptophan dynamics in chymotrypsin
inhibitor 2: correlations with structure and stability. Biochemical
Journal 32, 657–662.
Noe, W., Mollenschott, C., Berlin, J., 1984. Tryptophan decarboxylase
from Catharanthus roseus cell suspension cultures: purification,
molecular and kinetic data of the homogeneous protein. Plant
Molecular Biology 3, 281–288.
Numata, A., Takahashi, C., Ito, Y., Takada, T., Kawai, K., Usami, Y.,
Matsumura, E., Imachi, M., Ito, T., Hasegawa, T., 1993. Commune-
sins, cytotoxic metabolites of a fungus isolated from a marine alga.
Tetrahedron Letters 34, 2355–2358.
We thank R. Shepherd and J. Barton (Chemistry Depart-
ment, Imperial College) for obtaining NMR and MS data,
G. Goodwin (Pfizer, UK) for synthesis of indole-N-(13C-
methyl) tryptophan, and Prof. C. Moody, Loughborough
University, UK, for confirming our tryptophan synthesis.
We thank BBSRC for a research studentship (LJW) in col-
laboration with Pfizer Central Research (Sandwich, Kent,
UK) and acknowledge the advice of Dr. M. Haxell (Pfizer,
UK) concerning communesin fermentations.
Robinson, Sir R., Teuber, H.J., 1954. Calycanthine and calycanthidine.
Chemistry and Industry, 783–784.
References
Songstad, D.D., De Luca, V., Brisson, N., Kurz, W.G.W., Nessler, C.L.,
1990. High levels of tryptamine accumulation in transgenic tobacco
expressing tryptophan decarboxylase. Plant Physiology 94, 1410–
1413.
Taylor, D.C., Wightman, F., 1987. Metabolism of D,L-chloro-phenylala-
nines by phenylalanine aminotransferase isozymes purified from
bushbean shoots. Phytochemistry 26, 1279–1288.
Wigley, L.J. 1995. Biosynthesis of communesin alkaloids. Ph D thesis,
University of London.
Yamada, S., Shioiri, T., Itaya, T., Hara, T., Matsueda, P., 1965. Ind.-N-
alkylation of tryptophan and synthesis of 1-alkyltryptophan hydraz-
ides. Chemical and Pharmaceutical Bulletin 13, 88–93.
Adjibade, Y., Weniger, B., Quirion, J.C., Kuballa, B., Cabalion, P.,
Anton, R., 1992. Dimeric alkaloids from Psychotria fosteriana.
Phytochemistry 31, 317–319.
Aerts, R.J., Alarco, A.-M., De Luca, V., 1992. Auxins induce tryptophan
decarboxylase activity in radicles of Catharanthus seedlings. Plant
Physiology 100, 1014–1019.
Beck, J., Ripka, S., Siegner, A., Schiltz, E., Schweizer, E., 1990. The
multifunctional 6-methylsalicylic acid synthase gene of Penicillium
patulum. Its gene structure relative to that of other polyketide
synthases. European Journal of Biochemistry 192, 487–498.