Breaking the regioselectivity of indole prenyltransferases: Identification of regular C3-prenylated hexahydropyrrolo[2,3-b]indoles as side products of the regular C2-prenyltransferase FtmPT1

Article abstract of DOI:10.1039/c2ob26149a

The prenyltransferase FtmPT1 from Aspergillus fumigatus is involved in the biosynthesis of fumitremorgin-type alkaloids and catalysed the regular C2-prenylation of brevianamide F (cyclo-l-Trp-l-Pro). It has been shown that FtmPT1 also accepted a number of

Full text of DOI:10.1039/c2ob26149a

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Organic &
Biomolecular
Chemistry
Cite this: Org. Biomol. Chem., 2012, 10, 9262
www.rsc.org/obc
PAPER
Breaking the regioselectivity of indole prenyltransferases: identification of
regular C3-prenylated hexahydropyrrolo[2,3-b]indoles as side products of the
regular C2-prenyltransferase FtmPT1†
Beate Wollinsky,a Lena Ludwig,a Xiulan Xieb and Shu-Ming Li*a
Received 15th June 2012, Accepted 1st October 2012
DOI: 10.1039/c2ob26149a
The prenyltransferase FtmPT1 from Aspergillus fumigatus is involved in the biosynthesis of
fumitremorgin-type alkaloids and catalysed the regular C2-prenylation of brevianamide F (cyclo-
L-Trp-L-Pro). It has been shown that FtmPT1 also accepted a number of other tryptophan-containing
cyclic dipeptides and prenylated them, in the presence of dimethylallyl diphosphate, at C-2 of the indole
nucleus. Detailed analysis of the incubation mixtures of FtmPT1 with these cyclic dipeptides revealed
the presence of additional product peaks in the HPLC chromatograms. Seven regularly C3-prenylated
hexahydropyrrolo[2,3-b]indoles were isolated and identified by HR-ESI-MS and NMR analyses including
HMBC, HMQC and NOESY experiments. Further experiments proved that the C2- and C3-prenylated
products are both independent enzyme products. To the best of our knowledge, this is the first report on
the enzymatic formation of regularly C3-prenylated indolines. A reaction mechanism for both C2- and
C3-prenylated derivatives was proposed.
at N1, C-2 and C-3 of the indole ring.1,5 For C2-prenylation,
Introduction
both regular (FtmPT1) and reverse prenyltransferases (NotF and
BrePT) have been identified.5,7,8 Two C3-prenyltransferases
Prenylated indole alkaloids are a large family of secondary
metabolites with diverse biological activities.1,2 These sub-
AnaPT and CdpC3PT from Neosartorya fischeri were found to
stances are hybrid natural products containing prenyl moieties
derived from prenyl diphosphates and an indole or indoline ring
from tryptophan or its precursors.1 Prenyltransferases, especially
those from the DMATS superfamily, are responsible for connec-
tion of these two structural elements.1,3 Bioinformatic analysis
indicated the presence of at least 200 members from this family
in the database.4 More than 20 of such enzymes have been
characterized biochemically.1,3,5
These enzymes usually transfer a prenyl moiety regiospecifi-
cally to one of the seven positions of the indole nucleus. The
identified enzymes, which use tryptophan as a natural or the best
aromatic substrate and therefore act as dimethylallyltryptophan
synthases, e.g. FgaPT2, 5-DMATS, IptA and 7-DMATS,
catalysed regular prenylation at C-4 to C-7.1,3,6 In contrast,
indole prenyltransferases, which use tryptophan-containing
cyclic dipeptides as aromatic substrates, catalysed the prenylation
catalyse regiospecific prenylations at C-3, resulting in the for-
mation of hexahydropyrrolo[2,3-b]indoles carrying an α- and a
β-configured prenyl moiety, respectively.9–11 In both cases, the
prenyl moieties are attached via its C-3 as reverse ones. It has
also been demonstrated that CdpNPT from Aspergillus fumigatus
was able to convert cyclic dipeptides to reversely C3-prenylated
derivatives.12,13
Until now, no reports on enzymes were found in the literature,
which catalyse a regular prenylation at C-3 of the indole ring.
This could be explained by the fact that regularly C3-prenylated
derivatives of cyclic dipeptides occur relatively rare in the nature
and examples, e.g. nocardioazine B (Fig. 1), have only been
recently isolated and identified.14 In contrast, a number of rever-
sely C3-prenylated derivatives including aszonalenin and roque-
fortine C (Fig. 1) have been identified as mycotoxins for
decades.1
The aforementioned prenyltransferase FtmPT1 from Aspergillus
fumigatus is involved in the biosynthesis of verruculogen and
catalysed the prenylation of cyclo-L-Trp-L-Pro (1a, Fig. 2) at C-2,
resulting in the formation of tryprostatin B (1b, Fig. 2).7,15 Try-
prostatin B was reported to exhibit inhibitory effects on the cell
cycle progression in the G2/M phase.16 It has also been shown
that the prenyl moiety at C-2 in 1b was essential for the cytotoxici-
city.17,18 In the course of our search for cytotoxic compounds,
aPhilipps-Universität Marburg, Institut für Pharmazeutische Biologie
und Biotechnologie, Deutschhausstrasse 17A, D-35037 Marburg,
Germany. E-mail: shuming.li@Staff.uni-Marburg.de;
Tel: +49-6421-2822461; Fax: +49-6421-2825365
bPhilipps-Universität Marburg, Fachbereich Chemie, Hans-Meerwein-
Strasse, 35032 Marburg, Germany
†Electronic supplementary information (ESI) available. See DOI:
10.1039/c2ob26149a
9262 | Org. Biomol. Chem., 2012, 10, 9262–9270
This journal is © The Royal Society of Chemistry 2012

View Article Online
Fig. 1 Examples of naturally occurring C3-prenylated indolines.
