Total synthesis of the marine antibiotic pestalone and its surprisingly facile conversion into pestalalactone and pestalachloride A

Article abstract of DOI:10.1002/anie.201003755

Surprise, surprise! The total synthesis of the marine natural product pestalone (1), a highly substituted benzophenone with strong antibiotic acitivity, has provided insight into the surprising tendency of this and related molecules to undergo intramolecular Cannizzaro-Tishchenko-type reactions. Pestalone can be readily converted into pestalachloride A, a strongly antifungal metabolite isolated from an endophytic fungus, by simple treatment with ammonia at pH 8.

Full text of DOI:10.1002/anie.201003755

Communications
DOI: 10.1002/anie.201003755
Natural Products
Total Synthesis of the Marine Antibiotic Pestalone and its Surprisingly
Facile Conversion into Pestalalactone and Pestalachloride A**
ˇ
Nikolay Slavov, Jꢀn Cvengros, Jꢁrg-Martin Neudꢁrfl, and Hans-Gꢂnther Schmalz*
In 2001, Fenical and co-workers reported the isolation and
structure elucidation of pestalone (1), a chlorinated, highly
functionalized benzophenone produced by a marine fungus of
the genus Pestalotia in response to bacterial challenge.[1]
Besides a moderate in vitro cytotoxicity against various
tumor cell lines (mean GI50 = 6.0 mm), pestalone (1) was
reported to exhibit highly potent antibiotic activity against
methicillin-resistant
Staphylococcus
aureus
(MIC =
37 ngmLꢀ1) and vancomycin-resistant Enterococcus faecium
(MIC = 78 ngmLꢀ1). This prompted Fenical to emphasize
that “the potency of this agent toward drug-resistant pathogens
suggests that pestalone should be evaluated in more advanced,
whole animal models of infectious disease.”[1] Consequently,
pestalone (1) is considered a particularly promising molecule
with antibiotic properties.[2]
possible “biosynthetic” relationship of 1 with the pestala-
chlorides.
The strategy behind our synthesis is summarized in
Scheme 1. As a key consideration, we intended to assemble
the benzophenone scaffold from a 2,6-dibromobenzaldehyde
Recently, Che and co-workers identified a few strongly
antifungal metabolites (e.g. 3 and 4) from the plant endo-
phytic fungus Pestalotiopsis adusta which they named the
pestalachlorides.[3] These compounds are structurally closely
related to 1 and, interestingly, were obtained as racemates
indicating a non-enzymatic biosynthesis or a particularly
facile mode of racemization.[3]
Owing to their obvious biological potential, their limited
availability from natural sources, and their challenging
chemical structures,[4] pestalone (1) and its congeners are
interesting target molecules for chemical synthesis. Our first
study in 2003 resulted in the synthesis of deformylpestalone,[5]
and little later a total synthesis of 1 (and its demethylated
derivative 2)[6] was communicated by Nishiyama et al.[7]
Remarkably, no further biological studies have been reported
since then.
We describe herein a highly efficient and practicable
synthesis of 1 which makes it possible for the first time to
prepare substantial amounts of this natural product for
further chemical and biological studies. Moreover, we
report some surprising transformations (including the one-
step transformation of 1 to rac-3) shedding light on the
Scheme 1. Retrosynthetic analysis for pestalone (1).
