Progress towards the total synthesis of trichodermamides A and B: Construction of the oxazine ring moiety

Article abstract of DOI:10.1021/ol062993x

(Chemical Equation Presented) Trichodermamides are modified heterocyclic dipeptides that possess a unique 4H-5,6-dihydro-1,2-oxazine ring. Starting from affordable, easily available (-)-quinic acid, the enantioselective synthesis of this oxazine moiety wa

Full text of DOI:10.1021/ol062993x

ORGANIC
LETTERS
2007
Vol. 9, No. 6
977-980
Progress Towards the Total Synthesis
of Trichodermamides A and B:
Construction of the Oxazine Ring Moiety
Xiaobo Wan, Gabriel Doridot, and Madeleine M. Joullie´*
Chemistry Department, UniVersity of PennsylVania, 231 South 34th Street,
Philadelphia, PennsylVania 19104
mjoullie@sas.upenn.edu
Received December 11, 2006
ABSTRACT
Trichodermamides are modified heterocyclic dipeptides that possess a unique 4H-5,6-dihydro-1,2-oxazine ring. Starting from affordable,
easily available ( )-quinic acid, the enantioselective synthesis of this oxazine moiety was achieved by an intramolecular epoxide ring-opening
−
reaction by an oxime.
Trichodermamides A and B (Figure 1) are modified hetero-
cyclic dipeptides isolated from cultures of the marine-derived
found for the first time in a natural product. Although
trichodermamide A was found to be completely inactive in
biological assays, trichodermamide B displayed significant
in vitro cytotoxicity against HCT-116 human colon carci-
noma with an IC50 of 0.32 µg/mL and moderate antimicrobial
activities against amphoterocin-resistant Candida albicans,
methacillin-resistant Staphylococcus aureus, and vancomy-
cin-resistant Enterococcus faecium with MIC values in the
range of 15 µg/mL. More recently, Capon’s group isolated
aspergillazines A-E, which are structurally related to the
trichodermamides, from an Australian strain of Aspergillus
unilateralis.3 However, the bioactivity of these compounds
was not reported due to insufficient material for testing. To
date, no total synthesis of these compounds has been
reported. Therefore, this synthetically challenging and bio-
logically interesting new class of metabolites led us to
investigate their synthesis.
Figure 1. Trichodermamide family.
fungus Trichoderma Virens in 2003.1 Trichodermamide A
was believed to have the same structure as penicillazine,
which was isolated from a culture of the marine fungus
Penicillium sp. (strain #386).2 Both compounds possess a
unique 4H-5,6-dihydro-1,2-oxazine ring merged with a
highly functionalized cyclohexene ring, a heterocyclic core
Strategically, the most challenging problem is the con-
struction of the 4H-5,6-dihydro-1,2-oxazine ring incorporated
into the highly functionalized cyclohexane moiety. When we
began this study, we found few ways to form the oxazine
ring. The most common approach was the hetero Diels-
Alder reaction between vinylnitroso compounds generated
(1) Garo, E.; Starks, C. M.; Jensen, P. R.; Fenical, W.; Lobkovsky, E.;
Clardy, J. J. Nat. Prod. 2003, 66, 423.
(2) Lin, Y.; Shao, Z.; Jiang, G.; Zhou, S.; Cai, J.; Vrijmoed, L. L. P.;
Jones, E. B. G. Tetrahedron 2000, 56, 9607.
(3) Capon, R. J.; Ratnayake, R.; Stewart, M.; Lacey, E.; Tennant, S.;
Gill, H. J. Org. Biomol. Chem. 2005, 3, 123.
10.1021/ol062993x CCC: $37.00
© 2007 American Chemical Society
Published on Web 02/20/2007

