Cystomanamides: Structure and biosynthetic pathway of a family of glycosylated lipopeptides from myxobacteria

Article abstract of DOI:10.1021/ol500779s

Cystomanamides A-D were isolated as novel natural product scaffolds from Cystobacter fuscus MCy9118, and their structures were established by spectroscopic techniques including 2D NMR, LC-SPE-NMR/-MS, and HR-MS. The cystomanamides contain β-hydroxy amino acids along with 3-amino-9-methyldecanoic acid that is N-glycosylated in cystomanamide C and D. The gene cluster for cystomanamide biosynthesis was identified by gene disruption as PKS/NRPS hybrid incorporating an iso-fatty acid as starter unit and including a reductive amination step at the interface of the PKS and NRPS modules.

Full text of DOI:10.1021/ol500779s

Letter
pubs.acs.org/OrgLett
Cystomanamides: Structure and Biosynthetic Pathway of a Family of
Glycosylated Lipopeptides from Myxobacteria
Lena Etzbach, Alberto Plaza, Ronald Garcia, Sascha Baumann, and Rolf Muller*
̈
Department of Microbial Natural Products, Helmholtz-Institute for Pharmaceutical Research Saarland (HIPS), Helmholtz Centre for
Infection Research (HZI) and Pharmaceutical Biotechnology, Saarland University, Campus C2 3, 66123 Saarbrucken, Germany
̈
German Center for Infection Research (DZIF), Partner site Hannover-Braunschweig 38124, Germany
S
* Supporting Information
ABSTRACT: Cystomanamides A−D were isolated as novel
natural product scaffolds from Cystobacter fuscus MCy9118,
and their structures were established by spectroscopic
techniques including 2D NMR, LC-SPE-NMR/-MS, and
HR-MS. The cystomanamides contain β-hydroxy amino
acids along with 3-amino-9-methyldecanoic acid that is N-
glycosylated in cystomanamide C and D. The gene cluster for
cystomanamide biosynthesis was identified by gene disruption
as PKS/NRPS hybrid incorporating an iso-fatty acid as starter
unit and including a reductive amination step at the interface of the PKS and NRPS modules.
yxobacteria have proven to exhibit a fascinating capacity
10 double-bond equivalents (DBE). The 1H NMR spectrum of
Mto produce chemically intriguing natural products which cystomanamide B in CD3OD exhibited signals characteristic of
a peptide including three α-proton signals at δ 4.77 (1H, dd, J =
7.4, 6.3 Hz), 4.80 (1H, d, J = 3.75 Hz), and 4.49 (1H, d, J =
3.25 Hz). Moreover, a downfield pair of triplets at δ 7.28 (2H,
t, J = 7.45 Hz) and 7.20 (1H, t, J = 7.45 Hz) and a doublet at δ
7.34 (2H, d, J = 7.45 Hz) were observed (see Table S4,
Supporting Information). The HSQC spectrum revealed the
presence of several methylenes between δH 1.2 and 2.9 and two
oxygenated methines at δC 72.8/δH 4.29 (β-OH-Asn) and δC
74.9/δH 5.22 (β-OH-Phe). A detailed analysis of the 2D NMR
data obtained from HSQC, HMBC, COSY, and TOCSY
experiments indicated the presence of asparagine, β-hydrox-
yasparagine (β-OH-Asn), and β-hydroxyphenylalanine (β-OH-
Phe). Additionally, a 3-amino-9-methyldecanoic acid residue
(AMDA) was identified, and its partial structure was deduced
as follows. A sequential spin system starting from two methyl
groups at δ 0.89 (Me-10 and Me-11AMDA) attached to a
methine at δ 1.55 (H-9AMDA) and comprising five sequential
methylene groups H2-8 to H2-4 was obtained from COSY and
TOCSY correlations. COSY correlations extended this frag-
ment by an additional aminomethine group at δ 3.54 (H-
often show unique structural elements rarely produced by other
sources.1 Myxobacterial secondary metabolites belong to
multiple structural classes, and many of these diverse
compounds originate from mixed polyketide-nonribosomal
peptide biosynthetic pathways.2 Complex multimodular
enzymes involving various catalytic and structural domains
are responsible for the biosynthesis of these versatile
structures.3 The ability to generate a wide range of complex
natural products is expanded by enzymes introducing β-
branching and post-PKS and NRPS reactions such as
hydroxylation, glycosylation, and epimerization.4 Our discovery
strategy to determine natural products with novel structural
frameworks includes UHPLC/HRMS-based metabolomics for
strain selection and dereplication as well as hyphenated
chromatographic methods such as LC-SPE-NMR/-MS for
isolation driven by chemical and structural features. This
approach indicated the presence of unusual peptides in an
extract of Cystobacter fuscus MCy9118 and finally led to the
isolation of a new family of lipopeptides, the cystomanamides
(ctm). These new linear peptides comprise exclusively
nonproteinogenic amino acids and bear an unusual 3-amino-
9-methyldecanoic acid residue at the N-terminus. Cystomana-
mides C and D contain an N-linked glycosylation, which is one
of the rare examples of late-stage modification reactions of
myxobacterial metabolites.5 In addition to these four new
myxobacterial natural products, the strain was also found to
produce the known antibiotics althiomycin,6,7 roimatacene,8
and myxochelin A9 and B.10
3
AMDA) followed by a methylene at δ 2.68, 2.50 (H2-2AMDA).
