Complete Decoration of the Indolyl Residue in cyclo- l -Trp- l -Trp with Geranyl Moieties by Using Engineered Dimethylallyl Transferases

Article abstract of DOI:10.1021/acs.orglett.8b03124

Mutation of the gatekeeping residues for prenyl donor selectivity in six dimethylallyl transferases significantly increased their activities toward geranyl diphosphate. Forty-two geranylated derivatives were obtained from 15 cyclic dipeptides by using the engineered enzymes. Taking cyclo-l-Trp-l-Trp as an example, the geranyl moiety can be attached to all seven possible positions of the indole nucleus. This study demonstrates a convenient way to increase the structural diversity of geranylated products by structure-based engineering of the available dimethylallyl transferases.

Full text of DOI:10.1021/acs.orglett.8b03124

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
Complete Decoration of the Indolyl Residue in cyclo-L‑Trp‑L‑Trp with
Geranyl Moieties by Using Engineered Dimethylallyl Transferases
Ge Liao,†,§ Peter Mai,†,§ Jie Fan,† Georg Zocher,‡ Thilo Stehle,‡ and Shu-Ming Li*,†
̈
†Institut fu
Germany
̈
r Pharmazeutische Biologie und Biotechnologie, Philipps-Universitat Marburg, Robert-Koch Strasse 4, Marburg 35037,
‡
̈
̈
̈
̈
̈
Interfakultares Institut fur Biochemie, Eberhard Karls Universitat Tubingen, Tubingen 72076, Germany
S
* Supporting Information
ABSTRACT: Mutation of the gatekeeping residues for
prenyl donor selectivity in six dimethylallyl transferases
significantly increased their activities toward geranyl diphos-
phate. Forty-two geranylated derivatives were obtained from
15 cyclic dipeptides by using the engineered enzymes. Taking
cyclo-L-Trp-L-Trp as an example, the geranyl moiety can be
attached to all seven possible positions of the indole nucleus.
This study demonstrates a convenient way to increase the
structural diversity of geranylated products by structure-based
engineering of the available dimethylallyl transferases.
renylated natural products are hybrid structures contain-
indole ring can be achieved by the available indole PTs.13−15
Ping prenyl moieties (n × C ) or derivatives thereof and are However, the transfer of the C10 unit (geranyl) onto the indole
5
widely distributed in plants and microorganisms.1−3 Among
them, prenylated indole alkaloids including prenylated
tryptophan-containing cyclic dipeptides (CDPs) usually exhibit
a wide range of biological and pharmacological activities and,
therefore, are important sources for drug discovery and
development.2,4,5 Representative structures such as notoa-
mides,6,7 roquefortines,8 fumitremorgins,9 and felluta-
nines10−12 are commonly derived from CDPs and often
decorated with one or more dimethylallyl (C5) moieties
(Figure 1).
nucleus was rarely reported. Taking CDPs as examples, AnaPT
from Neosartorya fischeri catalyzes in the presence of
dimethylallyl diphosphate (DMAPP) reverse C3-prenylation
and also uses geranyl diphosphate (GPP) for their C6- and C7-
geranylation, as exemplified by cWW (1; for simplicity, one
letter codes are used for proteinogenic amino acids) in Scheme
1.16,17 AtaPT from Aspergillus terreus was reported to catalyze
the C4-geranylation of 1.18 Our recent reinvestigation on this
Scheme 1. Reverse Dimethylallyl C3-PT AnaPT from N.
a
fischeri Can Also Use GPP for C6- and C7-Geranylation
Figure 1. Representative examples of prenylated indole alkaloids
derived from cyclic dipeptides.
a
AtaPT from A. terreus attaches a geranyl moiety to C7 of 1. To
facilitate the reading flow, geranylated products are named by using
substrate number, prenylation pattern (a for regular or normal and b
for reverse prenylation), and prenylation position at the indole ring (1
to 7 for N1- to C7-alkylated products), e.g., 1a6 for C6-regularly
geranylated cWW.
