Reconstruction of pyrrolo[2,3-b]indoles carrying an α-configured reverse C3-dimethylallyl moiety by using recombinant enzymes

Article abstract of DOI:10.1039/b922440h

Nine reversely C3-prenylated pyrrolo[2,3-b]indoles were successfully prepared by using two recombinant enzymes involved in the biosynthesis of acetylaszonalenin from Neosartorya fischeri. The prenyltransferase AnaPT catalysed the conversion of six tryptop

Full text of DOI:10.1039/b922440h

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Reconstruction of pyrrolo[2,3-b]indoles carrying an a-configured reverse
C3-dimethylallyl moiety by using recombinant enzymes†
Wen-Bing Yin,a Xiu-Lan Xie,b Marco Matuscheka and Shu-Ming Li*a
Received 28th October 2009, Accepted 4th December 2009
First published as an Advance Article on the web 7th January 2010
DOI: 10.1039/b922440h
Nine reversely C3-prenylated pyrrolo[2,3-b]indoles were successfully prepared by using two
recombinant enzymes involved in the biosynthesis of acetylaszonalenin from Neosartorya fischeri. The
prenyltransferase AnaPT catalysed the conversion of six tryptophan-containing cyclic dipeptides to
reversely C3-prenylated indoline derivatives. Using cyclo-L-Trp-L-Trp as substrate, both mono- and
diprenylated indolines were obtained. Two of the AnaPT products were acetylated at position N1 by
the acetyltransferase AnaAT. The structures of the obtained compounds were characterised by
HR-ESI-MS, 1H- and 13C-NMR analyses as well as by long-range 1H-13C connectivities in
heteronuclear multiple-bond correlation (HMBC) spectra after preparative isolation. Their absolute
configurations were determined by analysing the 1H-1H spatial correlations in rotating-frame nuclear
overhauser effect spectroscopy (ROESY).
Introduction
Natural products from plants and microorganisms are among
the most important sources of leading compounds for drug
development.1 Prenylated indole alkaloids are mainly found in
the genera of Penicillium and Aspergillus of Ascomycota2,3 and
often possess biological activities clearly distinct from their
non-prenylated aromatic precursors.4,5 A subgroup of these
compounds, C3-prenylated indoline alkaloids, usually carry a
reverse prenyl moiety at position C3 of the indoline ring and
a five-membered ring system between the indoline and the
diketopiperazine ring. A prominent example is the mycotoxin
roquefortine C, which was isolated firstly from Penicillium roque-
forti6 and identified later in many Penicillium strains.7,8 Some
of these natural products showed a wide range of biological
and pharmacological activities. For example, brevicompanines
A, B and C produced by P. brevicompactum were identified as
plant growth regulators.9,10 Fructigenines A and B (verrucofortine)
Fig. 1 Examples of C3-prenylated pyrrolo[2,3-b]indoles.
(Fig. 1) from P. fructigenum and P. verrucosum var. cyclopium11,12
The C3-prenylated indoline derivatives mentioned above are be-
were reported to show an inhibitory effect on plant growth.
lieved to be formed biosynthetically by prenylation of tryptophan-
Rugulosuvines A and B (Fig. 1) from P. rugulosum and P. piscarium
containing cyclic dipeptides and subsequent modifications such as
had moderate cytotoxicity against L-929, K562 and Hela cells.13
acetylation at position N1.2 An experimental proof was provided
Amauromine (Fig. 1) from Amauroascus sp. was reported to have
recently by identification of the biosynthetic gene cluster of
a vasodilating activity.14 Its stereoisomer epiamauromine (Fig. 1)
acetylaszonalenin from N. fischeri.17 A putative prenyltransferase
from Aspergillus ochraceus, on the other hand, showed insecticidal
gene anaPT and a putative acetyltransferase gene anaAT from
activity.15
this cluster were cloned and overexpressed in E. coli, respectively.
Acetylaszonalenin is a mycotoxin, which was isolated initially
The overproduced recombinant proteins AnaPT and AnaAT
from an unidentified Aspergillus species.16 Together with its non-
were purified to homogeneity and characterised biochemically.
acetylated form aszonalenin, it was also identified in various fungal
In the biosynthesis of acetylaszonalenin, AnaPT catalysed the
strains including Neosartorya fischeri.17,18
reverse prenylation of (R)-benzodiazepinedione at position C3 of
the indole moiety in the presence of dimethylallyl diphosphate
aPhilipps-Universita¨t Marburg, Institut fu¨r Pharmazeutische Biologie,
Deutschhausstrasse 17A, D-35037, Marburg, Germany. E-mail: shuming.
li@Staff.uni-Marburg.de; Fax: +49-6421-2825365; Tel: +49-6421-2822461
(DMAPP) resulting in formation of aszonalenin containing a
pyrrolo[2,3-b]indole moiety. The acetyltransferase AnaAT catal-
ysed the conversion of aszonalenin to acetylaszonalenin in the
presence of acetyl-CoA.17
Recently, different strategies were developed for the synthe-
sis of biologically active natural products or analogues.19–22
bPhilipps-Universita¨t Marburg, Fachbereich Chemie, Hans-Meerwein-
Strasse, 35032, Marburg, Germany
† Electronic supplementary information (ESI) available: HPLC chro-
matograms of enzyme assays. See DOI: 10.1039/b922440h
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 1133–1141 | 1133
©

View Article Online
Scheme 1 Chemoenzymatic synthesis of C3-prenylated pyrrolo[2,3-b]indoles by recombinant enzymes.