Finally, an HMBC correlation from H2-2AMDA to a carbonyl
resonance at δ 173.2 (C-1AMDA) clearly indicated that AMDA
represents a β-amino acid. The sequence was established
through HMBC correlations from α-protons to carbonyl
carbons of adjacent residues to result in the sequence:
AMDA-Asn-(β-OH-Asn)-(β-OH-Phe). Long-range HBMC
correlations from the α-proton at δ 4.77 (H-2Asn) to C-1AMDA
HRESIMS of cystomanamide A (1) displayed an [M + H]+
peak at m/z 609.3248 (calcd for C28H45N6O9, 609.3243),
consistent with the molecular formula C28H44N6O9 containing
Received: March 14, 2014
Published: April 15, 2014
© 2014 American Chemical Society
2414
dx.doi.org/10.1021/ol500779s | Org. Lett. 2014, 16, 2414−2417

Organic Letters
Letter
connected AMDA to the N-terminus of asparagine, thereby
completing the structure of 1 as linear tetrapeptide.
HRESIMS of cystomanamide B (2) displayed an [M + H]+
peak at m/z 860.4039 (calcd for C40H58N7O14, 860.4036),
consistent with the molecular formula C40H57N7O14 containing
16 DBE. The 1H NMR spectra of 2 in comparison to that of 1
exhibited additional signals for one α-proton at δ 4.43 (1H, dd,
J = 7.5, 5.5 Hz) and two downfield doublets at δ 7.05 (2H, d, J
= 8.4 Hz) and δ 6.68 (2H, d, J = 8.4 Hz). The HSQC spectrum
revealed the additional presence of one methylene at δC 38.2/
δH 3.12, 2.97 (Tyr), one oxygenated methylene at δC 67.9/δH
4.32, 4.16 (glyceric acid GA), and one oxygenated methine at
δC 71.0/δH 4.24 (GA). This evidence in combination with the
DQF-COSY and HMBC correlations indicated that in
comparison to 1, cystomanamide B bears an additional glyceric
acid (GA) and tyrosine residue. HMBC long-range correlations
from α-protons to carbonyl carbons of adjacent residues and
from the methylene protons H2-3GA to C-1β−OH‑Phe resulted in
the linear depsipeptide 2 with the sequence AMDA-Asn-(β-
OH-Asn)-(β-OH-Phe)-GA-Tyr.