The key reactions in the biosynthesis of the prenylated
natural products are connections of the prenyl moiety with its
acceptor catalyzed by prenyltransferases (PTs). PTs belonging
to the dimethylallyltryptophan synthase (DMATS) superfamily
catalyze the regiospecific Friedel−Crafts alkylation of aromatic
substrates, mainly of indole derivatives.13,14 The attachment of
the C5 unit to N1 and the six nonbridgehead carbons at the
Received: October 1, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.8b03124
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
reaction revealed a C7- instead of a C4-geranylation (data not
shown). Other CDPs with geranyl residues have not yet been
reported. Furthermore, only a few members of PTs are able to
accept GPP as a prenyl donor and act as geranyl transferases,
e.g. TleC,19 VrtC,20 and AmbP1.21 Therefore, it is necessary to
find new geranyl transferases or to manipulate available
dimethylallyl transferases for GPP acceptance to enrich the
biocatalysis toolboxes.
Inspired by the activities of AtaPT and AnaPT toward GPP
for prenylation of CDPs, we first evaluated the GPP acceptance
of five dimethylallyl transferases, FtmPT1,22 CdpNPT,23
BrePT,24 CdpC2PT,25 and CdpC3PT,26 in the presence of
CDPs as acceptors. These enzymes catalyze regular (normal)
or reverse prenylations at C-2 or C-3 of the indole ring. The
tested CDPs include 1, cWP (2) and its stereoisomers (3−5),
cWA (6) and its stereoisomers (7−9), cWL (10), cWY (11),
cWF (12), cWG (13), (S)-(14), and (R)-benzodiazepinedin-
one (15). HPLC analysis of the incubation mixtures showed
that the five PTs well accepted their natural or best aromatic
substrates in the presence of DMAPP (Figure S1 in the
Supporting Information (SI)), confirming the presence of
active proteins. In the presence of GPP, CdpC2PT showed
clear activity toward 1, 2, 14, and 15 (Figure S2). The other
four enzymes exhibited very low (CdpC3PT toward 1, 2, 5,
and 14, Figure 2, Figure S3) or no activity (CdpNPT,
FtmPT1, and BrePT) (Figure 2, Figures S4−S6), when GPP
was used as a prenyl donor.
AtaPT was previously reported to have broad tolerance for
prenyl donors. The residue Gly326 was proven to contribute to
the enlarged pocket for donor binding.18 Mutation of Gly326
to methionine dramatically reduced the activity of G326M
toward GPP and FPP, but substantially enhanced the catalytic
activity with DMAPP. This proved that the size of the residue
at 326 controls the prenyl donor selectivity in AtaPT. Mutation
of Met328 in the tryptophan prenyltransferase FgaPT2 to
small amino acids such as cysteine, threonine, serine, glycine,
or alanine significantly increased the acceptance of GPP.27
Encouraged by these findings, structure-based protein
sequence alignments of AtaPT and FgaPT2 with the five
PTs mentioned above led to the identification of the
corresponding residues Met364 in FtmPT1, IIe337 in BrePT,
Thr351 in CdpC2PT, Met349 in CdpNPT, and Phe335 in
CdpC3PT (Figure 3, Figure S7).
Molecular modeling with the ternary crystal structures of
FtmPT1 and CdpNPT23,28 also uncovered Met364 in FtmPT1
and Met349 in CdpNPT as key residues for DMAPP
selectivity (Figure S8). The side chain of this residue seals
the active site to form a pocket for DMAPP. In the active site
of both enzymes, GPP would clash with the side chain of the
gatekeeping residue methionine. Therefore, reducing the steric
pressure in the DMAPP binding site by replacement of this
bulky residue with a small amino acid like glycine would
increase the enzymatic activity toward GPP.
Figure 2. HPLC chromatograms of the incubation mixtures of six
dimethylallyl transferases and their corresponding mutants with their
natural or best accepted substrates in the presence of GPP. Regular
geranylated products are labeled in red, and reverse, in blue. The
product yields are given in parentheses after the product number. See
please Figure S1 for their behavior toward DMAPP.