building block of type A by reaction with a lithiated arene of
type B and subsequent oxidation. The bromo functionalities
should then be used for the introduction of the prenyl and the
formyl side chains at a later stage in the synthesis. The
selective generation of the mono-O-methylated structure was
considered to be possible by appropriate protecting-group
manipulations.[5,7]
To probe the general feasibility of the concept, we first
elaborated
a
synthesis of per-O-methyl-pestalone 12
(Scheme 2). Starting from commercial 5-methylresorcinol
(5), we prepared building block 6 through successive halo-
genation[5] and double O-methylation. As the second building
block, the aldehyde 8 was prepared by bromination of
commercially available 7. When 6 was treated with nBuLi
in 2-Me-THF[8] and the resulting lithiated species (of type B)
was reacted with 8, the desired coupling product rac-9 was
obtained in high yield. Alcohol rac-9 was then oxidized under
Dess–Martin conditions[9] to give the pure crystalline benzo-
phenone 10. The prenyl side chain was introduced by
bromine–lithium exchange using phenyllithium,[10] followed
by addition of CuCN·2LiCl and coupling of the resulting
cuprate intermediate with prenyl bromide. The formylation of
ˇ
[*] N. Slavov, Dr. J. Cvengros, Dr. J.-M. Neudꢀrfl, Prof. Dr. H.-G. Schmalz
Department of Chemistry, University of Cologne
Greinstrasse 4, 50939 Kꢀln (Germany)
Fax: (+49)221-470-3064
E-mail: Schmalz@uni-koeln.de
Homepage: http://www.schmalz.uni-koeln.de
[**] This work was supported by Evangelisches Studienwerk e.V. Villigst
(PhD stipend to N. S.). We gratefully acknowledge generous gifts of
chemicals from Chemetall GmbH and Sanofi-Aventis Deutschland
GmbH. We also thank Prof. Dr. A. Griesbeck for support concerning
the photochemical experiments.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201003755.
7588
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 7588 –7591

Angewandte
Chemie
undergo isomerization (disproportionation) leading to lac-
tones (arylphthalides) of type rac-13 which are related to
pestalachloride A (Scheme 3). Thus, during our initial
attempts to achieve the formylation of 11 through reaction
Scheme 2. Synthesis of per-O-methyl-pestalone (12). Conditions:
a) SO2Cl2, CHCl3/CH3CN (5:1), 3 h, RT; b) Br2, CH3CN, RT, 6 h;
c) NaH, Me2SO4, DMF, RT, 12 h; d) NBS, CH3CN, RT, 6 h; e) nBuLi,
2-Me-THF, ꢀ358C, 30 min, then addition of 8, ꢀ358C!RT, 18 h;
f) DMP, CH2Cl2, RT, 15 h; g) PhLi, THF, ꢀ788C, 30 min, then CuCN·
2LiCl, prenyl bromide; h) nBuLi, MgBr2·OEt2, THF, ꢀ788C, 30 min,
then HCO2Et, ꢀ788C!RT. 2-Me-THF=2-methyltetrahydrofuran,
DMP=Dess–Martin periodinane, NBS=N-bromosuccinimide.
Scheme 3. Top: Unexpected formation of lactones of type rac-1;
bottom: structure of rac-13a in the crystal (C dark gray, H light gray, Cl
green, O red).[19] Conditions: a) nBuLi, MgBr2·OEt2, THF, ꢀ788C,
30 min, then DMF, ꢀ788C!RT, formation of rac-13a detected by
NMR; b) CO/H2 (15 bar), 1.4 mol% Pd(OAc)2, 4 mol% [cataCXium A],
TMEDA, 1008C, 20 h, 51% of rac-13a; c) LiSEt (4 equiv), DMF, 708C,
4 h, 37% rac-13b, 14% rac-13a. cataCXium A=di-1-adamantyl-n-butyl-
phosphane, TMEDA=N,N,N’,N’-tetramethylethylenediamine.
11 proved to be difficult (see below) but was finally achieved
in satisfying yield by reaction of an organomagnesium
intermediate (obtained from 11 with nBuLi and subsequent
transmetalation with MgBr2) with ethyl formate. The struc-
ture of 12 was confirmed by X-ray crystal structure analysis
(Figure 1), which also revealed the strongly twisted confor-
mation of the benzophenone core. This strong twist and the
resulting shielding of the keto function by the ortho-aryl
substituents may also explain its remarkable resistance
towards the organolithium reagents used in the transforma-
tions of 10 and 11.
of the Grignard intermediate with DMF (instead of ethyl
formate) or through Pd-catalyzed hydrocarbonylation
according to Beller et al.,[11] the lactone rac-13a appeared as
the main product. Also, treatment of 12 with LiSEt in DMF
resulted in the formation of rac-13a and its monodemethy-
lated derivative rac-13b.