The ESI-MS/MS fragmentation patterns of cystomanamide
C (3) and D (4) indicated a glycosylation. More precisely, 3
showed the loss of a 162 mass unit fragment compared to 1,
suggesting the presence of a hexose residue. The 1H NMR and
HSQC spectra in comparison with those of 1 showed various
additional signals between 3.1 and 4.2 ppm with different
intensities belonging to three rotamers of a sugar moiety. Based
on HMBC and ROESY NMR data, the sugar residue was
identified as fructose with β-D-fructopyranose as the most
abundant rotamer, a sugar moiety also found in kwansonine A
and B.11 The HMBC spectrum showed a key correlation from
the methylene at δ 3.27 (H-1FRU) to the aminomethine at δ
3.59 (H-3AMDA) indicating an N-linked glycosylation of the β-
amino fatty acid chain. Brabantamides are another example for
compounds showing a sugar attached to a fatty acid derived
moiety.12 Cystomanamide D showed similar NMR and MS
data with a mass difference of 16 Da compared to 3 that was
explained by the absence of the β-hydroxylation of the
phenylalanine residue. The absolute configuration of the
amino acid residues was elucidated by MS detected chromato-
graphic analysis of the L- and D-FDLA (1-fluoro-2,4-dinitro-
phenyl-5-L/D-leucinamide) derivatives of the acid hydrolysate
of cystomanamides A−C and comparison with respective
standards.13 The amino acid residues were assigned as D-
asparagine, L-erythro-β-OH-asparagine, D-threo-β-OH-phenyl-
alanine and D-tyrosine. The R configuration of C-3 of the
AMDA residue was also determined by this method. β-Amino
fatty acids derivatized with Marfey’s reagent show a behavior
analogue to α-amino acids where the L series elutes earlier than
the D series.14 The D configurations of glyceric acid in 2 and the
fructose in 3 were both established by chiral HPLC of the acid
hydrolysate and comparison with respective standards.
Figure 1. Structures of cystomanamide A (1), B (2), C (3), and D (4).
and analyzed using Pfam,16 NRPS predictor2,17 and PKS/
NRPS analysis.18 This analysis led to the prediction of one
loading module, six elongation modules, and one termination
module. Genes not resulting in predictions related to PKS or
NRPS domains were analyzed via the BLAST algorithm19 using
the nonredundant sequence database at the National Center for
Biotechnology Information (NCBI). CtmA encodes for a
complex protein that contains functional domains belonging
to fatty acid synthases, polyketide synthases, amino transferases,
and nonribosomal peptide synthetases. The gene shows
similarity to mycA encoding for mycosubtilin synthase subunit
A.20 Feeding experiments with L-[methyl-2H3]leucine indicated
a leucine-derived branched chain carboxylic acid starter unit.21
The assembly line starts with a CoA ligase domain responsible
for recognition and activation of the starter molecule 9-
methyldecanoic acid, an iso-odd fatty acid (see Figure 2). The
activated substrate is then transferred to the first acyl carrier
protein in the loading module.22 It undergoes one elongation
step using malonate as a substrate to form the β-ketothioester.
The amino transferase domain (AMT) located in module 1 at
the interface of the PKS and NRPS modules next reductively
aminates the β-ketothioester which is then passed on to the
NRPS module as shown for mycosubtilin biosynthesis.23 This
AMT shows significant (57%) similarity to the AMT found in
MycA. It was proven in vitro that the AMT catalyzes amine
transfer from an amino acid to a protein-bound β-ketothioester
to generate the corresponding protein-bound β-aminothioester
dependent on pyridoxal 5′-phosphate (PLP).23 Biosynthesis
continues with three NRPS based reaction cycles. In silico
analysis of the A domain specificities is consistent with the
incorporated amino acids. Modules 2, 4, and 6 contain
epimerization domains that are responsible for the trans-
formation of the L- into the respective D-amino acid. The
incorporation of D-Asn, D-Phe, and D-Tyr are in accordance
with the structure and the absolute configuration of 2. The
second NRPS module, module 3, is split into two proteins,
On the basis of the chemical structures of the cystomana-
mides (see Figure 1), it seemed likely that these compounds are
products of a PKS/NRPS hybrid megasynthetase. Genome
sequence data of C. fuscus MCy9118 were generated using
Illumina sequencing. A retrobiosynthetic approach in combi-
nation with antiSMASH 2.015 analysis of the draft genome
sequence led to the identification of the ctm biosynthetic gene
cluster. The predicted gene cluster consists of 10 open reading
frames (ORFs) and has an overall GC content of 69.8%.
To further analyze the catalytic domains and the A domain
substrate specificity, the open reading frames were translated
2415
dx.doi.org/10.1021/ol500779s | Org. Lett. 2014, 16, 2414−2417

Organic Letters
Letter
Figure 2. Gene cluster of cystomanamides in Cystobacter fuscus MCy9118 and the biosynthetic pathway of 2.