To switch the prenyl donor specificity from DMAPP to GPP
for FtmPT1, BrePT, CdpNPT, and CdpC3PT and to expand
the GPP acceptance of CdpC2PT, we replaced the mentioned
amino acids of the five dimethylallyl transferases by glycine,
resulting in the mutants FtmPT1_M364G, BrePT_I337G,
C d p N P T _ M 3 4 9 G , C d p C 2 P T _ T 3 5 1 G , a n d
CdpC3PT_F335G, respectively (see Table S1 for constructs
and SI for description). As expected, replacing bulky amino
acids by glycine significantly improved in most cases the CDP
consumption in the presence of GPP (Figure 2, Figures S2−
S6, S9). In comparison, the acceptance of DMAPP by the
mutants was reduced to different levels (Figure S1). In sharp
contrast to no activity in the presence of GPP by wild type
CdpNPT, FtmPT1, and BrePT, all the tested CDPs (1−15)
were accepted by CdpNPT_M349G with total product yields
from 13.7% to 95.8% (Figure 2, Figures S4 and S9). 1−13 with
a diketopiperazine ring were consumed by FtmPT1_M349G
with total product yields from 9.6% to 75.4%. The two
benzodiazepinediones 14 and 15 were not accepted (Figure 2,
Figures S5 and S9).
B
DOI: 10.1021/acs.orglett.8b03124
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 2. Structures of Enzyme Products with Different
Prenylation Patterns and Positions (Products of 1 Are
Presented in Scheme 3)
Figure 3. Selected region of structure-based alignments of PTs used
in this study with AtaPT and FgaPT2. See Figure S7 for complete
alignments.
The activity of BrePT toward GPP can be strongly increased
by mutation at the gatekeeping residue Ile337, at least for its
natural substrate 2 (79.8%) and the two stereoisomers 3
(71.4%) and 4 (28.4%) (Figure 2, Figure S6). In the case of
CdpC3PT_F335G, conversion of almost all substrates was
significantly increased, especially for 10. Its conversion was
improved from 2.9% by the wild type enzyme to 51.8% by this
mutant (Figure 2, Figure S3). CdpC2PT_T351G exhibited
different behaviors toward 1−15 from CdpC2PT, with distinct
products or different ratios of the same products (Figure 2,
Figures S2 and S9). Detailed inspection of the HPLC
chromatograms of 1−15 with FtmPT1_M364G, BreP-
T_I337G, CdpNPT_M349G, CdpC2PT_T351G, and
CdpC3PT_F335G (Figures S2−S6, S9) revealed the presence
of one to four product peaks with different ratios. Most
reactions have two or three products. This shows the relaxed
regiospecific geranylation by the mutants in comparison to the
wild type enzymes in the presence of DMAPP.
For structure elucidation, 41 representative products of
selected substrates were isolated by preparative HPLC from
different incubation mixtures (Table S2) and subjected to HR-
ESI-MS and 1H NMR analyses (Tables S2−S18, Figures S10−
S54). Structure elucidation (see the SI) confirmed reverse or
regular geranylation at different positions of the indole ring
(Scheme 2).
Three of the tested PTs catalyze regular (FtmPT1) or
reverse (BrePT and CdpC2PT) C2-prenylations of CDPs in
the presence of DMAPP.22,24,25 Structure elucidation proved
that BrePT_I337G catalyzes solely or predominantly still
reverse C2-geranylation of 2− 4. In the reaction mixtures of
FtmPT1_M364G with six substrates, regularly C2-prenylated
derivatives were obtained as major products, i.e. 1a2, 2a2, 3a2,
10a2, 11a2, and 12a2. Regularly C3-prenylated derivatives
such as 1a3, 3a3, 6a3, 8a3, 10a3, 11a3, 12a3, and 13a3 were
identified in at least eight incubation mixtures. This is not
surprising, because regularly C3-prenylated CDPs were also
reported for FtmPT1 in the presence of DMAPP.29 In that
study, however, regularly C2-prenylated derivatives have been
identified as the major products. Interestingly, in the assays of
FtmPT1_M364G with GPP and 6, 8, and 13, no C2-, but
regularly N1- and C3-geranylated derivatives were obtained
(Scheme 2, Figure S5). The behavior of CdpC2PT_T351G
toward GPP also differs significantly from that of the wild type
enzyme. For example, 1 was converted by CdpC2PT to 1b2,
1a5, and 1a7, while CdpC2PT_T351G mainly catalyzes C2-
geranylation, resulting in the formation of mono- (1b2) and
digeranylated (1b2-2) derivatives (Figure 2, Figure S9).