Chemoenzymatic synthesis by using purified enzymes can be
recognised as a potential method for this purpose. Based on
this strategy, four aszonalenin stereoisomers were synthesised
successfully by using two indole prenyltransferases AnaPT and
CdpNPT. The conversion rates of the one-step reactions reached
85–100%. By using HPLC, only one stereoisomer each was
detected from the incubation mixtures of different enzyme and
substrate combinations. The enzymatic products were isolated and
their structures were elucidated by NMR, MS and CD analyses.23
These results prompted us to test the possibility for synthesis of
other C3-prenylated pyrrolo[2,3-b]indoles by using AnaPT and
AnaAT as catalysts and dipeptides as substrates (Scheme 1).
1f were relative low. This indicated however that AnaPT can be in
principle used for synthesis of diverse C3-prenylated indolines. For
a biotechnological application, however, further experiments such
as mutagenesis should be carried out to create enzyme derivatives
with higher catalytic efficiency.
In a second series, the reaction mixtures of the six substrates
with AnaPT were further incubated with AnaAT in the presence of
acetyl-CoA. In the HPLC chromatograms of the tandem reaction
mixtures, additional detectable products were only observed
with cyclo-L-Trp-L-Leu (1a) (Fig. 2B) and cyclo-L-Trp-L-Phe (1c)
(Fig. 2F). In the tandem assay with 1a, the prenylated product
was almost completely converted to the peak 3a. In the case of 1c,
about two-third of the prenylated product was converted. The
low concentration of the prenylation products of other cyclic
dipeptides was likely responsible for their non-acceptance by
AnaAT.
Results and discussion
For this purpose, six tryptophan-containing cyclic dipeptides at
a final concentration of 1 mM were tested for acceptance by
AnaPT in the presence of DMAPP (1 mM). After incubation with
0.54 mM AnaPT for 24 h, the reaction mixtures were analysed on
HPLC. As shown in Fig. 2, all of the six substrates were accepted
by AnaPT with cyclo-L-Trp-L-Leu (1a) as the best substrate under
these conditions (Fig. 2A). One product peak each was detected
for cyclo-L-Trp-L-Leu (1a), cyclo-L-Trp-L-Phe (1c), cyclo-L-Trp-L-
Tyr (1d) and cyclo-L-Trp-Gly (1f) (Fig. 2A, 2E, 2G, and 2K). In the
chromatograms of the incubation mixtures of cyclo-L-Trp-L-Trp
(1b) (Fig. 2C) and cyclo-D-Trp-L-Tyr (1e) (Fig. 2I), two product
peaks were observed. Because the unknown extinction coefficients
of the enzymatic products and low amounts of the isolated
compounds, the conversion rates of the AnaPT reaction could
not be determined with acceptable accuracy. To determine the
dependence of product formation on incubation times, we carried
out incubations with all of the six mentioned cyclic dipeptides for
0, 1, 4 and 10 h and analysed the reaction mixtures with HPLC
(see Figure S1, ESI†). Increasing of product formation was clearly
observed for all of the substrates, although the yields for 1e and
For structural elucidation, the products of AnaPT were isolated
on a preparative scale and substances in the range of 0.2–1 mg
were subjected to HR-ESI-MS and NMR analyses. HR-ESI-MS
(Table 1) confirmed the presence of one dimethylallyl moiety in
the structures of 2a–2f by detection of [M+1]+ or [M+Na]+ ions,
which are 68 daltons larger than those of the respective substrates.
Comparison of the 1H-NMR data of the isolated products 2a–2f
(Table 2) with those of the respective substrates (data not shown)
revealed clearly the presence of signals for a reverse dimethylallyl
moiety in the spectra of the isolated compounds at 5.03–5.16
(d or dd) for H-1¢, 5.90–5.96 (dd) for H-2¢, 0.95–0.99 ppm (s)
for H-4¢ and 1.05–1.14 (s) for H-5¢, respectively. The signals of
H-2 were strongly upfield shifted from approximately 7.0–7.2 in
the spectra of the substrates to 5.4 ppm in the spectra of the
enzymatic products. These shifts were similar to those observed
with aszonalenins.23 This indicated that the prenylation has very
likely taken place at position C3 of the cyclic dipeptides and
pyrrolo[2,3-b]indole derivatives were formed during the enzymatic
reactions.