CtmA and CtmC, which can be regarded as unusual but is not
unprecedented.22 The condensation domain is encoded by
ctmA whereas the adenylation domain and the peptidyl carrier
protein are encoded by ctmC. The module bears an additional
condensation domain encoded by ctmC, which seems to be
inactive due to the missing catalytic histidine residues in the
active site.24 Genetically, the module is separated by ctmB
whose product is homologous to TauD from Streptomyces
auratus (44.9%), a well-studied nonheme iron hydroxylase.
CtmB contains the conserved 2-His-1-carboxylate facial triad
responsible for iron binding25,26 as well as the conserved Arg
residue that ligates α-ketoglutarate.27 It also shows similarity
(35.1%) to SyrP from Pseudomonas syringae, which is
responsible for the β-hydroxylation of an aspartyl residue in
syringomycin E biosynthesis.28 This indicates that CtmB is
most likely responsible for the hydroxylation of the second
asparagine residue to form L-erythro-β-OH-asparagine.
Module 5 has a similar organization as NRPS modules but
instead of an adenylation domain it contains an unusual domain
that exhibits the three conserved motifs specific for the HAD
superfamily.29 BLAST results show similarity to FkbH30 from
Streptomyces hygroscopicus subsp. ascomyceticus (29.5%) and
OzmB31 from Streptomyces albus (30.6%) which are responsible
for the formation of glyceryl-ACP in the biosynthesis of the
polyketide natural products FK520 and oxazolomycin. For
OzmB it was demonstrated that it acts as a bifunctional glyceryl
transferase/phosphatase that first binds D-1,3-bisphosphogly-
cerate from the glycolytic pool to form the D-3-phosphogly-
ceryl-S-OzmB intermediate. In the next step, it removes the
phosphate group to receive the D-3-glyceryl-S-OzmB species
(acting as a phosphatase), and finally it acts as glyceryl
transferase by transferring the glyceryl group to an acyl carrier
protein (ACP).31 To the best of our knowledge, CtmD is the
first protein containing such a domain integrated into an NRPS
module underlining the enormous potential of these modular
megaenzymes for combinatorial biosynthesis. Although the
condensation domain of module 5 is homologous to the typical
amide-forming condensation domains, it most likely catalyzes
the formation of an ester bond instead of a peptide bond. The
biochemical evidence for condensation domains being able to
catalyze ester bond formation was given in studies concerning
the mycotoxins fumonisins and the antitumor antibiotic C-
1027.32,33 The same mechanism with C domains embedded in
elongation modules that are responsible for chain extension via
ester bond formation has been proposed for the biosynthesis of
some other nonribosomal peptides.34−36 The last step in the
biosynthesis of the cystomanamides is the incorporation and
epimerization of tyrosine in module 6. The assembly line is
terminated by a thioesterase domain that releases the product
to receive the linear PKS/NRPS product 2. The glycosylation
of 3 and 4 requires enzymes encoded outside the aglycon
cluster which could not be identified as 41 genes were
annotated as glycosyltransferase in the genome of Cystobacter
fuscus MCy9118.
To further confirm that the candidate gene cluster is
responsible for biosynthesis of the cystomanamides, the PKS/
NRPS gene ctmA was inactivated using a single crossover
homologous recombination knockout strategy. The resulting
ctmA mutant lost the ability to produce 1 as well as its
analogues (see Figures S4 and S5, Supporting Information),
therefore validating that the proposed gene cluster is indeed
responsible for cystomanamide biosynthesis. To explore the
relation of the surrounding open reading frames in the
biosynthetic pathway, we inactivated orf 2, orf4 and orf5.
Cultivation of the resulting orf 2 and orf4 mutants and
subsequent HPLC analysis revealed that inactivation of both
genes abolished production of 1 and its analogues (see Figures
S4 and S5, Supporting Information). The product of orf1 shows
2416
dx.doi.org/10.1021/ol500779s | Org. Lett. 2014, 16, 2414−2417

Organic Letters
Letter
(9) Kunze, B.; Bedorf, N.; Kohl, W.; Hofle, G.; Reichenbach, H. J.
̈
similarity to the JmjC domain containing proteins that are
connected to transcription factors,37 whereas the products of
orf 2, orf 3, and orf4 show similarity to the very heterogeneous
group of β-lactamases. Orf1 to orf4 are most likely involved in
the regulation of gene expression. At the same time, HPLC
analysis of a cultivation of orf5 mutants showed a
cystomanamide production comparable to the wild type strain,
indicating no or only a minor role of orf5 in the biosynthesis.