CdpC3PT and CdpNPT both catalyze reverse C3-prenylation
of CDPs in the presence of DMAPP.23,26 Reverse C3-
geranylation was also detected for CdpNPT_M349G with
the selected substrates 1, 5, and 14. In addition, regularly C6-
geranylated products 1a6, 5a6, and 14a6 were also identified.
In the incubation mixture of CdpNPT_M349G with 15, the
product 15a6 was even the predominant peak (Figure S4).
CdpC3PT_F335G converted 10 to a mixture of at least four
products 10a1, 10b1, 10b3, and 10a7 with 10b3 as the major
one (Figure S3). These results indicate the trends of C6- and
C7-geranylation of CDPs by dimethylallyl transferases, which
catalyze C3-prenylation in the presence of DMAPP, as
observed for AnaPT (Scheme 1).16,17
In summary, our results presented in this study showed that
replacing the gatekeeping residue not only changes the
tolerance of PTs toward prenyl donor with longer chain but
C
DOI: 10.1021/acs.orglett.8b03124
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
also breaks their regiospecificity, leading to several prenylated
products in one incubation. To some extent, this change
broadens their application in creation of structure diversity of
geranylated indole derivatives. The possible reason for the
relaxed regioselective prenylation by the mutants could be that
the enlarged pocket allows the binding of aromatic substrates
in different orientations. However, the reduced regioselectivity
requires more efforts for product isolation. Therefore, further
mutagenesis experiments are necessary in the future to increase
the desired regio- and stereoselectivity. This would require
more structural information, e.g. of the available single mutants
of the gatekeeping residues. Nevertheless, the single mutants
constructed in this study clearly showed different preferences
toward the 15 tested CDPs and converted them to geranylated
derivatives at various positions. As shown in Scheme 2,
products obtained in this study also carry the geranyl moiety at
the N1-, C2-, and C3-position of the pyrrole ring in a regular or
reverse manner.
Scheme 3. Attachment of Geranyl Moiety to Different
Positions of 1 via One-Step Chemoenzymatic Reactions
To the best of our knowledge, known geranylations of
diverse substrates usually occur at the benzene ring of the
indole nucleus, e.g. C7-geranylation of indolactam V by TleC,
C6-geranylation of tryptophan derivatives by 6-DMATSSa, and
C6- and C7-geranylation of CDPs by AnaPT.16,19,30 As
mentioned above, the reported C4-geranylation of 1 by
AtaPT18 should be at C-7. Attachment of the geranyl moiety
has been reported at neither C-4 or C-5 nor at N-1, C-2, or C-
3 prior to this study.
