Synthesis and Antibacterial Activity of Four Stereoisomers of the Spider-Pathogenic Fungus Metabolite Torrubiellone D

Article abstract of DOI:10.1021/acs.orglett.6b00245

Four stereoisomers of the spider-pathogenic fungus metabolite torrubiellone D were synthesized for the first time in 10% overall yield starting from l-tyrosine or d-tyrosine. The 3-decatrienoyl side chain was assembled and attached via (E)-selective HWE and Wittig olefinations. Their antibiotic activities against drug-susceptible Escherichia coli strains differed considerably.

Full text of DOI:10.1021/acs.orglett.6b00245

Letter
pubs.acs.org/OrgLett
Synthesis and Antibacterial Activity of Four Stereoisomers of the
Spider-Pathogenic Fungus Metabolite Torrubiellone D
Sebastian Bruckner,† Ursula Bilitewski,‡ and Rainer Schobert*,†
†Organic Chemistry Laboratory, University Bayreuth, Universitaetsstrasse 30, 95440 Bayreuth, Germany
‡Helmholtz Centre for Infection Research, Inhoffenstrasse 7, 38124 Braunschweig, Germany
S
* Supporting Information
ABSTRACT: Four stereoisomers of the spider-pathogenic fungus
metabolite torrubiellone D were synthesized for the first time in 10%
overall yield starting from L-tyrosine or D-tyrosine. The 3-decatrienoyl side
chain was assembled and attached via (E)-selective HWE and Wittig
olefinations. Their antibiotic activities against drug-susceptible Escherichia
coli strains differed considerably.
any fungi of the order Hypocreales are pathogenic to
M
insects and feed on them.1 They are also a rich source of
structurally diverse metabolites that may contribute to the
infestation of the host and to the defense of its resources
against competitors.1,2 These metabolites are therefore of
particular interest as potential leads for new drugs and
insecticides. As part of a screening program in Thailand,3
Isaka et al.4 assessed the metabolite profiles of 16 Torrubiella
species,5 the most prolific of which, Torrubiella sp. BCC 2165,
was found to produce four hitherto unknown alkaloids, three 2-
pyridones and a tetramic acid, torrubiellone D (1).
Their structures were elucidated except for the configuration
of the stereocenters. A total synthesis of the pyridone
(+)-torrubiellone C (2) by Gademann et al. proved that the
natural (−)-enantiomer, the presumed metabolic product of the
tetramic acid torrubiellone D (1), has an (R)-configured
stereocenter in the side chain.6 Cursory tests of compounds 1
and 2 on Plasmodium falciparum, Mycobacterium tuberculosis,
and three cancer cell lines were negative.4 We have now
synthesized the four diastereomers 1a−d in order to assign the
stereochemistry of the natural product and also to evaluate their
antibacterial activities (Figure 1).
First, the N,O-bisprotected tetramic acids (S)-5 and (R)-5
were prepared via a previously published general route7 starting
from enantiopure tyrosine as shown exemplarily for (S)-5 in
Scheme 1. L-Tyrosine was Boc-protected to give carbamate (S)-
3 which, in turn, was silylated to afford amino acid derivative
(S)-4. This was cyclized to (S)-5 with Meldrum’s acid using a
modification of Hosseini’s protocol.8
Figure 1. Structures of diastereoisomers of torrubiellone D (1) and of
natural (−)-torrubiellone C (2).
outlined exemplarily for (R)-16 in Scheme 2. Acid (R)-6 was
reduced with LiAlH4 to give alcohol (R)-7, which was
converted to the acetate (R)-8 with acetic anhydride in the
presence of catalytic copper(II) triflate according to a method
by Firouzabadi.10 Oxidative cleavage of the phenyl ring with
NaIO4/RuCl3 gave the carboxylic acid (S)-9 in 56% yield. The
latter was treated with (trimethylsilyl)diazomethane and the
resulting diester was selectively saponificated without prior
purification to afford 2-(hydroxymethyl)butyrate (S)-10 in 98%
over the last two steps.11 Silylation of the hydroxy group
furnished TBS-ether (S)-11, the methoxycarbonyl residue of
which was reduced with DIBAL-H in THF at −78 °C to give
the aldehyde (S)-12 in 59% yield. This aldehyde was then
olefinated with the anion of phosphonate 13, generated with
LiHMDS in THF. The product ethyl dienoate (R)-14, obtained
in 64% yield, was reduced in 95% yield to the alcohol (R)-15
The 3-decatrienoyl side chain of 1 was then attached to the
tetramic acid 5 by first acylating the latter with the cumulated
phosphorus ylide Ph3PCCO to give a 3-acyl ylide which
would be used to olefinate a suitably protected octadienal. By a
similar approach, we previously synthesized ravenic acid.9 The
required octadienal 16 was prepared in both enantiomeric
forms from purchasable enantiopure 2-phenylbutyric acids 6 as
Received: January 25, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.6b00245
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 1. Synthesis of N,O-Bisprotected Tetramic Acid (S)-
5
Scheme 3. Attachment of the Side Chain via a 3-
Acylylidation−Wittig Olefination Sequence
Since all four synthetic stereoisomers of torrubiellone D
showed specific optical rotations which deviated from that
reported by Isaka et al.4 for their natural isolate (Table 1), we
Scheme 2. Synthesis of the Side-Chain Precursor (R)-16
Table 1. (Specific) Optical Rotations (c = 0.12, MeOH)
Isaka4
1a
1b
1c
1d
α
−0.62
−517
−0.63
−525
+0.64
+533
+0.65
+542
23
[α]D
−182
confirmed their stereochemical identity and purity by analytical
HPLC on a chiral Phenomenex Lux Amylose-1 column, in
comparison to authentic diastereomeric mixtures. Figure 2
shows this for the (5S)-torrubiellones D 1a and 1b and a
mixture of these synthesized from racemic aldehyde 16.
