Total Synthesis and Stereochemical Revision of the Anti-Tuberculosis Peptaibol Trichoderin A

Article abstract of DOI:10.1021/acs.orglett.6b01886

The first total synthesis of the postulated structure of the aminolipopeptide trichoderin A and its epimer are reported. A late-stage solution phase C-terminal coupling was employed to introduce the C-terminal aminoalcohol moiety. This methodology provides a foundation to prepare analogues of trichoderin A to establish a structure-activity relationship. NMR spectroscopic analysis established that the C-6 position of the 2-amino-6-hydroxy-4-methyl-8-oxodecanoic acid (AHMOD) residue in trichoderin A possesses an (R)-configuration as opposed to the originally proposed (S)-configuration.

Full text of DOI:10.1021/acs.orglett.6b01886

Letter
pubs.acs.org/OrgLett
Total Synthesis and Stereochemical Revision of the Anti-Tuberculosis
Peptaibol Trichoderin A
Iman Kavianinia,†,‡ Lavanya Kunalingam,† Paul W. R. Harris,†,‡ Gregory M. Cook,∥,‡
and Margaret A. Brimble*,†,‡,§
†School of Biological Sciences, The University of Auckland, 3A Symonds Street, Auckland 1010, New Zealand
‡Maurice Wilkins Centre for Molecular Biodiscovery, The University of Auckland, 3A Symonds Street, Auckland 1010, New Zealand
§School of Chemical Sciences, The University of Auckland, 23 Symonds Street, Auckland 1010, New Zealand
∥Department of Microbiology and Immunology, School of Medical Sciences, University of Otago, 720 Cumberland Street, Dunedin
9054, New Zealand
S
* Supporting Information
ABSTRACT: The first total synthesis of the postulated
structure of the aminolipopeptide trichoderin A and its epimer
are reported. A late-stage solution phase C-terminal coupling
was employed to introduce the C-terminal aminoalcohol
moiety. This methodology provides a foundation to prepare
analogues of trichoderin A to establish a structure−activity
relationship. NMR spectroscopic analysis established that the
C-6 position of the 2-amino-6-hydroxy-4-methyl-8-oxodecanoic
acid (AHMOD) residue in trichoderin A possesses an (R)-configuration as opposed to the originally proposed (S)-configuration.
uberculosis (TB) is an infectious bacterial disease caused
Tby Mycobacterium tuberculosis (Mtb), which mainly affects
the respiratory system.1 TB now ranks alongside human
immunodeficiency virus (HIV) as a leading cause of death
worldwide.2 Although several anti-TB drugs have been
identified and developed over the years, TB continues to
cause considerable morbidity and mortality worldwide. Globally
in 2014, an estimated 480,000 people developed multidrug-
resistant TB (MDR-TB).3 Naturally occurring antimicrobial
peptides are a promising source for the development of a new
Figure 1. Originally proposed structure of trichoderin A 1 and its
nonproteinogenic residues MDA 2, AMAE 3, and AHMOD 4.
class of drugs to prevent and treat systemic and topical
infections,4−6 and recent work in our group has focused on the
synthesis of antimicrobial peptides.7−9
Recently, Kobayashi et al.10 isolated a novel family of
and confirmation of its reported structure will enable further
structure−activity relationship (SAR) studies.
Trichoderin A 1 is composed of eight amino acids with an N-
terminal lipid chain, (R)-2-methyldecanoic acid (MDA) (2),
and a C-terminus comprising an amino alcohol moiety, (S)-2-
((2-aminopropyl(methyl)amino)ethanol (AMAE) (3). Struc-
aminolipopeptide antibiotics, the trichoderins A 1, A1 and B,
from a marine sponge-derived fungus trichoderma sp. Notably,
trichoderin A 1 (Figure 1) was found to exhibit a more potent
minimum inhibitory concentration activity (MIC, 0.12 μg/mL)
than the first-line anti-TB drug isoniazid (MIC, >100 μg/mL)
tural analysis revealed that trichoderin A 1 also possesses
against Mtb H37Rv under hypoxic conditions. The anti-
several unnatural amino acids, namely, α-aminoisobutyric acid
mycobacterial activity of the trichoderins has been suggested to
be attributed to the inhibition of ATP synthesis in the
mycobacteria,11 but further mode of action studies are required
to validate this hypothesis.