The compounds were tested in various bioactivity assays
including cytotoxicity against HCT-116 and CHO-K1 cells,
antibacterial tests against various Gram-negative and Gram-
positive bacterial strains, and antifungal assays against Candida
albicans and Mucor hiemalis and HIV-1 inhibition. Until now,
they have not shown biological activity. We continue functional
testing to find the often very specific biological activity of these
natural products and to further evaluate their biological
function.
In summary, we discovered a new family of glycosylated
lipopeptides using a structure-guided approach by LC-SPE-
NMR. The compounds were fully characterized, and a gene
cluster responsible for cystomanamide biosynthesis was
identified. Inactivation of three independent genes in this
cluster completely abolished cystomanamide production in the
mutants, verifying their essential role during biosynthesis.
Antibiot. 1989, 42, 14−7.
(10) Ambrosi, H.-D.; Hartmann, V.; Pistorius, D.; Reissbrodt, R.;
Trowitzsch-Kienast, W. Eur. J. Org. Chem. 1998, 1998, 541−551.
(11) Ogawa, Y.; Konishi, T. Chem. Pharm. Bull. 2009, 57, 1110−2.
(12) Schmidt, Y.; van der Voort, M.; Crusemann, M.; Piel, J.; Josten,
̈
M.; Sahl, H.-G.; Miess, H.; Raaijmakers, J. M.; Gross, H.
ChemBioChem 2014, 15, 259−66.
(13) Harada, K.; Fujii, K.; Mayumi, T.; Hibino, Y.; Suzuki, M.; Ikai,
Y.; Oka, H. Tetrahedron Lett. 1995, 36, 1515−1518.
(14) Gerwick, W. H.; Jiang, Z. D.; Agarwal, S. K.; Farmer, B. T.
Tetrahedron 1992, 48, 2313−2324.
(15) Blin, K.; Medema, M. H.; Kazempour, D.; Fischbach, M. A.;
Breitling, R.; Takano, E.; Weber, T. Nucleic Acids Res. 2013, 41,
W204−12.
(16) Punta, M.; Coggill, P. C.; Eberhardt, R. Y.; Mistry, J.; Tate, J.;
Boursnell, C.; Pang, N.; Forslund, K.; Ceric, G.; Clements, J.; Heger,
A.; Holm, L.; Sonnhammer, E. L. L.; Eddy, S. R.; Bateman, A.; Finn, R.
D. Nucleic Acids Res. 2012, 40, D290−301.
(17) Rottig, M.; Medema, M. H.; Blin, K.; Weber, T.; Rausch, C.;
̈
Kohlbacher, O. Nucleic Acids Res. 2011, 39, W362−W367.
(18) Bachmann, B. O.; Ravel, J. Methods Enzymol. 2009, 458, 181−
217.
(19) Altschul, S. F.; Gish, W.; Miller, W.; Myers, E. W.; Lipman, D. J.
J. Mol. Biol. 1990, 215, 403−10.
(20) Duitman, E. H.; Hamoen, L. W.; Rembold, M.; Venema, G.;
Seitz, H.; Saenger, W.; Bernhard, F.; Reinhardt, R.; Schmidt, M.;
Ullrich, C.; Stein, T.; Leenders, F.; Vater, J. Proc. Natl. Acad. Sci. U.S.A.
1999, 96, 13294−13299.
ASSOCIATED CONTENT
* Supporting Information
■
S
(21) Bode, H. B.; Dickschat, J. S.; Kroppenstedt, R. M.; Schulz, S.;
Muller, R. J. Am. Chem. Soc. 2005, 127, 532−3.
̈
Figures and tables giving configuration analysis, feeding studies,
(22) Silakowski, B.; Nordsiek, G.; Kunze, B.; Blocker, H.; Muller, R.
̈
̈
and inactivation of the ctm cluster in MCy9118 as well as
Chem. Biol. 2001, 8, 59−69.
1
experimental details, H and 13C NMR assignments, and 1D
(23) Aron, Z. D.; Dorrestein, P. C.; Blackhall, J. R.; Kelleher, N. L.;
Walsh, C. T. J. Am. Chem. Soc. 2005, 127, 14986−7.
(24) Stachelhaus, T. J. Biol. Chem. 1998, 273, 22773−22781.