in situ and alkenes with inverse electron demand.4-7 Un-
activated alkenes gave generally poor yields, and although
electron-rich alkenes such as silyl enol ethers reacted more
efficiently, the reaction could not be applied to our system
because the regioselectivity would be reversed and the
tertiary alcohol could not be installed properly. Furthermore,
the enantioselective hetero Diels-Alder reaction of vinyl-
nitroso compounds was not reported, and we wanted to syn-
thesize trichodermamides enantioselectively. Only recently,
the Zakarian group developed a novel strategy to construct
the oxazine ring by a 1,2-oxaza-Cope rearrangement.8 Here
we report our synthetic approach toward the enantioselective
construction of the oxazine ring, which will be further
manipulated to afford trichodermamides A and B.
stereoselectively by the directed epoxidation controlled either
by the tertiary hydroxyl group or by the secondary hydroxyl
group, and the R-functionality would be introduced by the
oxidation of the enolate of compound III. We found that
(-)-quinic acid was an appropriate starting material because
it provided a tertiary hydroxyl functional group with the
desired stereochemistry and a secondary hydroxyl group
which could be further functionalized to the epoxide ring.
In addition, it also had a diol moiety that could be
manipulated by a Corey-Winter olefination.
As shown in Scheme 1, (-)-quinic acid was converted to
the corresponding lactone acetonide, followed by the lactone
Intrigued by the report of an intramolecular epoxide ring-
opening reaction by oximes,9 we envisioned the installation
of the oxazine ring and the secondary alcohol with the desired
stereochemistry by the stereoselective formation of the
epoxide. As shown in Figure 2, we planned to introduce the
Scheme 1. Synthesis of the R-Functionalized Spiroketone
ring-opening reaction to give the ester, and then protection
of the liberated secondary alcohol with TBDPS-Cl afforded
1 in an overall yield of 80% in three steps. Ester 1 was
reduced to the corresponding alcohol using NaBH4/LiCl. The
diol was converted to epoxide 2 in two steps in good overall
yield. The epoxide was then opened in 95% yield using
acetonitrile as the nucleophile and LDA as the base to give
the side chain elongation product 3.12 After treatment with
NaOMe in MeOH, followed by neutralization with acetic
acid, nitrile 3 was converted to the spirolactone 4 in 80%
yield13 (97% yield based on the recovery of starting material).
Conversion of the nitrile to an ester and protection of the
tertiary alcohol were realized in one step. Treatment of 4
with TBAF gave the free alcohol that was converted to a
mesylate. Elimination of the mesylate group under micro-
wave conditions gave the desired product 5 in an overall
70% yield. However, introduction of the azide at the
R-position of lactone 5 could not be accomplished.
Figure 2. Retrosynthetic analysis of the oxazine-containing moiety.
double bond at a later stage by a Corey-Winter olefination,10
and the hydroxyl group at the C5 position could be installed
by allylic oxidation. The key intermediate I would be
synthesized by the epoxide ring-opening reaction. The oxime
could be introduced either by the oxidation of R-amino ester
IIa11 or by the traditional reaction between hydroxylamine
and R-keto ester IIb. The epoxide moiety would be obtained
Evans’ azidation14 failed under all conditions tried. Other
azidation conditions also failed. Bromination only gave a
trace amount of R-bromo lactone. Finally, treatment of
compound 5 with 3 equiv of KHMDS to form the corre-
(4) Gilchrist, T. L.; Lingham, D. A.; Roberts, T. G. J. Chem. Soc., Chem.
Commun. 1979, 1089.
(5) Gilchrist, T. L.; Roberts, T. G. J. Chem. Soc., Perkin Trans. 1 1983,
1283.
(6) Gilchrist, T. L.; Stretch, W. J. Chem. Soc., Perkin Trans. 1 1987,
2235.
(7) Gilchrist, T. L.; Lemos, A. J. Chem. Soc., Perkin Trans. 1 1993,
1391.
(8) Zakarian, A.; Lu, C.-D. J. Am. Chem. Soc. 2006, 128, 5356.
(9) Al-Qawasmeh, R. A.; Al-Tel, T. H.; Abdel-Jalil, R. J.; Voelter, W.
Chem. Lett. 1999, 541.
(11) Boehlow, T. R.; Harburn, J. J.; Spilling, C. D. J. Org. Chem. 2001,
66, 3111.
(12) Takagi, Y. T. I. Bull. Chem. Soc. Jpn. 1993, 66, 2949.
(13) Kim, Y. J.; Kido, M.; Bando, M.; Kitahara, T. Tetrahedron 1997,
53, 7501.
(14) Evans, D. A.; Britton, T. C.; Ellman, J. A.; Dorrow, R. L. J. Org.
Chem. 1990, 112, 4011.
(10) Chiara, J. L.; Valle, N. Tetrahedron: Asymmetry 1995, 6, 1895.
978
Org. Lett., Vol. 9, No. 6, 2007