Fig. 2 Conversion of L-tryptophan-containing cyclic dipeptides by FtmPT1 to regular C2- and C3β-prenylated derivatives.
14 regularly C2-prenylated tryptophan-containing diketopipera-
zines were prepared by incubation of the respective cyclic dipep-
tides with FtmPT1 in the presence of dimethylallyl
diphosphate.19 In at least seven of these incubation mixtures, regu-
larly C3-prenylated indolines were identified as side products
of the enzymatic reactions. In this study, we describe the iso-
lation and structure elucidation of these products and postulate a
plausible reaction mechanism for the formation of both C2- and
C3-prenylated derivatives.
illustrated in Fig. 4. The structures of 1b–14b were unequivo-
cally identified by NMR and MS analyses as regularly C2-preny-
lated diketopiperazines.19 It showed clearly that the natural
substrate of FtmPT1, cyclo-L-Trp-L-Pro (1a), was converted to
tryprostatin B (1b) as the predominant product in this enzyme
assay, corresponding very well to the results reported pre-
viously.7 Due to the increased amount of the injected sample
under the preparative conditions, one additional minor peak
(approximately 0.7%) with
a large retention time was
detected. Inspection of other 13 chromatograms revealed the
presence of two or more product peaks. Similar to that of 1a,
C2-prenylated diketopiperazines 3b, 7b and 8b were predomi-
nant products in the incubation mixtures of cyclo-L-Trp-L-Leu
(3a) and the two cyclo-Trp-Pro isomers 7a and 8a (Fig. 2 and 3).
The second product peak (c series) in other reaction mix-
tures was significantly higher than those of 1a, 3a, 7a and 8a. In
the cases of 5a and 9a, a third product (d series) each was
detected.
Results and discussion
Detection of additional products in the incubation mixtures of
FtmPT1
During the isolation of 1b–14b (Fig. 2 and 3) for cytotoxic
testing, the preparative HPLC chromatograms of the incubation
mixtures of 14 cyclic dipeptides (1a–14 a) were recorded and
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 9262–9270 | 9263

View Article Online
Fig. 3 Conversion of D-tryptophan-containing cyclic dipeptides by FtmPT1 to regular C2- and C3α-prenylated derivatives.
For better quantification, we carried out incubations at 37 °C in
100 μl scales with 5 μg FtmPT1 for 2 h. The yields of products b
and c are summarized in Table 1. Under these conditions, yields
of more than 10% were detected for the products of the c series in
five of the used substrates, i.e. 2a, 9a, 10a, 11a and 13a. The
product yield of 18.6% for 2c is about one-third of that of 2b. The
product yields of the c series in incubation mixtures of 9a, 10a,
11a and 13a were found to be about one-fourth of those of the b
series. In the incubation mixture of 9a, the product yield of an
additional product 9d reached 24% of that of 9b.
To determine the prenylation position, connectivities in
HMBC of 2c, 5c, 11c and 14c were inspected in detail and sum-
marized in Fig. S10 (ESI†). In these HMBC spectra, connectiv-
ities from H-1′ of the prenyl moiety at 2.21–2.46 ppm to C9 at
δc 131.5–132.4 ppm and to C-3 at δc 56.1–56.5 ppm confirmed
the attachment of the prenyl residue at position C-3. In HMBC
spectra of 2c and 5c connectivities from H-10 to C-2′ were also
observed.
Unambiguous proof of the β-configuration of the prenyl and
H-2 in 5c, 10c, 11c and 14c was provided by NOESY experiments
(Fig. S11, ESI†). In all NOESY spectra, interactions between
H-10anti and H-1′, H-1′ and H-2 as well as H-10syn and H-4 were
clearly detected (Table 4 and Fig. S11, ESI†). Strong NOE corre-
lations between H-2 and the protons H-1′ of the prenyl moiety
proved the cis-configuration between H-2 and C3-prenyl moieties
of the indoline rings. For all of the analysed compounds, strong
NOE correlations were observed between H-11 and H-10syn as well
as H-1′ and H-10anti. For 5c, a strong correlation between H-10anti
and H-2 was also observed. In contrast, no correlation was detected
for H-11 with H-1′ and H-2′. From these results, we concluded that
the regular C3-prenyl moiety in products c must be situated on the
opposite side to H-11, i.e. with a β-configuration, as summarized
FtmPT1 catalysed also regular C3-prenylation as side reaction
For structure elucidation, 12 products of the c series as well as 5d
and 9d were isolated on HPLC from the respective incubation
mixtures and subjected to spectroscopic analysis. Positive high reso-
lution electrospray ionization mass spectrometry (HR-ESI-MS)
data (Table 2) confirmed the monoprenylation in the isolated
enzyme products by detection of [M + Na]+ or [M + H]+ ions,
which are 68 daltons larger than those of the respective substrates.