Since Mtb can lie dormant for years before being activated
leading to TB, the intriguing activity of trichoderin A against
Mtb in its dormant states suggests that trichoderin A can
potentially serve as a new peptide-based lead compound for the
treatment of TB. Thus, an efficient synthesis of trichoderin A 1
(Aib) and the unusual β-hydroxyketone (AHMOD) 4. All
chiral proteinogenic amino acids were reported to be L-amino
acids.10 The C-6 position of the AHMOD residue was also
assumed to exhibit the (S)-configuration by comparison with
the NMR data for structurally similar peptaibol trichopolyn I.10
However, in order to determine the absolute stereochemistry of
Received: June 29, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.6b01886
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
C-6 in trichoderin A, we herein report the first synthesis of the
postulated structure of trichoderin A 1 and its epimer at the C-
6 position of the AHMOD unit.
Scheme 1. Initial Synthetic Strategies Towards Trichoderin
A 1
An Fmoc-based SPPS strategy was envisaged for the
synthesis of the peptaibol framework of trichoderin A due to
the presence of the highly acid-sensitive β-hydroxyketone
AHMOD residue. 2-Chlorotrityl-functionalized aminomethyl
polystyrene resin (2-ClTrt) was also selected for Fmoc-SPPS,
as it is cleavable under mild conditions using hexafluoropropan-
2-ol (HFIP) thus preventing undesired elimination of the β-
hydroxy ketone moiety in the AHMOD residue.12
Incorporation of a (2S,4S,6S)-AHMOD 5 or (2S,4S,6R)-
AHMOD 6 residue into the peptaibol framework of trichoderin
A is required to determine the absolute configuration of C-6 in
the AHMOD residue. Our group recently reported the total
synthesis of the related anticancer peptaibol culicinin D and
confirmed the (6R)-stereochemistry of the hydroxyl group in its
AHMOD unit.12 The synthesis of the two Fmoc-protected
AHMOD residues that are epimeric at the C-6 position were
carried out following our earlier work.13 The synthesis of Fmoc-
protected (S)-AMAE 7 was also achieved using a similar
strategy to that employed for our synthesis of 2-(2-
aminopropylamino)ethanol (APAE) present in culicinin D
(see Supporting Information).12 The N-terminal lipid residue
(R)-MDA 2 was obtained in a three-step procedure using an
Evans chiral auxiliary mediated asymmetric alkylation (see
Supporting Information).14 Having prepared the requisite
building blocks, we proceeded to incorporate these into
Fmoc-SPPS to access trichoderin A 1 and its C-6 AHMOD
epimer 22.
Optimal conditions for the solid phase synthesis of
trichoderin A were investigated by carrying out the synthesis
of an analogue of trichoderin A wherein L-alanine, L-leucine,
and dodecanoic acid were used as substitutes for the
synthetically challenging AMAE 7, AHMOD 5, and MDA 2
building blocks respectively (see Supporting Information). Our
initial attempts using the powerful coupling reagent
tetramethylfluoroformamidinium hexafluorophosphate
(TFFH) to form acyl fluorides in situ, which we had
successfully used for the total synthesis of culicinin D did not
enable coupling of the two consecutive Aib residues in the
simplified trichoderin analogues.12 Several coupling conditions
were therefore evaluated to investigate the difficult coupling of
the sterically hindered Aib residues. A mixture of the uronium-
type coupling reagent 1-[(1-(cyano-2-ethoxy-2-oxoethylidene-
aminooxy)dimethylaminomorpholinomethylene)]-
methanaminium hexafluorophosphate (COMU)15 and 2-
cyano-2-(hydroxyimino)acetate (Oxyma) minimized Aib dele-
tion (see Supporting Information). Having established
conditions to enable formation of the problematic Aib-Aib
unit, two different pathways were investigated for the synthesis
of trichoderin A 1 (Scheme 1).