1134 | Org. Biomol. Chem., 2010, 8, 1133–1141
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Fig. 2 HPLC chromatograms of incubation mixtures of six tryptophan-containing cyclic dipeptides with recombinant AnaPT or AnaPT and
AnaAT. Detection was carried out with a photo diode array detector and illustrated for absorption at 254 nm.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 1133–1141 | 1135
©

View Article Online
Table 1 HR-ESI-MS data of the enzymatic products
The absolute configuration of fructigenine A was determined
by comparison of the CD spectra with those of related known
compounds.12 Taken together, both verrucofortine11 and 3a
are N1-acetylated and C3-prenylated pyrrolo[2,3-b]indole from
HR-ESI-MS data
Deviation
Measured (ppm)
Comp. Chemical Formula Calculated
L-tryptophan and L-leucine. Both 3c and fructigenine
A
are N1-acetylated and C3-prenylated pyrrolo[2,3-b]indole from
L-tryptophan and L-phenylalanine.
2a
2b
2c
2d
2e
2f
3a
3c
4
C22H29N3O2
C27H28N4O2
C25H27N3O2
C25H27N3O3
C25H27N3O3
C18H21N3O2
C24H31N3O3
C27H29N3O3
C32H36N4O2
C25H27N3O3
368.2338 [M+1]+
441.2291 [M+1]+
402.2182 [M+1]+
418.2131 [M+1]+
368.2370 8.7
441.2234 12.9
402.2209 6.7
418.2145 3.3
To determine the stereochemistry of 3a and 3c at C2 and C3, we
compared the 1H-NMR data of 3a with those of verrucofortine11
and 3c with those of fructigenine A,12 all taken in CDCl3. It
turned out that the NMR data of 3a and 3c differed clearly from
those of verrucofortine and fructigenine A, respectively, indicating
that 3a and verrucofortine as well as 3c and fructigenine A are
diastereomers to each other by differences of stereochemistry at
positions C2 and C3. One of the remarkable differences of the
NMR data is the chemical shift for H2, which was reported to
be at 6.06 and 6.04 ppm for verrucofortine and fructigenine A,
respectively. The chemical shifts of H2 of 3a and 3c were strongly
upfield shifted to 5.76 and 5.26 ppm, respectively, being similar to
the observed values for other prenylated products reported in this
study. This means that both 3a and 3c have very likely a (2R,3S)-
configuration, which is in accordance with the results obtained
from the prenylation of (R)-and (S)-benzodiazepinedinone by
AnaPT.23 Because stereochemical retention was expected during
an acetyl transfer reaction, the configuration of 2a and 2c at
positions C2 and C3 must be the same as those of 3a and 3c,
i.e. 2R and 3S.
440.1950 [M+Na]+ 440.1962 2.7
312.1712 [M+1]+
410.2444 [M+1]+
444.2287 [M+1]+
509.2917 [M+1]+
418.2131 [M+1]+
312.1732 6.4
410.2399 11.0
444.2309 4.9
509.2932 2.9
418.2118 3.1
5
The assignments listed in Table 2 were verified by analysing
the 1H-13C heteronuclear single-quantum correlation (HSQC) and
heteronuclear multiple bond correlation (HMBC) spectra. Cross
peaks in the HSQC spectra of 2a–2f revealed that singlets of H-2 at
5.38–5.43 ppm correlated with the signals of C-2 at 77.5–79.5 ppm,
indicating the disappearance of the double bond between C-2
and C-3 of the indole ring in the structures of 2a–2f. HMBC
spectra showed clear connectivity from H-2 to C-11 and C-13.
These connectivities are conducted by the formation of a chemical
bond between C-2 and N-12 of the diketopiperazine ring. These
results proved that the structures of 2a–2f are indeed C3-prenylated
indolines with a fused five-membered ring between the indoline
and diketopiperazine ring. This means that AnaPT catalysed, in
analogy to its natural substrate (R)-benzodiazepinedinone, the
reverse C3-prenylation of tryptophan-containing cyclic dipeptides
and meanwhile the formation of a pyrrolo[2,3-b]indole structure.
In analogy to the prenylated products, the acetylated products
of the tandem reactions of cyclo-L-Trp-L-Leu (1a) and cyclo-L-
Trp-L-Phe (1c) with AnaPT and AnaAT were also isolated on
a preparative scale and subjected to HR-ESI-MS (Table 1) and
NMR analyses. HR-ESI-MS results confirmed the presence of
one acetyl moiety in 3a and 3c by detection of [M+1]+ ions, which
are 42 daltons larger than those of the respective substrates, i.e.