UV spectra of the isolated products of the c series showed
similar characteristic maximal absorptions at 200, 240 and
290 nm to each other and clearly differed from those of the
C2-prenylated products of the b series with maxima at 220 and
280 nm (data not shown). The UV spectra of the products of the
c series were very similar to those of the C3-prenylated indolines
isolated from incubation mixtures of AnaPT or CdpC3PT,9,11
indicating the presence of hexahydropyrrolo[2,3-b]indole
systems in the structures of the products of the c series.
1
in Fig. 2. The H-NMR data of 10c corresponded well to a syn-
1
thetic product reported previously.22 The H-NMR data of 9c are
identical to its enantiomer synthesized chemically.22 The 1H-NMR
data of 13c are identical to its enantiomer 11c reported in this study
(Table 3).These results proved that hexahydropyrrolo[2,3-b]indoles
with an α-configuration were products of the c series of D-trypto-
phan-containing cyclic dipeptides, as summarized in Fig. 3.
5d was identified as cyclo-N1-dimethylallyl-L-Trp-L-Tyr by
comparison of the 1H-NMR data with those published pre-
viously.21 9d showed a nearly identical NMR spectrum as its
enantiomer cyclo-N1-dimethylallyl-L-Trp-L-Pro.21
1
In the H-NMR spectra of 2c, 5c, 9c–12c and 14c (Fig. S1.1–
S9.4, ESI†), signals for a regular dimethylallyl moiety at δH
2.21–2.46 (d or dd, 2H-1′), 4.84–5.14 (br t, H-2′), 1.49–1.62 (s,
3H-4′), 1.61–1.67 (s, 3H-5′) were clearly observed (Table 3). The
signals for the two protons of H-1′ of the prenyl moiety at
2.21–2.46 ppm indicted its attachment to a C- rather than to an
O- or N-atom.9,11,19–21 In comparison to those of the respective
substrates, the signals for H-2 of the products of the c series were
upfield shifted from about δH 7.2 to 5.18–5.41 (s, 1H), which are
in a similar range to those of reversely C3-prenylated indolines.9,11
FtmPT1 products of the b and c series are independently formed
during the enzyme reaction
As described above, regularly C2- and C3-prenylated products
were identified in the reaction mixtures of FtmPT1. FtmPT1
9264 | Org. Biomol. Chem., 2012, 10, 9262–9270
This journal is © The Royal Society of Chemistry 2012

View Article Online
Fig. 4 HPLC chromatograms of incubation mixtures of FtmPT1 with 14 cyclic dipeptides. The reaction mixtures (50 ml) containing 1 mg FtmPT1,
10 mM CaCl2, 1 mM cyclic dipeptide and 2 mM DMAPP were incubated at 37 °C for 2 h. The substances were detected with a Photo Diode Array
detector and illustrated for absorption at 296 nm.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 9262–9270 | 9265

View Article Online
Table 1 Conversion of diketopiperazines to prenylated products by the
of the five substrates with FtmPT1 and DMAPP were used as
positive controls. It has been shown that the products of the
b and c series were chemically stable and no meaningful
changes were detected in the incubation mixtures without
FtmPT1. No additional peaks were observed in the incubation
mixtures of these substances with FtmPT1, proving that no con-
version had taken place between the isolated products of the b
and c series (Fig. S13–S17, ESI†). It can however not be
excluded that conversion between b and c takes place as
enzyme-bound intermediates before their release from FtmPT1.
To compare the behavior of FtmPT1 towards different sub-
strates and to understand the two prenylation possibilities,
kinetic parameters were determined for eight substrates (Table 5,
Fig. S18†). As shown in Fig. S18,† substrate inhibition at 2 mM
was observed for the formation of 11b and 13b (Fig. S18†). In
several cases, e.g. 9a, 11a and 14a, the kinetic parameters for
products of the c series cannot be determined, because the
product formation was found still in the lineal region for sub-
strate concentrations of up to 2 mM. From the data listed in
Table 5, it is obvious that tryptophan-containing cyclic dipep-
tides with a smaller amino acid than proline, e.g. glycine in 2a
or alanine in 10a, 11a and 13a, were poor substrates for
FtmPT1. In the case of 14a, electrostatic difference between
proline and histidine seems to be responsible for their different
affinities to FtmPT1. Furthermore, it can be interpreted that sub-
strates with high affinity to FtmPT1 were mainly converted to
products of the b series. As expected, FtmPT1 showed the
(nearly) highest affinity and velocity towards its natural substrate
1a with 1b as the predominant product. The catalytic efficiency
of approximately 25 312 s−1 M−1 for 1a is much higher than
other tested substrates. The two other well accepted substrates 5a
and 9a, i.e. with high affinity to FtmPT1, were converted mainly
to products of the b series with relatively high catalytic efficien-
cies of 6407 and 2214 s−1 M−1, respectively. In contrast, the pro-
ducts of the c series were clearly detected in the incubation
mixtures of substrates with low affinity, e.g. 2a and 10a, with
KM values of 0.77 and 0.54 mM for 2b and 10b as well as 0.69
and 0.48 mM for 2c and 10c, respectively.