Our initial approach to trichoderin A 1 (Scheme 1, route A)
began with anchoring the hydroxyl group of Fmoc protected
(S)-AMAE 7 to 2-ClTrt resin resulting in a resin loading of
29% (see Supporting Information). The synthesis of trichoder-
in A was then continued with removal of the Fmoc group on
resin 8 using 20% piperidine in DMF followed by the use of 1-
[bis(dimethylamino)-methylene]-1H-1,2,3-triazolo[4,5-b]-
pyridinium hexafluorophosphate 3-oxide (HATU) and
diisopropylethylamine (DIPEA) to effect coupling of all the
natural amino acids. A combination of COMU/Oxyma/DIPEA
was used for the difficult coupling of the Aib amino acids to the
peptide chain. Fmoc-protected (2S,4S,6S)-AHMOD 5 was
effectively coupled to the growing peptidyl resin using HATU/
DIPEA without hydroxyl group protection (see Supporting
Information). Coupling of the N-terminal MDA residue 2 was
then achieved using COMU/Oxyma/DIPEA. After completion
of the peptide sequence, a solution of HFIP/CH2Cl2 (v/v, 1:4)
was used to release the peptide from the solid support. Analysis
of the crude product by LC-MS only showed the presence of a
trace amount of desired product (see Supporting Information);
therefore this approach (Scheme 1, route A) was deemed to be
inefficient. Thus, an alternative anchoring strategy wherein
Fmoc-protected AMAE 7 was attached to the resin via the
more reactive amino group was investigated (Scheme 1, route
B).
It was anticipated that a final O,N-intramolecular acyl
migration would afford the desired peptaibol.16 Briefly, Fmoc-
AMAE-OH 7 was attached to 2-ClTrt resin using 2% DBU in
DMF to effect concomitant removal of the Fmoc group and
attachment of the resultant amino group17 to afford loaded
resin 9 (see Supporting Information). Fmoc-Aib-OH was next
coupled to the free hydroxyl group on resin using DIC/DMAP.
The Fmoc protecting group was then removed, and the loading
for the resin-bound dipeptide was 37%. It was unclear whether
the low loading was due to the inefficient attachment of AMAE
to 2-ClTrt resin or due to a low yielding esterification between
the Aib and AMAE residue. Despite the low loading, peptide
elongation proceeded in the same manner as route A. After
complete assembly of the desired sequence, the resin-bound
peptide 10 was released from the solid support using a solution
of HFIP/CH2Cl2 (v/v, 1:4). Peptide 11 was next dissolved in
DMF and treated with 10 equiv of DBU to give trichoderin A 1.
B
DOI: 10.1021/acs.orglett.6b01886
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 2. Total Synthesis of Trichoderin A and Its C-6 AHMOD Epimer
LC-MS analysis of the mixture confirmed the presence of the
desired product, but only in a trace amount similar to that
observed for route A (see Supporting Information). The O,N-
intramolecular acyl rearrangement strategy was therefore also
abandoned.
We suspected that the low yield and poor recovery of the full
length peptide chain were due to poor loading of the AMAE
unit to the 2-ClTrt resin or the instability of the peptidyl resin
linkage during the synthesis. In order to investigate the
possibility of premature cleavage of peptide, Fmoc-AMAE-OH
7 was attached to 2-ClTrt resin via the hydroxyl group (route
A) followed by acetylation after each coupling step. Analysis of
the crude product by LC-MS after each coupling step showed
N-acetylated AMAE ([M + H]+: 175.2) which increased with
each coupling step (see Supporting Information). This finding
confirms the instability of the AMAE bond to the first Aib
residue during peptide elongation.
We therefore developed a combined solid-phase and
solution-phase strategy involving preparation of peptide 18
using SPPS, followed by cleavage of the peptide chain from the
support and attachment of the fully unprotected AMAE 3 to
the C-terminal carboxylic group of peptide 20 (Scheme 2). We
envisaged that the amino group of 3 would react faster than the
alcohol with the carboxylic acid of peptide 20. Advantageously,
given that the C-terminal amino acid in peptide 20 is achiral,
racemization is not a concern.
The synthesis of trichoderin A 1 began with formation of
resin 12 by loading of Fmoc-Aib-OH onto 2-ClTrt resin using
DIPEA resulting in a higher loading of 49%. The second Fmoc-
Aib-OH was then coupled to the resin using COMU/Oxyma/
DIPEA effecting the difficult formation of the Aib-Aib unit to
give 13. Elongation of the peptide sequence was continued
using the previously optimized conditions with HATU/DIPEA
as a coupling reagent. The amino acid (2S,4S,6S)-AHMOD 5
was then effectively coupled to resin 15 using HATU/DIPEA
to form 16. Finally, L-proline and fatty acid (2R)-MDA 2 were
sequentially coupled to complete the synthesis of linear
peptidyl resin 18. Treatment of resin-bound peptide 18 with
HFIP/CH2Cl2 (v/v, 1:4), followed by analysis of the cleaved
mixture by LC-MS, confirmed the presence of the desired
peptide 20 as the major product (see Supporting Information).