(25) Hegg, E. L.; Que, L. Eur. J. Biochem. 1997, 250, 625−9.
(26) Ryle, M. J.; Koehntop, K. D.; Liu, A.; Que, L.; Hausinger, R. P.
Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 3790−5.
and 2D NMR spectra for 1−4. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
(27) Hausinger, R. P. Crit. Rev. Biochem. Mol. Biol. 2004, 39, 21−68.
(28) Singh, G. M.; Fortin, P. D.; Koglin, A.; Walsh, C. T. Biochemistry
2008, 47, 11310−20.
(29) Koonin, E. V.; Tatusov, R. L. J. Mol. Biol. 1994, 244, 125−32.
(30) Wu, K.; Chung, L.; Revill, W. P.; Katz, L.; Reeves, C. D. Gene
2000, 251, 81−90.
*E-mail: rolf.mueller@helmholtz-hzi.de.
Notes
The authors declare no competing financial interest.
(31) Dorrestein, P. C.; Van Lanen, S. G.; Li, W.; Zhao, C.; Deng, Z.;
Shen, B.; Kelleher, N. L. J. Am. Chem. Soc. 2006, 128, 10386−7.
(32) Zaleta-Rivera, K.; Xu, C.; Yu, F.; Butchko, R. a E.; Proctor, R.
H.; Hidalgo-Lara, M. E.; Raza, A.; Dussault, P. H.; Du, L. Biochemistry
2006, 45, 2561−9.
(33) Lin, S.; Van Lanen, S. G.; Shen, B. Proc. Natl. Acad. Sci. U.S.A.
2009, 106, 4183−8.
ACKNOWLEDGMENTS
■
Research in R.M.’s laboratory was funded by the Bundesmi-
nisterium fur Bildung and Forschung. We thank Kirsten
̈
Harmrolfs for support with LC-SPE-NMR and Thomas and
Michael Hoffmann (all from R.M.’s laboratory) for MS
measurements.
(34) Magarvey, N. A.; Beck, Z. Q.; Golakoti, T.; Ding, Y.; Huber, U.;
Hemscheidt, T. K.; Abelson, D.; Moore, R. E.; Sherman, D. H. ACS
Chem. Biol. 2006, 1, 766−79.
REFERENCES
■
(1) Wenzel, S. C.; Muller, R. Curr. Opin. Drug Discovery Devel. 2009,
̈
(35) Fujimori, D. G.; Hrvatin, S.; Neumann, C. S.; Strieker, M.;
Marahiel, M. A.; Walsh, C. T. Proc. Natl. Acad. Sci. U.S.A. 2007, 104,
16498−503.
12, 220−30.
(2) Weissman, K. J.; Muller, R. Nat. Prod. Rep. 2010, 27, 1276−95.
̈
(3) Fischbach, M. A.; Walsh, C. T. Chem. Rev. 2006, 106, 3468−96.
(4) Walsh, C. T.; Chen, H.; Keating, T. A.; Hubbard, B. K.; Losey, H.
C.; Luo, L.; Marshall, C. G.; Miller, D. A.; Patel, H. M. Curr. Opin.
Chem. Biol. 2001, 5, 525−34.
(36) Xu, Y.; Orozco, R.; Wijeratne, E. M. K.; Gunatilaka, A. A.; Stock,
S. P.; Molnar
(37) Clissold, P. M.; Ponting, C. P. Trends Biochem. Sci. 2001, 26, 7−
9.
́
, I. Chem. Biol. 2008, 15, 898−907.
(5) Weissman, K. J.; Muller, R. Bioorg. Med. Chem. 2009, 17, 2121−
̈
36.
(6) Yamaguchi, H.; Nakayama, Y.; Takeda, K.; Tawara, K.; Maeda,
K.; Takeuchi, T.; Umezawa, H. J. Antibiot. 1957, 10, 195−200.
(7) Bycroft, B. W.; Pinchin, R. J. Chem. Soc., Chem. Commun. 1975,
121.
(8) Zander, W.; Gerth, K.; Mohr, K. I.; Kessler, W.; Jansen, R.;
Muller, R. Chem.Eur. J. 2011, 17, 7875−81.
̈
2417
dx.doi.org/10.1021/ol500779s | Org. Lett. 2014, 16, 2414−2417