prenyltransferase FtmPT1a
Non-reacted
substrate (%)
Yield of
product b [%]
Yield of
product c [%]
Substrate
1a
2a
3a
4a
5a
6a
7a
8a
7.0
24.4
28.2
29.2
26.5
56.1
20.9
70.8
44.8
33.8
44.0
86.3
42.0
25.1
82.3
57.0
70.9
67.0
72.4
39.3
75.7
27.4
43.4
53.5
43.9
8.3
0.7
18.6b
0.9c
3.8c
1.1b
4.6c
3.4
1.8c
9a
11.8b
12.7b
12.1b
5.4c
10a
11a
12a
13a
14a
46.4
68.1
11.6b
6.8b
a The reaction mixtures in 100 μl scales containing 5 μg of FtmPT1,
10 mM CaCl2, 2 mM DMAPP and 1 mM diketopiperazine were
incubated for 2 h at 37 °C. The substances were detected at 296 nm.
b The products were isolated on HPLC and characterized by NMR and
MS analyses. c The products were isolated on HPLC and identified by
HR-ESI-MS analysis.
Table 2 HR-ESI-MS data of the enzyme products c and d series
HR-ESI-MS data
Chemical
Comp. formula
Deviation
Measured (ppm)
Calculated
2c
3c
4c
5c
5d
6c
8c
C18H21N3O2 334.1531 [M + Na]+ 334.1559 8.2
C22H29N3O2 390.2157 [M + Na]+ 390.2172 3.6
C27H28N4O2 463.2110 [M + Na]+ 463.2131 4.5
C
25H27N3O3 440.1951 [M + Na]+ 440.1915 7.9
C25H27N3O2 440.1950 [M + Na]+ 440.1999 1.9
C25H27N3O2 424.2001 [M + Na]+ 424.2008 1.7
C21H25N3O2 374.1844 [M + Na]+ 374.1834 2.7
9c
C
21H25N3O2 374.1844 [M + Na]+ 374.1840 1.3
9d
10c
11c
12c
13c
14c
C21H25N3O2 374.1844 [M + Na]+ 374.1844 0.1
C19H23N3O2 348.1688 [M + Na]+ 348.1672 4.7
C19H23N3O2 348.1688 [M + Na]+ 348.1667 6.0
C
19H23N3O2 348.1688 [M + Na]+ 348.1690 0.5
C19H23N3O2 348.1688 [M + Na]+ 348.1671 4.8
C22H25N5O2 392.2087 [M + H]+
392.2087 0.2
Reaction mechanisms of C2- and C3-prenylation by FtmPT1
catalysed the conversion of 1a to 1b by regular C2-prenylation,
which serves as an intermediate in the biosynthesis of fumitre-
morgin-type alkaloids.23 It would be not surprising that other
cyclic dipeptides can also be prenylated at C-2 of the indole
ring. However, a regular C3-prenylation was an unexpected
event.
To prove the relationships between the products of the b and c
series, we carried out time dependence of their formation by
using 2a, 9b and 11a as substrates. As shown in Fig. S12
(ESI†), total yields of 60–80% were reached after incubation for
4 h. It seems that the reactions have reached their equilibrium
after 4 h under these conditions. The formation of the products b
and c was found however to be independent from each other in
all of the cases.
C3-prenylation by a C2-prenyltransferase sounds somewhat sur-
prising. It can be however explained by inspection of the sub-
strate binding pocket of the FtmPT1 structure.24 It has been
described that brevianamide F (1a) is placed in the PT barrel of
FtmPT1 and forms a hydrogen bond with its N-1 nitrogen to the
side chain carboxylate of the E102 residue of FtmPT1 (Fig. 5).
The indole ring of 1a is furthermore stabilised by tyrosine 203.
As proposed previously,24–26 a dimethylallyl cation with deloca-
lisation of the positive charge between its C-1 and C-3 will be
firstly created by elimination of the pyrophosphate group with
the help of several basic amino acid residues and stabilised by
cation-π interactions with the aromatic ring system, i.e. with its
substrate 1a from one side and the tyrosine residue Y382 from
the other side (Fig. 5). Stabilisation of this cation by substrates
with a smaller residue than 1a, e.g. 2a, 10a, 11a and 13a or with
a different electrostatic density as in the case of 14a, can differ
from that of 1a, so that clear variation in kinetic parameters was
Furthermore, the isolated products b and c from 2a, 5a, 10a,
11a and 13a were incubated at 37 °C for 2 h in the presence of
DMAPP with and without FtmPT1 (Fig. S13–S17†). Incubations
9266 | Org. Biomol. Chem., 2012, 10, 9262–9270
This journal is © The Royal Society of Chemistry 2012

Table 3 1H-NMR and 13C-NMR data of the enzyme products in CD3OH (2c, 5c and 10c, 11c, 13c, 14c) or CDCl3 (9c)
cyclo-C3β-
cyclo-C3β-
cyclo-C3β-
cyclo-C3β-
cyclo-C3β-
dimethylallyl-
L-Trp-Gly (2c)
dimethylallyl-
L-Trp-L-Tyr (5c)
cyclo-C3α-
dimethylallyl-D-
Trp-D-Pro (9c)
dimethylallyl-
L-Trp-L-Ala (10c)
dimethylallyl-
L-Trp-D-Ala (11c)
cyclo-C3α-
dimethylallyl-
D-Trp-L-Ala (13c)
dimethylallyl-
L-Trp-L-His (14c)
δH, multi,
J in Hz
δH, multi,
J in Hz
δH, multi,
J in Hz
δH, multi,
J in Hz
δH, multi,
J in Hz
δH, multi,
J in Hz
δH, multi,
J in Hz
Pos.