The lyophilized crude peptide was then purified by semi-
preparative RP-HPLC using 0.1% formic acid to afford peptide
20 in 14% yield. Pleasingly, use of mild formic acid instead of
0.1% TFA avoided elimination of the β-hydroxy ketone to an
enone byproduct.12 The resulting peptide 20 was then
subjected to solution phase C-terminal installation of
unprotected AMAE. Use of HATU, COMU, or PyBOP in
DMF or DIC in CH2Cl2 resulted in minimal C-terminal
addition of AMAE. The coupling efficiency increased
significantly using DIC/6-Cl-HOBt in DMF (see Supporting
Information). Subsequent purification of the crude reaction
mixture by RP-HPLC afforded 7 mg (6.0%) of the postulated
structure of trichoderin A 1 (>97% purity). The synthetic route
to the postulated structure of trichoderin A 1 outlined in
Scheme 2 was adapted for the synthesis of the C-6 AHMOD
epimer of trichoderin A 22 wherein (S)-AHMOD 5 residue was
substituted for (R)-AHMOD 6. Adopting this procedure,
peptide 22 (4 mg, 3.4%) was synthesized from resin 12
(loading 47%) in >97% purity (see Supporting Information).
The absolute stereochemistry at C-6 of the AHMOD residue in
trichoderin A was determined by comparing the deviations of
the 1H and 13C NMR chemical shifts between isolated
trichoderin A and the two synthetic trichoderin A peptides 1
and 22 (see Supporting Information). The deviations of the 1H
and 13C NMR chemical shifts between isolated trichoderin A
and synthetic trichoderin A 1 were larger than the values
obtained between isolated trichoderin A and 22. It was
therefore concluded that the natural product trichoderin A is
the same structure as synthetic peptide 22 where the
configuration at C-6 of the AHMOD residue is assigned as
(R), rather than (S) as reported by Kobayashi et al.10 This is
identical to the stereochemistry exhibited by the AHMOD
residue in culicinin D.12
The two C-6 AHMOD epimers of trichoderin A 1 and 22
were tested for their antimicrobial activity (MIC) against a
range of bacteria (Table 1).18 Though the anti-TB activity
measured for both synthetic trichoderins were less potent than
C
DOI: 10.1021/acs.orglett.6b01886
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Table 1. Minimum Inhibitory Activity (MIC) of Synthetic
Trichoderin A 1 and 22 against Selected Bacterial Pathogens
REFERENCES
(1) Russell, D. G. Nat. Rev. Microbiol. 2007, 5, 39−47.
(2) Barry, C. E.; Boshoff, H. I.; Dartois, V.; Dick, T.; Ehrt, S.; Flynn,
J.; Schnappinger, D.; Wilkinson, R. J.; Young, D. Nat. Rev. Microbiol.
2009, 7, 845−855.
■
MIC (μg/mL)
bacterium
synthetic peptide, 1
synthetic peptide, 22
(3) Zumla, A.; Davies, P. Int. J. Tuberc. Lung Dis. 2016, 20, 285−286.
(4) Cochrane, S. A.; Vederas, J. C. Med. Res. Rev. 2016, 36, 4−31.
(5) Rodriguez, L. M. D. L.; Kaur, H.; Brimble, M. A. Org. Biomol.
Chem. 2016, 14, 1177−1187.
M. tuberculosis
M. smegmatis
S. aureus
9.3
9.3
9.3
9.3
2.3
2.3
E. coli
>595.4
4.6
>595.4
2.3
(6) Reddy, K.; Yedery, R.; Aranha, C. Int. J. Antimicrob. Agents 2004,
24, 536−547.
S. uberis
(7) Hung, K.y.; Harris, P. W.; Heapy, A. M.; Brimble, M. A. Org.
Biomol. Chem. 2011, 9, 236−242.
(8) Brimble, M. A.; Edwards, P. J.; Harris, P. W.; Norris, G. E.;
Patchett, M. L.; Wright, T. H.; Yang, S. H.; Carley, S. E. Chem. - Eur. J.