the enzymatic products 2a and 2c of AnaPT. A singlet for three
CD spectra of 2a–2f (Fig. 3) provided further evidence for their
stereochemistry at C2 and C3. Fig. 3 showed clearly that all of
these compounds have a positive Cotton effect at 296 nm. A
positive Cotton effect at 296 nm was detected in the CD spectra
of the two aszonalenins with a configuration of (2R,3S) and a
negative Cotton effect at this wavelength in other two aszonalenin
stereoisomers with a configuration of (2S,3R). Almost no signal
was found for the both precursors of the four aszonalenins without
1
protons at 2.53 ppm in the H spectrum of 3a and 2.56 ppm in
that of 3c were observed, respectively (Table 3). These data are
in the range of signals for the N1-acetyl groups of C3-prenylated
pyrrolo[2,3-b]indoles, e.g. 2.63 ppm for NCOCH3 of fructigenine
A12 and verrucofortine.11 Spatial correlations of H-2 and H-7 with
the methyl protons of the acetyl group were clearly observed in a
ROESY spectrum of 3a (spectrum not shown). This proved that
the acetyl moiety was attached to position N1 of the indoline ring
and corresponded very well to the N1 acetylation of aszonalenin
by AnaAT, the natural substrate of this enzyme.17
Until now, three C3-monoprenylated pyrrolo[2,3-b]indoles de-
rived from the cyclic dipeptides tested in this study were isolated
from various fungal cultures and identified as N1-acetylated and
non-acetylated derivatives. These compounds include verrucofor-
tine (Fig. 1),11 termed also fructigenine B,12 which is an N1-
acetylated derivative of L-tryptophan and L-leucine. Rugulosuvine
A and its N1-acetylated derivative rugulosuvine B, termed also
fructigenine A,12 (Fig. 1) are derived from L-tryptophan and L-
phenylalanine.13 All of these compounds have a configuration
of 2S and 3R, i.e. a cis-configuration between C2 and C3.
Fig. 3 CD spectra of C3-prenylated pyrrolo[2,3-b]indoles.
1136 | Org. Biomol. Chem., 2010, 8, 1133–1141
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 1133–1141 | 1137
©

View Article Online
Table 3 1H-NMR and 13C-NMR data of enzymatic products
Compound
Position
dC
dH, multi., J in Hz in CDCl3 dH, multi., J in Hz in CDCl3 dH, multi., J in Hz in CDCl3
dH, multi., J in Hz in DMSO-d6
1
2
3
4
5
6
7
8
9
—
80.8
60.8
—
5.76,br s
—
—
5.26, s
—
7.24, d, 7.1
7.11, td, 7.5, 1.0
7.22, t, 7.6
7.08, d, 7.0
—
7.52, s
5.27, s
—
7.12, d, 7.3
6.70, td, 7.5, 1.0
7.01, td, 7.8, 1.2
6.30, d, 7.8
—
10.65, s
—
—
7.40, d, 8.1
6.91, td, 10.3, 2.4
7.18, t, 7.9
7.06, d, 8.1
—
125.5 7.30, d, 7.5
124.0 7.10, td, 7.6, 1.1
128.6 7.25, td, 7.6, 1.3
118.6 7.91, br s
141.1
132.9
30.1
—
—
—
—
—
a
10syn
10anti
2.93, m
2.66, dd, 13.9, 10.3
4.16, dd, 10.2, 4.5
—
3.85, dd, 9.2, 3.9
5.45, s
—
3.46, dd, 13.5, 4.9
2.65, m
4.12, dd, 11.6, 3.3
—
4.09, dd, 10.6, 3.9
—
—
2.67, dd, 10.2, 5.8
2.63, dd, 10.2, 6.6
—
7.23, d, 7.6
7.28, t, 7.6
7.25, m
7.28, t, 7.6
7.23, d, 7.6
—
2.56, br s
5.15, d, 10.8
5.12, d, 17.4
5.83, dd, 17.4, 10.8
—
2.71, dd, 13.8, 9.3
2.71, dd, 13.8, 9.3
4.08, br t, 8.8
—
4.08, br t, 8.8
5.57, s
3.08, dd, 14.3, 4.4
2.87, m
4.42, dd, 11.2, 3.6
—
4.30, dd, 10.6, 5.6
5.45, s
a
11
13
14
15
16
17
57.6
168.0
53.8
—
168.9
37.7
—
24.3
22.8
21.0
169.9
23.7
—
—
—
1.89, ddd, 14.0, 9.6, 3.9
1.42, ddd, 14.5, 9.2, 5.3
2.55, dd, 13.8, 7.3
2.55, dd, 13.8, 7.3
—
5.27, s
7.52, s
2.84, m
2.59, m
—
6.60, t, 8.5
6.87, t, 8.5
9.17, s
6.87, t, 8.5
6.60, t, 8.5
—
—
3.27, m
—
5.32, t, 6.1
—
1.69, s
18
19
20
21
22
23
24
25
1.56, m
0.88, d, 6.6
0.86, d, 6.6
—
2.53, s
—
—
6.30, d, 7.8
7.01, td, 7.8, 1.2
6.70, td, 7.5, 1.0
7.12, d, 7.3
5.12, dd, 10.8, 0.9
5.08, dd, 17.4, 1.1
5.92, dd, 17.4, 10.8
—
—
—
—
—
1¢ (1¢¢)
114.5 5.14, d, 10.8
5.11, d, 17.3
143.3 5.83, dd, 17.3, 10.8
—
2¢(2¢¢)
3¢(3¢¢)
4¢(4¢¢)
5¢(5¢¢)
41.4
22.2
22.2
—
0.92, s
1.14, s
0.93, s
1.14, s
0.99, s
1.14, s
1.70, s
a not determined for 5.