sponding enolate, followed by oxidation with a molybdenum
pentoxide-pyridine-HMPA complex (MoOPH), gave the
corresponding R-hydroxyl lactone 6 in 70% yield as a
mixture of diastereomers.
Unfortunately, the five-membered spirolactone 6 was very
stable and could not be esterified under several conditions.
Alternatively, after protection of the secondary alcohol with
TBDPS, 6 was opened easily when reduced by LiBH4, with
the migration of the TBDPS group to the primary alcohol to
afford 7 in almost quantitative yield, as shown in Scheme 2.
compound 10 was treated with TCDI, undesired dithiocar-
bonate 11b was also formed with the desired monothiocar-
bonate 11a, which made this approach less favorable.
A model study was then carried out to optimize the
conditions for the directed epoxidation (Scheme 3a). The
Scheme 3. Model Study of the Directed Epoxidation and Its
Application in the Real System
Scheme 2. Attempts to Install the Epoxide with the Desired
Stereochemistry
acetonide moiety on lactone 5 was cleaved, and the liberated
hydroxyl group was used as the directing group. All of the
directed epoxidation methods mentioned above were tried,
but again all failed. However, we were delighted to find that
Our initial strategy was to use the tertiary alcohol at C-4 to
direct the epoxidation; however, compound 7 was inert to
Sharpless epoxidation conditions.15 When treated with mCP-
BA in CH2Cl2, followed by Dess-Martin oxidation, a single
diastereomer 8 was obtained. However, a detailed 1H NMR
study showed that the epoxide was formed anti to the tertiary
alcohol. This stereochemistry was verified when oxidation
with DMDO, which is reported to have no hydroxyl directing
effect, provided the same diastereomer 8. We attribute the
failure of the mCPBA to direct epoxidation to the weak
H-bonding-directing effect of the tertiary alcohol that was
unable to control the approach of the oxidant from the
concave face.
20
with TFAA/90% H2O2 a single diastereomer 13 was
obtained in 90% yield. The stereochemistry was confirmed
by X-ray structure analysis of the corresponding diol-
protected compound 14 (Figure 3). The optimized conditions
One way to circumvent this problem was to remove the
acetonide protecting group and use the secondary alcohol
as the directing group (Scheme 2). The acetonide was cleaved
under acidic conditions to give 9 with retention of the TBDPS
group on the primary alcohol.16 However, Sharpless epoxi-
dation failed again even when more reactive VO(OEt)3/t-
BuOOH was used.17
When treated with mCPBA under various conditions, 9
decomposed and epoxide 10 was formed in only 25% yield.
Other conditions, such as H2WO4/30% H2O218 and MnSO4/
30% H2O2/NaHCO3,19 were also fruitless. Furthermore, when
Figure 3. Crystal structure of the epoxide 14.
were modified by adding Na2HPO4 as the buffer21 and
applied to compound 15, which was synthesized in two steps
from compound 6. The desired epoxide 16 was obtained in
excellent yield (Scheme 3b).
The diol moiety was then reprotected to give compound
17 in quantitative yield (Scheme 4). The reduction of the
(15) Sharpless, K. B.; Michaelson, R. C. J. Am. Chem. Soc. 1973, 95,
6136.
(16) Nicolaou, K. C.; Seitz, S. P.; Pavia, M. R.; Petasis, N. A. J. Org.
Chem. 1979, 44, 4011.
(17) Nicolaou, K. C.; Harrison, S. T. Angew. Chem., Int. Ed. 2006, 45,
3256.
(18) Prat, D.; Left, R. Tetrahedron Lett. 1986, 27, 707.
(19) Lane, B. S.; Vogt, M.; Derose, V. J.; Burgess, K. J. Am. Chem.
Soc. 2002, 124, 11946.
(20) McKittrick, B. A.; Ganem, B. Tetrahedron Lett. 1985, 26, 4895.
(21) Johnson, C. R.; Miller, M. W. J. Org. Chem. 1995, 60, 6674.
Org. Lett., Vol. 9, No. 6, 2007
979

secondary alcohol of 18 was then oxidized to the ketone by
the Dess-Martin reagent in excellent yield to give a single
diastereomer 19. The ketone 19 was converted to the oxime
(trans/cis ) 2:1) in excellent yield when treated with
hydroxylamine. This cis isomer can be partially isomerized
to the trans isomer when heated in EtOH/H2O. The trans
isomer 20a cyclized easily when treated with LDA in THF
to give the desired oxazine 21 in 85% yield.
Scheme 4. Construction of the Oxazine Moiety
In conclusion, starting with affordable, easily available
(-)-quinic acid, we stereoselectively synthesized the oxazine
moiety, which will be further manipulated to complete the
total synthesis of trichodermamides A and B.
Acknowledgment. We thank the NSF (CHEM-0130958)
and the University of Pennsylvania for support of this work.
We also acknowledge additional financial support in the form
of an undergraduate internship (Gabriel Doridot) from Re´gion
Rhoˆne-Alpes, Ecole Normale Supe´rieure, and Wyeth.
lactone in the presence of the epoxide moiety was challenging
as the epoxide would not survive most conditions used for
lactone reduction. After several conditions were tried and
failed, the reduction of the lactone to lactol using NaBH4/
CeCl3 as the reducing reagent22 followed by NaBH4 reduction
successfully solved the problem, as shown in Scheme 4. The
Supporting Information Available: Experimental pro-
cedures and characterization of compounds 17, 19, and 21.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(22) Wipf, P.; Kim, Y.; Fritch, P. C. J. Org. Chem. 1993, 58, 7195.
OL062993X
980
Org. Lett., Vol. 9, No. 6, 2007