The achieved relaxed regioselective geranylation by the
mutants encouraged us to decorate all nucleophilic reactive
positions (N-1 and the six nonbridgehead carbons) at the
indole ring of one CDP with the geranyl moiety. For this
purpose, 1 has been selected as the favorable CDP. HPLC
analysis of the incubation mixture of 1 and GPP with
FtmPT1_M364G, CdpNPT_M349G, CdpC2PT, and
CdpC2PT_T351G revealed the presence of two to four
product peaks each with product yields between 37.4 and
68.0% (Figure S9). Isolation and structure elucidation
confirmed geranylations at N-1 (1a1), C-2 (1a2 and 1b2),
C-3 (1a3 and 1b3), C-5 (1a5), C-6 (1a6), and C-7 (1a7) of
the indole ring (Scheme 3). To get the remaining C4-position
filled by geranylation, we applied the mutant FgaPT2_M328G,
which uses GPP for C4-geranylation of tryptophan.27 HPLC
analysis of the incubation mixture of 1 and GPP with
FgaPT2_M328G indicated product formation with a low
yield (7.2%, Figure 2). In another previous study, we
demonstrated that mutation on Arg244 of FgaPT2 significantly
increased the catalytic activity toward CDPs in the presence of
DMAPP.31 To enhance the geranylation of 1, we combined
mutations at Met328 and Arg244 and created the double
mutant FgaPT2_M328G_R244Q. As shown in Figure 2, one
peak with a product yield of 29.7% was detected. Structure
elucidation confirmed the desired geranylated product 1a4
(Scheme 3). In this way, nine products with a geranyl moiety
at all nucleophilic reactive positions of the indole ring can be
obtained from 1 by one-step chemoenzymatic reactions
(Scheme 3, Figure S9).
an excellent example for rational engineering of desirable
enzymes. Mutation of the gatekeeping residues in six PTs
turned on or improved the acceptance of GPP for geranylation
of diverse CDPs. Forty-two geranylated products were
obtained from in vitro enzymatic reactions of CDPs and
GPP. Using 1 as an acceptor and GPP as a donor, N-1 and the
six nonbridgehead carbons of the indole ring were decorated
with geranyl moieties. Derivatives of other CDPs could also be
obtained by the mutants described in this study or by mutation
of additional available dimethylallyl transferase at the
corresponding positions.
ASSOCIATED CONTENT
■
With the tremendously increasing availability of protein
structures, structure-based enzyme engineering becomes more
important for modification of drug-like molecules. Indolactam
geranyl transferase (TleC) can be changed to dimethylallyl
transferase by small-to-large mutation.19 MpnD and PagF were
changed from selective dimethylallyl to geranyl transferases by
large-to-small substitution.19,32 In analogy, our study provides
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.8b03124.
Experimental procedures, structural elucidation, and
NMR spectra of geranylated products (PDF)
D
DOI: 10.1021/acs.orglett.8b03124
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(27) Mai, P.; Zocher, G.; Stehle, T.; Li, S.-M. Org. Biomol. Chem.
2018, 16, 7461.
(28) Jost, M.; Zocher, G.; Tarcz, S.; Matuschek, M.; Xie, X.; Li, S.-
M.; Stehle, T. J. Am. Chem. Soc. 2010, 132, 17849.
(29) Wollinsky, B.; Ludwig, L.; Xie, X.; Li, S.-M. Org. Biomol. Chem.
2012, 10, 9262.
(30) Winkelblech, J.; Li, S.-M. ChemBioChem 2014, 15, 1030.
(31) Fan, A.; Li, S.-M. Appl. Microbiol. Biotechnol. 2016, 100, 5389.
(32) Estrada, P.; Morita, M.; Hao, Y.; Schmidt, E. W.; Nair, S. K. J.
Am. Chem. Soc. 2018, 140, 8124.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: shuming.li@staff.uni-marburg.de.
ORCID
Shu-Ming Li: 0000-0003-4583-2655
Author Contributions
§G.L. and P.M. contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Rixa Kraut, Stefan Newel (University of Marburg)
for synthesis of prenyl donors, taking MS and NMR spectra,
respectively. S.M.L. acknowledges the DFG for funding the
Bruker microTOF QIII mass spectrometer (INST 160/620-1).
Ge Liao (201607565014) and Jie Fan (201507565006) are
scholarship recipients from the China Scholarship Council.
REFERENCES
(1) Heide, L. Curr. Opin. Chem. Biol. 2009, 13, 171.
(2) Li, S. M. Nat. Prod. Rep. 2010, 27, 57.
■
(3) Fan, A.; Winkelblech, J.; Li, S.-M. Appl. Microbiol. Biotechnol.
2015, 99, 7399.
(4) Borthwick, A. D. Chem. Rev. 2012, 112, 3641.