chiral atoms at these positions.23 Therefore, the Cotton effect at
296 nm is very likely caused by the chiral centres at positions
C2 and C3. This proved that the prenylated products described
in this study should have the same configuration at C2 and C3
as those of (2R,3S,11R)- and (2R,3S,11S)-aszonalenin as well
as of 3a and 3c, i.e. 2R and 3S. This corresponded also to the
observation that AnaPT introduced a a-dimethylallyl group and
catalysed the formation of an fused ring between the indoline and
the diketopiperazine rings.23
Unambiguous proof of the stereochemistry was provided by
ROESY experiments for 2a–2f and 3a (Table 4). Strong NOE cor-
relations between H-2 and H-2¢, H-4¢and H-5¢of the dimethylallyl
moiety proved the cis-configuration between C-2 and C3 of the
indoline rings.
Table 4 NOE results of C3-prenylated pyrrolo[2,3-b]indoles (with excep-
tion for 3c and 4)
Protons
Strength
H-2 to H-1¢
Weak
H-2 to H-2¢
Strong
Strong
Strong
Mediuma
Weak
H-2 to H-4¢
H-2 to H-5¢
H-2 to H-11
H-10syn to H-1¢
H-10syn to H-2¢
H-10syn to H-4¢
H-10syn to H-5¢
H-10syn to H-11
H-10anti to H-4
H-11 to H-2¢
H-11 to H-4¢
H-11 to H-5¢
Medium
Strong
Strong
Strongb
Strong
Very weakc
Weakc
With an exception for 2e, NOE correlation with a medium
intensity has also been observed between H-2 and H-11. In the
case of 2e, only weak NOE was detected for H-2 and H-11. More
Weakc
importantly, strong NOEs were found between H-11 and H-10syn
,
a Only weak correlation was observed in 2e. b Not observed in 2e. For
2e, a correlation between H-10anti and H-11 was observed instead. c Not
observed in 2e.
which was not observed in 2e. The same proton (H-10syn) showed
medium or strong NOE correlations with H-2¢, H-4¢and H-5¢of
the dimethylallyl moiety. For 2e, a correlation between H-11 and
1138 | Org. Biomol. Chem., 2010, 8, 1133–1141
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Scheme 2 Possible mechanism of AnaPT reaction.
H-10anti was detected instead. These results proved that, with an
exception of 2e, the dimethylallyl moiety was attached from the
same side of H-11. This means that H-2, C3-prenyl and H-11 are
on the same side of the ring systems as illustrated in Scheme 1 and
Tables 2 and 3. In the case of 2e, H-11 is located at the opposite
side of H-2 and C3-dimethylallyl moiety. This conclusion was
also supported by the weak NOEs between H-11 and H-2¢, H-4¢or
H-5¢of the prenyl moiety in 2a–2d, 2f and 3a. No NOE correlations
for these protons were detected in 2e.
of a diprenylated derivative with a molecular mass 136 daltons
larger than that of the substrate. The H-NMR spectrum of 4
1
showed signals of two identical tryptophanyl moieties (Table 3),
which are reversely prenylated at position C3 of the indoline rings.
C3-diprenylated derivatives of cyclo-L-Trp-L-Trp containing two
pyrrolo[2,3-b]indole systems have been reported, e.g. amauromine
(Fig. 1) from Amauroascus sp14 and epiamauromine (Fig. 1) from
Aspergillus ochraceus.15 However, the 1H-NMR data of 4 differed
clearly from those reported for amauromine14 or epiamauromine.15
Thus, 4 has different configurations at C2 and C3 of the indoline
ring as those in amauromine or epiamauromine. The absolute
configuration of 4 was not determined in this study. However,
it can be expected that the symmetrical 4 was formed by a
second prenylation and cyclisation of 2b under the catalysis of
AnaPT. Therefore, both tryptophanyl moieties must have the
same configuration as in 2b, i.e. 2R and 3S as well as 19R and
18S (Scheme 1 and Table 3). Identification of 4 as a diprenylated
product of AnaPT indicated that the two prenyl moieties of some
natural products derived from cyclo-Trp-Trp, e.g. amauromine and
epiamauromine mentioned above, could also be introduced by one
prenyltransferase.
From the reaction mixture of 1e (Fig. 2I), the second product
peak 5 was also isolated and subjected to 1H-NMR analysis. 1H-
NMR indicated the presence of a regular prenyl moiety with
signals at 3.27 (m) for H-1¢, 5.32 (t) for H-2¢, 1.70 (s) for H-5¢ and
1.69 ppm (s) for H-4¢, respectively (Table 3). The signal of proton
H-2 disappeared in comparison to that of the substrate (data not
shown). Therefore, this compound was identified as a regular C2-
prenylated cyclo-D-Trp-L-Tyr (Table 3). Assignments of NMR
signals (Table 3) for this compound were also confirmed by H–
H-COSY (data not shown). The structure of 5 was confirmed by
detection of the [M+1]+ ion at m/z 418.2118 in HR-ESI-MS.