As a noteworthy fact, we discovered a strong tendency of
the formylated product, that is, the pestalone derivative 12, to
This surpringly facile lactone formation can be rational-
ized in terms of a Cannizzaro–Tishchenko-type reaction[12]
(Scheme 4). We suppose that an intermediate of type 15
(either formed by nucleophilic addition to an aldehyde
precursor of type 14 or as an intermediate during the reaction
of a Grignard precursor with a formyl derivative) undergoes
intramolecular hydride transfer. The resulting alkoxide (16)
then readily cyclizes to the lactone 17.
Figure 1. Structure of per-O-methyl-pestalone 12 (left) and the arylin-
dene 19 (right) in the crystal. C dark gray, H light gray, Cl green,
O red.[19]
Scheme 4. Nucleophile-induced conversion of ortho-formylbenzophe-
nones 14 into arylphthalides 17 in a Cannizzaro–Tishchenko-type
reaction.
Angew. Chem. Int. Ed. 2010, 49, 7588 –7591
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7589

Communications
Having succeeded in synthesizing per-O-methyl-pestalone
protecting-group change at the stage of the dibromobenzo-
phenone 10 (Schemes 2 and 7). First, treatment of 10 with a
slight excess of BBr3 gave rise to the triply demethylated
derivative 25 in high yield.[16] After reprotection of all three
12 (Scheme 2), we tried to remove the protecting groups
(!1). However, while nucleophilic reagents such as LiSEt[13]
could not be employed for the previously mentioned reasons
(lactone formation), treatment of 12 with BBr3 only led to a
[14]
complex mixture (decomposition). With BF3·SMe2
the
mono-demethylated product 18 was obtained (Scheme 5),
Scheme 7. Completion of the total synthesis of pestalone (1). Condi-
tions: a) BBr3 (5.5 equiv), CH2Cl2, RT, 3 h; b) dimethoxymethane, AcCl,
cat. ZnBr2, CH2Cl2, DIPEA, RT 3 h; c) PhLi, THF, ꢀ788C, 30 min, then
CuCN·2LiCl, prenyl bromide; d) nBuLi, LiCl (3 equiv), THF, ꢀ788C,
30 min, then HCO2Et, ꢀ788C!RT; e) HCl (6m), 1,4-dioxane, 638C,
100 min. DIPEA=diisopropylethylamine, MOM=methoxymethyl.
Scheme 5. Mono-demethylation versus arylindane formation on treat-
ment of per-O-methyl-pestalone (12) with BF3. Conditions: a) BF3·SMe2
(1.7 equiv), CH2Cl2/SMe2 (1:2), 08C, 2 h, 40% 18, 20% 19;
b) BF3·SMe2 (6 equiv), CH2Cl2/SMe2 (1:2), ꢀ308C, 75 min; c) BF3·SMe2
(1.7 equiv), CH2Cl2, 08C, 45 min.
however, only if the reagent was used in excess at ꢀ308C in
CH2Cl2/SMe2 (1:2). Interestingly, treatment of 12 with a
smaller excess of the same reagent at 08C in CH2Cl2 afforded
the arylindene 19 in high yield, the structure of which was
again secured by means of X-ray crystallography (Figure 1).
The unexpected and facile formation of the indene
derivative 19 from 12 actually represents a rare and (to the
best of our knowledge) by far the most efficient example of a
metal-free carbonyl–olefin metathesis reaction.[15] A plausible
mechanism is given in Scheme 6. It involves the cyclization of
phenolic functions as MOM ethers,[17] both the prenylation (of
26) and the subsequent formylation (of 27) proceeded
smoothly under the optimized conditions. Fortunately, the
MOM protecting groups were cleaved from 28 in good yield
(with 6m aqueous HCl in dioxane) to give the natural product
pestalone (1), as unambiguously proven by the analytical
data.