2015, 21, 3556−3561.
that reported for the natural product, this is possibly due to
peptide 1 and 22 being assessed as the free base as compared to
the natural product which was evaluated as a salt. Importantly,
the anti-TB MIC values for synthetic trichoderins 1 and 22 (9.3
μg/mL) were found to be 10-fold more potent than the first-
line anti-TB drug isoniazid (MIC, >100 μg/mL). Thus, these
results highlight the importance of trichoderins 1 and 22 as
promising new candidates for the treatment of TB.
Despite the biological importance of AHMOD-containing
aminolipopeptides, only a limited number of synthetic studies
have been reported due to the presence of the sensitive
AHMOD residue and the sterically hindered Aib amino acids.
As a consequence of our synthetic studies, stereochemical
assignment of the C-6 position of the AHMOD moiety of
trichoderin A was determined to be (R) as opposed to the
originally proposed (S)-configuration.
Our successful first total synthesis of the postulated structure
of the antituberculosis aminolipopeptide trichoderin A 1 and its
C-6 AHMOD epimer 22 required use of a late-stage solution
phase C-terminal coupling to effectively introduce the (S)-
AMAE moiety 3, and this proved to be the significant factor to
enable an efficient synthesis. This methodology provides a
foundation and framework to prepare further analogues of
trichoderin A in order to construct a structure−activity
relationship. Given the broad-spectrum activity of these
compounds against Gram-positive bacterial pathogens and
mycobacterial species, the mechanism of action of trichoderin A
requires further investigation.
(9) Medini, K.; Harris, P. W.; Menorca, A.; Hards, K.; Cook, G. M.;
Brimble, M. A. Chem. Sci. 2016, 7, 2005−2010.
̀
(10) Pruksakorn, P.; Arai, M.; Kotoku, N.; Vilcheze, C.; Baughn, A.
D.; Moodley, P.; Jacobs, W. R.; Kobayashi, M. Bioorg. Med. Chem. Lett.
2010, 20, 3658−3663.
(11) Pruksakorn, P.; Arai, M.; Liu, L.; Moodley, P.; Jacobs, W. R., Jr.;
Kobayashi, M. Biol. Pharm. Bull. 2011, 34, 1287−1290.
(12) Hung, K.y.; Harris, P. W.; Brimble, M. A. Org. Lett. 2012, 14,
5784−5787.
(13) Ko, K. Y.; Wagner, S.; Yang, S.-H.; Furkert, D. P.; Brimble, M.
A. J. Org. Chem. 2015, 80, 8631−8636.
(14) Darley, D. J.; Butler, D. S.; Prideaux, S. J.; Thornton, T. W.;
Wilson, A. D.; Woodman, T. J.; Threadgill, M. D.; Lloyd, M. D. Org.
Biomol. Chem. 2009, 7, 543−552.
(15) El-Faham, A.; Albericio, F. J. Pept. Sci. 2010, 16, 6−9.
́
(16) Tailhades, J.; Gidel, M. A.; Grossi, B.; Lecaillon, J.; Brunel, L.;
Subra, G.; Martinez, J.; Amblard, M. Angew. Chem., Int. Ed. 2010, 49,
117−120.
(17) Stach, M.; Weidkamp, A. J.; Yang, S. H.; Hung, K. y.; Furkert, D.
P.; Harris, P. W.; Smaill, J. B.; Patterson, A. V.; Brimble, M. A. Eur. J.
Org. Chem. 2015, 2015, 6341−6350.
(18) Dunn, E. A.; Roxburgh, M.; Larsen, L.; Smith, R. A.; McLellan,
A. D.; Heikal, A.; Murphy, M. P.; Cook, G. M. Bioorg. Med. Chem.
2014, 22, 5320−5328.
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.6b01886.
Experimental procedures and spectral data (PDF)
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: m.brimble@auckland.ac.nz.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We acknowledge financial support from the Maurice Wilkins
Centre for Molecular Biodiscovery and The Auckland Medical
Research Foundation (I.K.). The authors would also like to
thank Dr. Louise A. Stubbing for the synthesis of AHMOD and
MDA, Dr. David Rennison for preparation of AMAE, and
Ayana Menorca for antimicrobial compound testing.
D
DOI: 10.1021/acs.orglett.6b01886
Org. Lett. XXXX, XXX, XXX−XXX