(5) Wollinsky, B.; Ludwig, L.; Hamacher, A.; Yu, X.; Kassack, M. U.;
Li, S. M. Bioorg. Med. Chem. Lett. 2012, 22, 3866.
(6) Kato, H.; Yoshida, T.; Tokue, T.; Nojiri, Y.; Hirota, H.; Ohta, T.;
Williams, R. M.; Tsukamoto, S. Angew. Chem., Int. Ed. 2007, 46, 2254.
(7) Tsukamoto, S.; Kato, H.; Samizo, M.; Nojiri, Y.; Onuki, H.;
Hirota, H.; Ohta, T. J. Nat. Prod. 2008, 71, 2064.
́
́
́
(8) García-Estrada, C.; Ullan, R. V.; Albillos, S. M.; Fernandez-
̈
Bodega, M. A.; Durek, P.; von Dohren, D. H.; Martín, J. F. Chem. Biol.
2011, 18, 1499.
(9) Li, S.-M. J. Antibiot. 2011, 64, 45.
(10) Nuber, B.; Hansske, F.; Shinohara, C.; Miura, S.; Hasumi, K.;
Endo, A. J. Antibiot. 1994, 47, 168.
(11) Kozlovsky, A. G.; Vinokurova, N. G.; Adanin, V. M.; Burkhardt,
̈
G.; Dahse, H. M.; Grafe, U. J. Nat. Prod. 2000, 63, 698.
(12) Kozlovsky, A. G.; Vinokurova, N. G.; Adanin, V. M.; Burkhardt,
̈
G.; Dahse, H.-M.; Grafe, U. J. Nat. Prod. 2001, 64, 553.
(13) Winkelblech, J.; Fan, A.; Li, S.-M. Appl. Microbiol. Biotechnol.
2015, 99, 7379.
(14) Tanner, M. E. Nat. Prod. Rep. 2015, 32, 88.
(15) Yu, X.; Liu, Y.; Xie, X.; Zheng, X.-D.; Li, S.-M. J. Biol. Chem.
2012, 287, 1371.
(16) Pockrandt, D.; Li, S.-M. ChemBioChem 2013, 14, 2023.
(17) Yin, W.-B.; Xie, X.-L.; Matuschek, M.; Li, S.-M. Org. Biomol.
Chem. 2010, 8, 1133.
(18) Chen, R.; Gao, B.; Liu, X.; Ruan, F.; Zhang, Y.; Lou, J.; Feng,
K.; Wunsch, C.; Li, S.-M.; Dai, J.; Sun, F. Nat. Chem. Biol. 2017, 13,
226.
(19) Mori, T.; Zhang, L.; Awakawa, T.; Hoshino, S.; Okada, M.;
Morita, H.; Abe, I. Nat. Commun. 2016, 7, 10849.
(20) Chooi, Y. H.; Wang, P.; Fang, J.; Li, Y.; Wu, K.; Wang, P.;
Tang, Y. J. Am. Chem. Soc. 2012, 134, 9428.
(21) Liu, X.; Hillwig, M. L.; Koharudin, L. M.; Gronenborn, A. M.
Chem. Commun. (Cambridge, U. K.) 2016, 52, 1737.
(22) Grundmann, A.; Li, S.-M. Microbiology 2005, 151, 2199.
(23) Schuller, J. M.; Zocher, G.; Liebhold, M.; Xie, X.; Stahl, M.; Li,
S.-M.; Stehle, T. J. Mol. Biol. 2012, 422, 87.
(24) Yin, S.; Yu, X.; Wang, Q.; Liu, X. Q.; Li, S.-M. Appl. Microbiol.
Biotechnol. 2013, 97, 1649.
(25) Mundt, K.; Li, S.-M. Microbiology 2013, 159, 2169.
(26) Yin, W.-B.; Yu, X.; Xie, X.-L.; Li, S.-M. Org. Biomol. Chem.
2010, 8, 2430.
E
DOI: 10.1021/acs.orglett.8b03124
Org. Lett. XXXX, XXX, XXX−XXX