This synthesis of 1 requires only 10 steps and proceeds
with an overall yield of 16% starting from commercially
available orcinol (5). Notably, most intermediates are crys-
talline solids and only the last three steps require chromato-
graphic purification. Thus, the synthesis of 1 could be
performed on a multigram scale enabling us to also inves-
tigate some aspects of the reactivity of this interesting
molecule. When a solution of 1 in [D6]DMSO was irradiated
with UV light (350 nm) a new product was formed in a clean
(albeit rather slow) process. To our surprise, this photo-
product turned out to be lactone rac-30, which could be easily
assigned by NMR in comparison to the Cannizzaro–Tish-
chenko product rac-13b. A plausible mechanism for this
remarkable transformation (Scheme 8) involves cyclization of
the primary photo-enol 31 to give the isobenzofurane 32.
Finally, tautomerization leads to the isolated product rac-30,
which we named pestalalactone.
Intrigued by the almost voluntary formation of 30 and its
close structural relationship to pestalachloride A (3), we also
probed the possibility of converting pestalone (1) directly into
3. And indeed, when 1 was treated with NH3 in aqueous
NH4Cl (pH 8.0), pestalachloride A (3) was formed as the only
major product (Scheme 8). We assume that cyclic imino-
hemiaminal (33) forms initially,[18] which then tautomerizes
via the isoindol 34 to the product (rac-3).
In conclusion, we have elaborated a short and efficient
total synthesis of pestalone (1), its antifungal relative
pestalachloride A (3), and some structural analogues, which
are now available for detailed biological studies. Moreover,
Scheme 6. Possible mechanism for the BF3-catalyzed, metal-free car-
bonyl–olefin metathesis of ortho-prenyl benzophenones.
20 to form the tertiary carbenium ion 21, which isomerizes to
the benzylic cation 23 via an oxetane intermediate 22. Finally,
the product 24 is liberated together with acetone in an
entropically favored fragmentation step.
Considering the pitfalls associated with the use of either
nucleophilic or Lewis acidic reagents in the end game, the
total synthesis of pestalone was successfully completed after a
7590 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 7588 –7591

Angewandte
Chemie
context is the synthesis of mumbaistatin, see a) L.
Vertesy, M. Kurz, E. F. Paulus, D. Schummer, P.
Hammann, J. Antibiot. 2001, 54, 354 – 363; b) F. Kaiser,
L. Schwink, J. Velder, H.-G. Schmalz, J. Org. Chem.
2002, 67, 9248 – 9256; c) K. Krohn, J. Diederichs, M.
Riaz, Tetrahedron 2006, 62, 1223 – 1230; d) D. Sucunza,
D. Dembkowski, S. Neufeind, J. Velder, J. Lex, H.-G.
Schmalz, Synlett 2007, 2569 – 2573; e) T. S. Lee, A. Das,
C. Khosla, Bioorg. Med. Chem. 2007, 15, 5207 – 5218.
[5] F. Kaiser, H.-G. Schmalz, Tetrahedron 2003, 59, 7345 –
7355.
[6] Compound 2, also called SB87-Cl 1, is an inhibitor of
testosterone-5a-reductase, see: Y. Wachi, T. Yamashita,
K. Komatsu, S. Yoshida, JP Patent JKXXAF JP
07061950A2 19950307, 1995.
[7] D. Iijima, D. Tanaka, M. Hamada, T. Ogamino, Y.
Ishikama, S. Nishiyama, Tetrahedron Lett. 2004, 45,
5469 – 5471.
[8] D. F. Aycock, Org. Process Res. Dev. 2007, 11, 156 – 159.
[9] a) D. B. Dess, J. C. Martin, J. Org. Chem. 1983, 48, 4155 –
4156; b) D. B. Dess, J. C. Martin, J. Am. Chem. Soc.
1991, 113, 7277 – 7287.
[10] G. Wittig, U. Pockels, Ber. Dtsch. Chem. Ges. 1939, 72,
89.
Scheme 8. Conversion of pestalone (1) into pestalalactone (30) or pestalachlor-
ide A (3). Conditions: a) hn (350 nm), DMSO, RT, 6 d; b) NH3/NH4Cl, H2O, 1,4-
dioxane, RT, 80 min.
[11] a) S. Klaus, H. Neumann, A. Zapf, D. Strꢀbing, S.
Hꢀbner, J. Almena, T. Riermeier, P. Groß, M. Sarich, W.-R.