As the topmost chromatogram, recorded of the diastereo-
meric mixture of (5S)-torrubiellones D, turned out to be an
overlay of the chromatograms recorded of the pure synthetic
(5S)-diastereomers 1a and 1b, we can rule out a side-chain
racemization during the synthesis of the four stereoisomers.
The additional peaks at earlier retention times in the
chromatograms of 1a and 1b are additive in the chromatogram
of the diastereomeric mixture and thus are very likely not
impurities but tautomers or rotamers with respect to the C3−
C7 bond of the 3-acyltetramic acid moiety. This assumption is
also supported by the fact that all peaks showed the same
characteristic UV absorption.
The optical rotation of Isaka’s natural isolate deviates
significantly from those of our pure synthetic stereoisomers.
Optical rotations of 3-acyltetramic acids depend decisively on
the solvent12−14 and on the age of the sample solutions since
these parameters govern the ratio of tautomers and rotamers
whose individual specific optical rotations may vary consid-
erably. Hence, it is hard to tell whether Isaka’s natural isolate
contained impurities, artifacts, several stereoisomers, or merely
a different combination of tautomers or rotamers of one
particular of the four possible stereoisomers. It is also worth
noting that the configuration of the stereogenic center in the
side chain has virtually no influence on the magnitude of the
specific optical rotations of the four stereoisomers. Moreover,
they all gave rise to virtually identical NMR spectra which are
also congruent to the NMR data published by Isaka. The
optical rotation of −182 quoted for his natural product isolate
would best agree with a mixture of (5S)- and (5R)-
stereoisomers since racemization at C5 is a well-known aspect
of tetramic acid chemistry.
with DIBAL-H in dichloromethane. (R)-15 was oxidized with
MnO2 to the dienal (R)-16 almost quantitatively.
Finally, in a sequence of three consecutive reactions in one
pot, this aldehyde was converted to the bisprotected tetramic
acid (5S,14R)-18b which gave the torrubiellone D isomer
(−)-(5S,14R)-1b in 55% overall yield after deprotection
(Scheme 3). First, tetramic acid (S)-5 was 3-acylated with
Ph3PCCO to afford the acyl ylide (S)-17 which was
deprotonated right away with potassium tert-butoxide to give
a Wittig-active species of hitherto unknown structure. This, in
turn, was treated with aldehyde (R)-16.9 The resulting mixture
was heated at reflux to afford compound (5S,14R)-18b as the
product of an (E)-selective Wittig alkenation. It was
deprotected stepwise, first with trifluoroacetic acid in dichloro-
methane and then with the same reagent in a methanol/water
mixture to afford the target compound (−)-(5S,14R)-1b. The
other three stereoisomers 1a, 1c, and 1d were prepared
analogously (cf. the Supporting Information).
B
DOI: 10.1021/acs.orglett.6b00245
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
a
Table 2. IC50 Values (μg/mL) of 1a−d for Various Bacteria
1a
1b
1c
1d
S. aureus
37
40
53
38
44
49
55
39
E. faecium
E. coli K12
E. coli ΔTolC
E. coli D21f2
>100
83
>100
30
>100
13
>100
14
62
41
37
39
a
S. aureus: Gram-positive. E. faecium: Gram-positive. E. coli K12: wild-
type, Gram-negative. E. coli ΔTolC: mutant lacking the ArcAB−TolC
efflux system. E. coli D21f2: supersusceptible mutant with truncated
lipopolysaccharide core.15
The (S,S)-isomer 1a was least efficacious against both E. coli
mutants.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.6b00245.