These results and those from our previous study17 showed clearly
that the reaction catalysed by AnaPT includes at least three steps,
i.e. attachment of a reverse prenyl moiety to C3 of the indole
ring, breaking of the double bond between C2 and C3 and the
formation of a C–N bond between C2 and N12. The detailed
mechanism of AnaPT reaction is unknown. We speculate that
generating of a dimethylallyl cation would initiate the enzymatic
reaction (Scheme 2), as in the case of other prenyltransferases, e.g.
FgaPT2.24 Attacking of this intermediate via its C3 to C3 of the
indole ring of the cyclic dipeptides would result in formation of
the intermediate 6 with a positive charge at C2. Two possible fates
could be expected for 6. Formation of a C–N bond between C2 and
N12 leads to 7, which is then converted to the enzymatic products
2 by releasing of a proton. The second possibility would be the
formation of a double bond between N1 and C2 results in cation 8,
which would be then converted to 9 by releasing of a proton. Using
cyclo-L-Trp-L-Leu (1a) and DMAPP as substrates and AnaPT as
catalyst, we obtained indeed a product, which could be interpreted
1
by H-NMR analysis as a structure like 9. However, the NMR
data were insufficient for an unequivocal structure elucidation.
Furthermore, the 1H-NMR spectrum of the obtained compound
in CDCl3 was changed to that of 2a, when the sample was analyzed
again after storage at room temperature for four days. We have
tried to isolate enough substance for an unambiguous structure
elucidation by spectroscopic methods including 13C-NMR and
HMBC. Chemical shift of C2 in 9 should clearly differ from
that of 2. Unfortunately, the initial observed results could not
be reproduced. Instead, 2a was isolated directly from the reaction
mixture. Plausible explanation for this phenomenon would be the
low stability of 9, which was rapidly converted to 2a during the
incubation or isolation. We are now looking for suitable conditions
to isolate this compound for evaluation of the reaction mechanism.
The second product peak 4 of cyclo-L-Trp-L-Trp (1b) (Fig. 2C)
was also isolated after repeated chromatography and subjected
to MS analysis. The positive HR-ESI-MS of 4 showed an ion
at m/z 509.2932 (Table 1), which can be interpreted as [M+1]+
Conclusions
In this study, we reported the successful synthesis of nine
reversely C3-prenylated pyrrolo[2,3-b]indoles by chemoenzymatic
strategy using recombinant and purified enzymes of secondary
metabolism. The structures of these compounds were char-
acterized by HR-ESI-MS and NMR analyses. NOE results
(Table 4) confirmed the absolute configuration of the enzymatic
products. Similar structures have been identified in different fungal
strains.17,25 The approach described in this study could therefore
also find usage for synthesis of natural products. Conversion
of cyclo-L-Trp-L-Trp to the diprenylated derivative 4 provided
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 1133–1141 | 1139
©

View Article Online
evidence that only one prenyltransferase is necessary for intro-
ducing of two prenyl moieties. It would be interesting to find
out the substrate binding sites of the enzyme and to clarify the
two prenylation steps. Although the efficiency of the enzymatic
conversion is still low for a biotechnological application, such
approaches offer however the potential to significantly increase
the structural diversity in drug discovery programmes. The low
catalytic efficiency could be improved by using enzyme derivatives
from mutagenesis experiments.
was used. The column was then washed with 100% solvent B for
5 min and equilibrated with 10% (v/v) solvent B for 5 min.
Circular dichroism spectra
The samples were dissolved in ethanol. CD spectra were recorded
on a JASCO J-810 CD spectropolarimeter by using Spectra
Manager Software (JASCO Inc). Spectral data were collected from
400 to 200 nm with a data pitch of 0.1 nm. A band width of 1 nm
was used with a detector response time of 0.5 s and scanning
speed of 200 nm min-1. Each spectrum or data point was acquired
5 times, and mean values were used.
Experimental section
Chemicals
Enzymatic assays and syntheses of reversely C3-prenylated
pyrrolo[2,3-b]indoles
DMAPP was prepared according to the method described for ger-
anyl diphosphate by Woodside.26 Acetylcoenzyme A (Trilithium
Salt) was obtained from Calbiochem (Darmstadt, Germany).
Cyclo-L-Trp-L-Leu, cyclo-L-Trp-L-Trp, cyclo-L-Trp-L-Phe, cyclo-
L-Trp-L-Tyr, cyclo-D-Trp-L-Tyr and cyclo-L-Trp-Gly were ob-
tained from Bachem (Bubendorf, Switzerland).