δC
δC
δC
δC
δC
NH-1
—
5.34, s
—
7.10, d, 7.5
6.70, td, 7.5, 0.8
7.01, td, 7.6, 1.2
6.58, d, 7.6
—
—
—
—
—
—
5.31, s
—
7.10, br d, 7.5
6.71, td, 7.7, 0.9
7.01, td, 7.7, 1.2
6.59, br d, 7.8
—
—
—
—
—
—
5.27, s
—
7.09, br d, 7.4
6.70, td, 7.4, 0.9
7.01, td, 7.6, 1.2
6.57, d, 7.7
—
—
80.5
56.5
124.1
119.8
129.4
110.0
150.8
132.0
39.8
2
3
4
5
6
7
8
9
80.6 5.18, s
80.3 5.41, s
56.5 —
124.2 7.07, mb
80.4 5,37 s
—
80.4 5.38, s
56.1
124.0 7.11, dd, 7.5, 1.2
119.6 6.72, td, 7.5, 1.0
129.3 7.03, td, 7.7, 1.3
109.9 6.60, dt, 7.8, 0.6
h
56.5
—
124.3 6.97, d, 7.2b
—
124.0 7.10, d, 7.5
120.0 6.65, td, 7.4, 0.8
129.6 6.97, td, 7.3, 1.2b
110.2 6.52, d, 7.5
119.6 6.75, td, 7.6, 0.8
119.8 6.70, td, 7.5, 0.7
129.3 7.02, td, 7.5, 1.1
110.2 6.59, d, 7.6
129.3 7.07, mb
109.5 6.57, d, 7.7
151.4
131.5
41.8 2.50, m
2.50, m
h
151.0
132.4
—
—
—
—
—
—
150.8
132.0
—
—
h
—
—
2.57, dd, 12.6, 6.3
2.23, dd, 12.6, 11.3
—
10syn
10anti
11
2.58, dd, 12.5, 6.1
2.22, dd, 12.5, 11.5
4.00, m
—
4.04, dd, 17.1, 2.3
3.73, dd, 17.0, 3.3
40.3 2.26, dd, 12.1, 5.5c
1.19, t, 12.0
39.6 2.60, dd, 12.5, 6.1 40.5 2.61, dd, 12.5, 6.1 2.57, dd, 12.6, 6.1
2.20, dd, 12.3, 11.7 2.22, dd, 12.5, 11.5 2.09, dd, 12.5, 11.5
a
a
59.2 3.79, ddd, 12.0, 5.4, 1.7 58.7 4.35, t, 8.7
166.2 166.0
46.9 4.33, m 57.7 4.15, m
4.03, ddd, 11.3, 6.2, 1.9 59.6 4.02, dd, 11.4, 6.0 58.4 4.02, dd, 11.5, 6.1 4.05, ddd, 12.1, 6.1, 1.8 59.6
—
4.11, qd, 6.9, 1.8
h
13
14
—
—
—
51.9 3.90, m
168.7
—
—
4.45, td, 6.0, 1.7
—
166.4
55.3
—
53.9 3.91, qd, 7.2, 0.8
NH-15 8.01, s
—
171.3
—
7.96, s
—
3.16, dd, 13.9, 4.1
2.92, dd, 13.9, 4.4
—
—
169.6
—
—
8.07, br s
—
1.36, d, 6.9
—
h
8.22, s
—
—
170.4
19.9 1.36, d, 7.1
—
—
—
—
—
171.8
26.0
16
17
—
—
—
—
38.3 3.46, mc
3.46, mc
15.6 1.36, d, 7.1
3.19, dd, 15.5, 5.7
3.33, dd, 4.3g
—
18
19
—
—
126.7 1.88–2.20, md
1.88–2.20, md
—
—
—
—
—
—
—
—
—
—
130.5
—
7.01, d, 8.5e
132.1 1.88–2.20, md
1.88–2.20, md
—
—
20
21
22
23
1′
—
—
—
—
—
—
—
—
6.72, dt, 8.5, 1.9d
—
116.0
157.6
116.0
132.1
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
8.75, s
—
7.32, s
—
134.8
—
118.8
—
6.72, 8.5, 1.9d
7.01, d, 8.5e
2.45, dd, 14.5, 8.0
2.41, dd, 14.5, 7.3
5.14, br t, 7.2
—
1.52, s
1.66, s
36.9 2.21, dd, 14.3, 8.3
2.24, dd,c
37.5 2.31, m
2.39,dd, 14.0, 7.0
2.44, dd, 14.0, 8.1
5.14, br t, 7.3
—
1.51, s
1.66, s
36.6 2.41, dd 14.6, 7.4
2.45, dd, 14.5, 8.1
36.8 2.42, dd, 14.4, 7.2 2.37, d, 7.6
2.46, dd, 14.3, 7.9
119.7 5.14, br t, 7.9
135.8
17.5 1.53, s
25.6 1.67, s
36.7
2′
3′
4′
5′
120.0 4.84, mf
119.8 5.10, t, 8.1
135.3
17.8 1.62, s
25.6 1.67, s
119.8 5.13, br t, 8.2
—
5.10, br t, 7.4
—
1.49, s
119.7
136.0
17.6
h
136.1
17.8 1.51, s
25.9 1.61
—
—
—
17.5 1.52, s
25.6 1.66, s
1.65, s
25.7
a H-10syn has a cis- and H-10anti a trans-configuration to H-11. b Overlapping signals. c Overlapping signals. d Overlapping signals. e Overlapping signals. f Overlapping with water signal. g Overlapping with
solvent signal. h Due to low purity, signals were not detected.