Experimental details of chemical syntheses and biological
tests, characterizations, and NMR spectra of new
compounds (PDF)
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: rainer.schobert@uni-bayreuth.de. Fax: +49 (0)921
552671. Tel: +49 (0)921 552679.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
Figure 2. HPLC chromatograms for (5S)-torrubiellones D. (Top)
Diastereomeric mixture of 1a and 1b; (middle) pure 1a; (bottom)
pure 1b (Phenomenex Lux Amylose-1 100 × 4.6 mm chiral column,
mobile phase 40% n-hexane, 60% ethanol with 0.1% TFA, flow rate 1
mL/min).
■
We thank Yasmin Wenzel and Max Niehage (Helmholtz
Centre for Infection Research, Braunschweig, D) for their
technical assistance with the microbiological tests.
DEDICATION
■
This paper is dedicated to Professor Steven Victor Ley
(University of Cambridge) on the occasion of his 70th birthday.
The four synthetic stereoisomers 1a−d of torrubiellone D
were finally tested for antibacterial activity against five different
bacteria: the Gram-positive strains Staphylococcus aureus
(DSM346) and Enterococcus faecium (DSM20477) and the
Gram-negative strains Escherichia coli K12 wild-type, Escherichia
coli ΔTolC mutant (JW5503), which lacks the ArcAB−TolC
efflux system, and Escherichia coli D21f2 with truncated
lipopolysaccharide (LPS) core (cf. the Supporting Information
for experimental details). The four isomers displayed only weak
activity against the Gram-positive bacteria with little variance
between the compounds and the two strains. The S. aureus was
slightly more susceptible to the (14S)-isomers 1a and 1c (Table
2). A more nuanced picture emerged from the tests with the
Gram-negative E. coli strains. Wild-type E. coli K12 was not
susceptible to any of the compounds, which was obviously due
to insufficient penetration through the outer LPS layer and to
efficient drug efflux pumps of the ArcAB−TolC type. The E.
coli mutants which had a truncated LPS layer (D21f2) or lacked
the TolC efflux pump (ΔTolC) were more susceptible than the
K12 wild-type. The (5R)-isomers 1c and 1d gained most
strongly from the absence of efflux pumps and reached IC50
values of ca. 13 μg/mL (i.e., ca. 35 μM) against E. coli ΔTolC.
REFERENCES
■
(1) Molnar, I.; Gibson, D. M.; Krasnoff, S. B. Nat. Prod. Rep. 2010,
27, 1241−1275.
(2) Humber, R. A. J. Invertebr. Pathol. 2008, 98, 262−266.
(3) Isaka, M.; Kittakoop, P.; Kirtikara, K.; Hywel-Jones, N. L.;
Thebtaranonth, Y. Acc. Chem. Res. 2005, 38, 813−823.
(4) Isaka, M.; Chinthanom, P.; Supothina, S.; Tobwor, P.; Hywel-
Jones, N. L. J. Nat. Prod. 2010, 73, 2057−2060.
(5) Johnson, D.; Sung, G.-H.; Hywel-Jones, N. L.; Luangsa-ard, J. J.;
Bischoff, J. F.; Kepler, R. M.; Spatafora, J. W. Mycol. Res. 2009, 113,
279−289.
(6) Jessen, H. J.; Schumacher, A.; Schmid, F.; Pfaltz, A.; Gademann,
K. Org. Lett. 2011, 13, 4368−4370.
(7) Barnickel, B.; Schobert, R. J. Org. Chem. 2010, 75, 6716−6719.
(8) Hosseini, M.; Kringelum, H.; Murray, A.; Tønder, J. E. Org. Lett.
2006, 8, 2103−2106.
(9) Schlenk, A.; Diestel, R.; Sasse, F.; Schobert, R. Chem. - Eur. J.
2010, 16, 2599−2604.
(10) Firouzabadi, H.; Iranpoor, N.; Sobhani, S.; Amoozgar, Z.
Synthesis 2004, 295−297.
(11) Kuo, G.-H.; Zhang, R.; Wang, A.; Deangelis, A. R. PCT Int.
Appl. WO 2005042478 A2, May 12, 2005.
C
DOI: 10.1021/acs.orglett.6b00245
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(12) Steyn, P.; Wessels, P. L. Tetrahedron Lett. 1978, 19, 4707−4710.
(13) Nolte, M. J.; Steyn, P.; Wessels, P. L. J. Chem. Soc., Perkin Trans.
1 1980, 1057−1065.
(14) Berkley, J. V.; Markopoulos, J.; Markopoulou, O. J. Chem. Soc.,
Perkin Trans. 2 1994, 1271−1274.
(15) Vaara, M. Antimicrob. Agents Chemother. 1993, 37, 2255−2260.
D
DOI: 10.1021/acs.orglett.6b00245
Org. Lett. XXXX, XXX, XXX−XXX