Enzyme assays (100 mL) contained substrates 1a to 1f (1 mM),
DMAPP (1 mM), CaCl2 (5 mM), Tris-HCl (50 mM, pH 7.5),
glycerol 1.5% (v/v) and AnaPT (0.54 mM) were incubated at
37 ◦C. For tandem reaction, the reaction mixtures of AnaPT were
incubated with acetylcoenzyme A (1 mM) and AnaAT (0.88 mM)
at 37 ◦C for further 24 h. The reaction was then terminated with
the same volume of methanol. After removal of the protein by
centrifugation (15,000 x g, 10 min, 4 ◦C), the enzymatic products
were analysed on a HPLC system described above. For synthesis
of C3-prenylated pyrrolo[2,3-b]indoles, 1a to 1f (1 mM), DMAPP
(2 mM), CaCl2 (5 mM), Tris-HCl (50 mM, pH 7.5), glycerol
1.5% (v/v) and AnaPT (0.54 mM) in total volumes of 10 mL
were incubated at 37 ◦C for 24 h. For synthesis of acetylated
indole alkaloids, the reaction mixture of AnaPT with 1a or 1c was
incubated with acetylcoenzyme A (2 mM) and AnaAT (0.88 mM)
at 37 ◦C for further 24 h. The reaction mixtures were extracted
with ethyl acetate subsequently. After evaporation of the solvent,
the residues were dissolved in methanol and purified on HPLC
under conditions described above.
Bacterial strains, plasmids and cultural conditions
Plasmids pWY22 and pWY2317 were used for overproduction of
AnaPT and AnaAT, respectively.
Escherichia coli XL1 Blue MRF¢ (Stratagene) was used for
overexpression experiments and grown in liquid or on solid Luria-
Bertani medium with 1.5% (w/v) agar at 37 ◦C. Carbenicillin
(50 mg mL-1) was used for selection of recombinant E. coli strains.
Preparation of AnaPT and AnaAT
For preparation of AnaPT, E. coli XL1 Blue MRF¢ cells harbour-
ing pWY22 were induced by 0.5 mM of IPTG at 37 ◦C. His6-
AnaPT was purified with Ni-NTA agarose to homogeneity as
judged by SDS-PAGE and a protein yield of 25 mg of purified
His6-tagged AnaPT per litre of culture was obtained.17
NMR experiments
For preparation of AnaAT, E. coli XL1 Blue MRF¢ cells har-
bouring pWY23 were induced by 0.7 mM of IPTG at 37 ◦C. His6-
AnaAT was purified with Ni-NTA agarose to homogeneity as
judged by SDS-PAGE and a protein yield of 5 mg of purified
His6-tagged AnaAT per litre of culture was obtained.27
Small amount (less than 1 mg) of each sample was dissolved in
0.2 mL of CDCl3. Samples were filled into Wilmad 3 mm tubes
from Rototec Spintec. Spectra were recorded at room temperature
on a Bruker Avance 600 MHz spectrometer equipped with an
inverse probe with z-gradient. The HSQC and HMBC spectra
were recorded with standard methods.28 ROESY experiment29 was
performed in phase-sensitive mode using State-TPPI technique.30
For all two-dimensional spectra, 32 to 64 transients were used. For
ROESY spectra, a mixing time of 300 ms and a relaxation delay of
3.0 s. 1H spectra were acquired with 65 536 data points, while 2D
spectra were collected using 4096 points in the F2 dimension and
512 increments in the F1 dimension. Typical experiment time for
the HMBC and ROESY measurements was about 12 h. Chemical
shifts were referenced to CDCl3. All spectra were processed with
Bruker TOPSPIN 2.1.
HPLC conditions for analysis and isolation of enzymatic products
The enzymatic products of the incubation mixtures of AnaPT
and tandem incubation of AnaPT and AnaAT were analysed by
HPLC on an Agilent series 1200 by using a LiChrospher RP 18-5
column (125 ¥ 4 mm, 5 mm, Agilent) at a flow rate of 1 mL min-1.
50% methanol in water (solvent A) and pure methanol (solvent
B) were used as solvents. For quantitative analysis of enzymatic
products, a linear gradient of 10–50% (v/v) solvent B in solvent
A in 15 min was used. The column was then washed with 100%
solvent B for 5 min and equilibrated with 10% (v/v) solvent B for
5 min. Detection was carried out by a Photo Diode Array detector.
For isolation, the same HPLC equipment with a Multospher
120 RP-18 column (250 ¥ 10 mm, 5 mm, C+S Chromatographie
Service, Langenfeld, Germany) was used. A linear gradient of 10–
50% (v/v) solvent B in A in 15 min at a flow rate of 2.5 mL min-1
HR-ESI-MS
The isolated products were analysed by high resolution (HR)
electrospray ionization (ESI) mass spectrometry (MS) with a
Q-Trap Quantum (Applied Biosystems). Positive HR-ESI-MS
data of the enzymatic products are in Table 1.
1140 | Org. Biomol. Chem., 2010, 8, 1133–1141
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
13 A. G. Kozlovsky, V. M. Adanin, H. M. Dahse and U. Gra¨fe, Appl.
Biochem. Microbiol., 2001, 37, 253–256.