View Article Online
determined in this study (Table 5). In the co-structure of FtmPT1
with a non-hydrolysable substrate analogue DMASP, both C-2
and C-3 of the indole ring are within a distance of 3.8 Å to C-1
of DMASP and therefore prenylation would be possible for both
positions. However, C2-prenylation of 1a by FtmPT1 would
result in the formation of a more favorable product.24 A similar
behavior can be expected for the three stereoisomers of 1a. This
hypothesis is supported by the fact that C2-prenylated deriva-
tives were detected predominantly in the incubation mixtures of
1a, 7a and 8a. Other diketopiperazines might be located, due to
the given geometric and electrostatic features, in a slightly
unfavourable manner in the reaction site, so that the prenylation
at C-3 would be easier and indeed the yields of C3-prenylated
derivatives have increased in the incubation mixtures of these
compounds (Fig. 4). These observations were supported by
detailed inspection of the kinetic parameters obtained for eight
compounds (Table 5). As mentioned above, the three best
accepted substrates 1a, 5a and 9a with KM values of 0.16, 0.27
and 0.14 mM and catalytic efficiencies (kcat/KM) of 25 312, 6407
and 2214 s−1 M−1, respectively, were converted dominantly to
products of the b series. In contrast, the poorly accepted substrate
2a with KM values of 0.77 and 0.69 mM for the products of the
b and c series was converted to both products with a ratio of
3 : 1. A similar phenomenon was also observed for 14a
(Fig. S18†).
It can be proposed that both C2- and C3-prenylations begin
simultaneously with the attack of the electron-rich indole ring at
C2 and C-3 to C-1′ of the dimethylallyl cation (Fig. 5). Their
ratio would be dependent on the position of a given aromatic
substrate and accordingly the distance of C-2 to C-1′ as well as
C-3 to C-1′. Attacking at C-2 and C-3 would result in the for-
mation of the intermediate I via pathway A and II via pathway
B, respectively. For formation of the C2-prenylated product b,
just one deprotonating step to restore the indole system is necess-
ary. The fate of II would be attacking the cation at position C-2
of the indole ring by the electron-rich N12 nitrogen, as proposed
for reversely C3-prenylated indolines,9 resulting in the formation
of protonated form III of the end product c. Releasing of one
proton leads then to regularly C3-prenylated derivative c
(Fig. 5).
Table 4 NOE results of 10c as an example for regular C3-prenylated
diketopiperazines
Protons
Strength
H-11–H-10syn
H-1′–H-10anti
H-1′–H-2
H-4–H-10syn
H-1′–H-11
Strong
Medium
Strong
Medium
Not observed
Not observed
H-2′–H-11
Table 5 Kinetic parameters for the b and c-series
KM [mM]
b series
kcat [s−1
b series
]
kcat/KM [s−1 M−1
]
Sub
c series
c series
b series
c series
Conclusions
1a
2a
5a
9a
10a
11a
13a
14a
0.16
0.77
0.27
0.14
0.54
0.41
0.39
0.45
—
4.05
3.14
1.73
0.31
1.47
0.71
0.24
2.14
—
25 312
4078
6407
2214
2722
1732
615
—
In this study, we proved that the prenyltransferase FtmPT17,27
converted tryptophan-containing diketopiperazines to their pre-
nylated derivatives. For three of the four cyclo-Trp-Pro isomers
and cyclo-L-Trp-L-Leu, regularly C2-prenylated derivatives were
found as predominant products. For seven diketopiperazines,
regularly C3-prenylated products were detected in the enzyme
0.69
0.37
—
0.48
—
2.55
0.09
—
0.13
—
3696
246
—
271
—
0.14
—
0.06
—
429
—
4756
Fig. 5 Proposed reaction mechanism of the formation of C2- and C3-β-prenylated products catalysed by FtmPT1 with L-tryptophan-containing
cyclic dipeptides as examples. In analogy, attacking of the prenyl moiety from the opposite side of H-11 of the D-tryptophan-containing cyclic dipep-
tides would result in formation of C3-α-prenylated derivatives. For R1 and R2 moieties see Fig. 2.