Acknowledgements
14 S. Takase, Y. Kawai, I. Uchida, H. Tanaka and H. Aoki, Tetrahedron
Lett., 1984, 25, 4673–4676.
15 F. S. De Guzman and J. B. Glober, J. Nat. Prod., 1992, 55, 931–939.
16 G. A. Ellestad, P. Mirando and M. P. Kunstmann, J. Org. Chem., 1973,
38, 4204–4205.
17 W.-B. Yin, A. Grundmann, J. Cheng and S.-M. Li, J. Biol. Chem., 2008,
284, 100–109.
This work was supported by a grant from the Deutsche
Forschungsgemeinschaft (S.-M. Li). We thank Li Feng for his
help to take the CD spectra. Xie acknowledges the Deutsche
Forschungsgemeinschaft for funding the Bruker AVANCE 600
spectrometer.
18 D. Wakana, T. Hosoe, T. Itabashi, K. Nozawa, K. Okada, G. M. d. C.
Takaki, T. Yaguchi, K. Fukushima and K. I. Kawai, Mycotoxins, 2006,
56, 3–6.
References
19 F. Felluga, W. Baratta, L. Fanfoni, G. Pitacco, P. Rigo and F. Benedetti,
J. Org. Chem., 2009, 74, 3547–3550.
20 L. K. Thale´n, D. B. Zhao, J. B. Sortais, J. Paetzold, C. Hoben and J. E.
Ba¨ckvall, Chem.–Eur. J., 2009, 15, 3403–3410.
21 T. Ueda, K. Tomita, Y. Notsu, T. Ito, M. Fumoto, T. Takakura, H.
Nagatome, A. Takimoto, S. I. Mihara, H. Togame, K. Kawamoto, T.
Iwasaki, K. Asakura, T. Oshima, K. Hanasaki, S. I. Nishimura and H.
Kondo, J. Am. Chem. Soc., 2009, 131, 6237–6245.
22 K. C. Seo, Y. G. Kwon, D. H. Kim, I. S. Jang, J. W. Cho and S. K.
Chung, Chem. Commun., 2009, 1733–1735.
1 D. J. Newman and G. M. Cragg, J. Nat. Prod., 2007, 70, 461–477.
2 R. M. Williams, E. M. Stocking and J. F. Sanz-Cervera, Top. Curr.
Chem., 2000, 209, 97–173.
3 E. M. Stocking, R. M. Williams and J. F. Sanz-Cervera, J. Am. Chem.
Soc., 2000, 122, 9089–9098.
4 B. Botta, A. Vitali, P. Menendez, D. Misiti and M. G. Delle, Curr. Med.
Chem., 2005, 12, 717–739.
5 T. Usui, M. Kondoh, C. B. Cui, T. Mayumi and H. Osada, Biochem. J.,
1998, 333, 543–548.
6 P. M. Scott and B. P. C. Kennedy, J. Agric. Food Chem., 1976, 24,
865–868.
7 T. Rundberget, I. Skaar and A. Flaoyen, Int. J. Food Microbiol., 2004,
90, 181–188.
8 C. Finoli, A. Vecchio, A. Galli and I. Dragoni, J. Food Prot., 2001, 64,
246–251.
9 M. Kusano, G. Sotoma, H. Koshino, J. Uzawa, M. Chijimatsu, S.
Fujioka, T. Kawano and Y. Kimura, J. Chem. Soc., Perkin Trans. 1,
1998, 2823–2826.
10 K. Sprogøe, S. Manniche, T. O. Larsen and C. Christophersen,
Tetrahedron, 2005, 61, 8718–8721.
23 W.-B. Yin, J. Cheng and S.-M. Li, Org. Biomol. Chem., 2009, 7, 2202–
2207.
24 L. Y. Luk and M. E. Tanner, J. Am. Chem. Soc., 2009, 131, 13932–
13933.
25 S.-M. Li, Phytochemistry, 2009, 70, 1746–1757.
26 A. B. Woodside, Z. Huang and C. D. Poulter, Org. Synth., 1988, 66,
211–215.
27 W.-B. Yin, H.-L. Ruan, L. Westrich, A. Grundmann and S.-M. Li,
ChemBioChem, 2007, 8, 1154–1161.
28 S. Berger and S. Braun, 200 and More NMR Experiments. A Practical
Course, Weiley-VCH, Weinheim, Germany, 2004.
29 T.-L. Hwang and A. J. Shaka, J. Am. Chem. Soc., 1992, 114, 3157–3159.
30 D. Marion, M. Ikura, R. Tschudin and A. Bax, J. Magn. Reson., 1989,
85, 393–399.
11 R. P. Hodge, C. M. Harris and T. M. Harris, J. Nat. Prod., 1988, 51,
66–73.
12 K. Arai, K. Kimura, T. Mushiroda and Y. Yamamoto, Chem. Pharm.
Bull., 1989, 37, 2937–2939.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 1133–1141 | 1141
©