9268 | Org. Biomol. Chem., 2012, 10, 9262–9270
This journal is © The Royal Society of Chemistry 2012

View Article Online
assays with total yields of up to 19%. Their structures were elu-
cidated unequivocally by NMR and MS analyses. This indicated
that the previously observed regioselectivity of many of the
identified prenyltransferases can be interrupted, at least in parts,
by using non-natural substrates. This reduces the convenience
for product isolation, increases however the structure diversity
obtained by a chemoenzymatic synthesis approach.
purified on HPLC (Agilent series 1200) with detection at
296 nm using a Multospher 120 RP-18-5 column (250 ×
10 mm, 5 μm, C+S Chromatographie Service) and a flow rate of
2.5 ml min−1. Water (solvent A) and methanol (solvent B) were
used as solvents. For isolation of 14c, 0.5% trifluoroacetic acid
was added to solvents A and B. For analysis of enzyme products,
a linear gradient of 65 to 100% (v/v) solvent B in 15 min was
used. The column was then washed with 100% solvent B for
5 min and equilibrated with 65% (v/v) solvent B for 5 min.
The fractions containing the prenylated diketopiperazines were
collected. After evaporation of the solvents, the obtained pro-
ducts were analysed spectroscopically.
Experimental section
Chemicals and protein purification
Synthesis and availability of dimethylallyl diphosphate
(DMAPP) and cyclic dipeptides were described previously.19
To get recombinant FtmPT1 for enzyme assays, Escherichia
coli M15 cells harbouring the plasmid pAG0127 were cultivated
in 2 l Erlenmeyer flasks containing 1 l liquid Luria–Bertani
medium supplemented with carbenicillin (50 μg ml−1) and kana-
mycin (25 μg ml−1). The cultures were grown at 37 °C for 4 h at
220 rpm until an absorption of 0.7 at 600 nm was reached. For
induction, IPTG was added to the cultures to a final concen-
tration of 0.1 mM and the bacteria were cultivated for a further
16 h at 30 °C before harvest. Purification of the recombinant
FtmPT1 was carried out according to the QIAexpressionist™
protocol from Qiagen.
ESI-MS and NMR analysis of the enzyme products
The obtained products were analyzed by NMR (in CD3OH,
CD3OD or CDCl3) and ESI-MS analyses. NMR spectra were
recorded at room temperature on a Bruker Avance 600 MHz or a
JEOL ECA-500 spectrometer. The heteronuclear single quantum
correlation (HSQC) and heteronuclear multiple bond correlation
(HMBC) spectra were recorded with standard methods.28
Gradient-selection for nuclear Overhauser effect spectroscopy
(NOESY) experiment was performed in phase-sensitive mode29
and processed with a Bruker TOPSPIN 2.1 or a MestReNova.5.2.2.
The positive electrospray ionization mass spectrometry
(HR-ESI-MS) was carried out on an AutoSpec instrument
(Micromass Co. UK Ltd).
Enzyme assay and HPLC analysis
Standard enzyme assays (100 μl) containing 50 mM Tris-HCl,
pH 7.5, 2 mM DMAPP, 1 mM cyclic dipeptide, 10 mM CaCl2
and 5 μg FtmPT1 were incubated at 37 °C for 2 h. KM values
were determined in 100 μl assays containing 0.4 μg FtmPT1,
2 mM DMAPP and cyclic dipeptides at concentrations of up to
2 mM. The assays were incubated at 37 °C for 30 min. The reac-
tions were terminated by addition of 100 μl methanol.
Acknowledgements
This work was supported by a grant from the Deutsche
Forschungsgemeinschaft (Li844/1-3 to S.-M.L.).
References
1 S.-M. Li, Nat. Prod. Rep., 2010, 27, 57–78.
After removal of the protein by centrifugation (15000g,
10 min, 4 °C), the enzyme products were analyzed on an Agilent
HPLC-Series 1200 by using a Multospher 120 RP-18-5 column
(250 × 4 mm, 5 μm, Chromatographie Service, Langerwehe,
Germany) at a flow rate of 1 ml min−1. Water (solvent A) and
methanol (solvent B) were used as solvents. For analysis of the
assay with cyclo-L-Trp-L-His, both solvents contained 0.5% (v/v)
trifluoroacetic acid. The assays were analyzed with a gradient
from 30% to 100% B over 30 min and a washing phase with
100% solvent B for 5 min was used. The column was equili-
brated with 30% solvent B for 5 min. The substances were
detected with a Photo Diode Array detector and illustrated at
296 nm in this study.
2 R. M. Williams, E. M. Stocking and J. F. Sanz-Cervera, Top. Curr.
Chem., 2000, 209, 97–173.
3 X. Yu, Y. Liu, X. Xie, X.-D. Zheng and S.-M. Li, J. Biol. Chem., 2012,
287, 1371–1380.
4 T. Bonitz, V. Alva, O. Saleh, A. N. Lupas and L. Heide, PLoS One, 2011,
6, e27336.
5 S. Yin, X. Yu, Q. Wang, X.-Q. Liu and S.-M. Li, Appl. Microbiol.
Biotechnol., 2012, DOI: 10.1007/s00253-012-4130-0.
6 S. Takahashi, H. Takagi, A. Toyoda, M. Uramoto, T. Nogawa, M. Ueki,
Y. Sakaki and H. Osada, J. Bacteriol., 2010, 192, 2839–2851.
7 A. Grundmann and S.-M. Li, Microbiology, 2005, 151, 2199–2207.
8 Y. Ding, J. R. Wet, J. Cavalcoli, S. Li, T. J. Greshock, K. A. Miller,
J. M. Finefield, J. D. Sunderhaus, T. J. McAfoos, S. Tsukamoto,
R. M. Williams and D. H. Sherman, J. Am. Chem. Soc., 2010, 132,
12733–12740.
9 W.-B. Yin, X.-L. Xie, M. Matuschek and S.-M. Li, Org. Biomol. Chem.,
2010, 8, 1133–1141.
10 W.-B. Yin, A. Grundmann, J. Cheng and S.-M. Li, J. Biol. Chem., 2009,
284, 100–109.
11 W.-B. Yin, X. Yu, X.-L. Xie and S.-M. Li, Org. Biomol. Chem., 2010, 8,
2430–2438.
12 W.-B. Yin, J. Cheng and S.-M. Li, Org. Biomol. Chem., 2009, 7,
2202–2207.
13 J. M. Schuller, G. Zocher, M. Liebhold, X. Xie, M. Stahl, S.-M. Li and
T. Stehle, J. Mol. Biol., 2012, 422, 87–99.
14 R. Raju, A. M. Piggott, X. C. Huang and R. J. Capon, Org. Lett., 2011,
13, 2770–2773.
15 N. Steffan, A. Grundmann, A. Afiyatullov, H. Ruan and S.-M. Li, Org.
Biomol. Chem., 2009, 7, 4082–4087.
Isolation of the enzyme products
The preparative enzymatic synthesis for the structural elucidation
was carried out as follows. To a 50 ml reaction tube, DMAPP
(2 mM), cyclic dipeptide (1 mM), CaCl2 (10 mM), Tris-HCL
(50 mM, pH = 7, 5) and FtmPT1 (1 mg) were added to a final
volume of 50 ml. The reaction mixtures were incubated at 37 °C
for 2 h and subsequently extracted twice with equal volume of
ethyl acetate and evaporated in vacuo. The products were
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 9262–9270 | 9269

View Article Online
16 C. B. Cui, H. Kakeya, G. Okada, R. Onose and H. Osada, J. Antibiot.,
1996, 49, 527–533.
22 E. Caballero, C. Avendañno and J. C. Menéndez, J. Org. Chem., 2003,
68, 6944–6951.
17 H. D. Jain, C. Zhang, S. Zhou, H. Zhou, J. Ma, X. Liu, X. Liao,
A. M. Deveau, C. M. Dieckhaus, M. A. Johnson, K. S. Smith,
T. L. Macdonald, H. Kakeya, H. Osada and J. M. Cook, Bioorg. Med.
Chem., 2008, 16, 4626–4651.
18 J. F. Sanz-Cervera, E. M. Stocking, T. Usui, H. Osada and
R. M. Williams, Bioorg. Med. Chem., 2000, 8, 2407–2415.
19 B. Wollinsky, L. Ludwig, A. Hamacher, X. Yu, M. U. Kassack and
S.-M. Li, Bioorg. Med. Chem. Lett., 2012, 22, 3866–3869.
20 H.-X. Zou, X. Xie, X.-D. Zheng and S.-M. Li, Appl. Microbiol. Biotech-
nol., 2011, 89, 1443–1451.
23 S.-M. Li, J. Antibiot., 2011, 64, 45–49.
24 M. Jost, G. Zocher, S. Tarcz, M. Matuschek, X. Xie, S.-M. Li and
T. Stehle, J. Am. Chem. Soc., 2010, 132, 17849–17858.
25 U. Metzger, C. Schall, G. Zocher, I. Unsöld, E. Stec, S.-M. Li, L. Heide
and T. Stehle, Proc. Natl. Acad. Sci. USA, 2009, 106, 14309–14314.
26 L. Y. P. Luk and M. E. Tanner, J. Am. Chem. Soc., 2009, 131,
13932–13933.
27 L. Wang, W.-B. Yin, S.-M. Li and X.-Q. Liu, Chin. J. Biochem. Mol.
Biol., 2009, 25, 580–584.
28 S. Berger and S. Braun, 200 and More NMR Experiments. A Practical
Course, Wiley-VCH, Weinheim, Germany, 2004.
21 W.-B. Yin, H.-L. Ruan, L. Westrich, A. Grundmann and S.-M. Li, Chem-
BioChem, 2007, 8, 1154–1161.
29 T.-L. Hwang and A. J. Shaka, J. Am. Chem. Soc., 1992, 114, 3157–3159.
9270 | Org. Biomol. Chem., 2012, 10, 9262–9270
This journal is © The Royal Society of Chemistry 2012