Krahnert, K. Rossen, M. Beller, Angew. Chem. 2006, 118, 161 –
165; Angew. Chem. Int. Ed. 2006, 45, 154 – 158; b) A. Brenn-
fꢀhrer, H. Neumann, S. Klaus, T. Riermeier, J. Almena, M.
Beller, Tetrahedron 2007, 63, 6252 – 6258.
our work has revealed some unique reactivities of pestalone-
type compounds (such as nucleophile- or photoinduced
Cannizzaro–Tishchenko-type reactions and metal-free car-
bonyl–olefin metathesis), which exploit the intense interac-
tion of functional substituents at the twisted benzophenone
core of 1 (or 12). The surprisingly facile conversion of 1 into
rac-3 under almost neutral conditions might explain the
formation of the racemic natural product rac-3 in nature
(under non-enzymatic conditions) and might also be of
relevance for understanding the molecular mechanism of
action of 1 as an antibiotic compound.
[12] a) S. Cannizzaro, Justus Liebigs Ann. Chem. 1853, 88, 129;
b) V. E. Tishchenko, J. Russ. Phys. Chem. Soc. 1906, 38, 355; for a
review, see: c) O. P. Tꢁrmꢂkangas, A. M. P. Koskinen, Recent
Res. Dev. Org. Chem. 2001, 225; see also: d) L. Cronin, F.
Manoni, C. J. Oꢃ Connor, S. J. Connon, Angew. Chem. 2010, 122,
3109 – 3112; Angew. Chem. Int. Ed. 2010, 49, 3045 – 3048.
ˇ
[13] J. Cvengros, S. Neufeind, A. Becker, H.-G. Schmalz, Synlett 2008,
1993 – 1998.
[14] M. T. Konieczny, G. Maciejewski, W. Konieczny, Synthesis 2005,
1575 – 1577.
Received: June 19, 2010
[15] Compare: a) A. C. Jackson, B. E. Goldman, B. B. Snider, J. Org.
Chem. 1984, 49, 3988 – 3994; b) H. J. Carless, H. S. Trivedi, J.
Chem. Soc. Chem. Commun. 1979, 382 – 383.
[16] For the use of BBr3 for the cleavage of ArOMe ethers, see:
a) J. F. W. McOmie, D. E. West, Org. Synth. Coll. Vol. V 1973,
412; the selective formation of the triply
demethylated product 25 can be understood
in terms of the effective shielding of the
intact methoxy group in an intermediate of
type 29.
[17] M. A. Berliner, K. Belecki, J. Org. Chem.
2005, 70, 9618 – 9621.
[18] J. W. Bode, K. Suzuki, Tetrahedron Lett.
2003, 44, 3559 – 3563; J. W. Bode, K. Suzuki,
Tetrahedron Lett. 2003, 44, 5557.
Published online: August 31, 2010
Keywords: antibiotics · formylation · metathesis ·
.
natural products · photoenolization
[1] M. Cueto, P. R. Jensen, C. Kauffman, W. Fenical, E. Lobkovsky,
J. Clardy, J. Nat. Prod. 2001, 64, 1444 – 1446.
[2] a) H. Rahman, B. Austin, W. J. Mitchell, P. C. Morris, D. J.
Jamieson, D. R. Adams, A. M. Spragg, M. Schweizer, Mar. Drugs
2010, 8, 498 – 518; b) F. A. Villa, L. Gerwick, Immunopharmacol.
Immunotoxicol. 2010, 32, 228 – 237; c) R. K. Pettit, Appl. Micro-
biol. Biotechnol. 2009, 83, 19; d) K. Scherlach, C. Hertweck, Org.
Biomol. Chem. 2009, 7, 1753 – 1760; e) M. Donia, M. T. Hamann,
Lancet Infect. Dis. 2003, 3, 338 – 348.
[3] E. Li, L. Jiang, L. Guo, H Zhang, Y. Che, Bioorg. Med. Chem.
2008, 16, 7894 – 7899; pestalachloride A appeared in the NMR
spectrum as a mixture of two atropdiastereomers (because of
hindered rotation); X-ray crystal structure analysis proved it to
be racemic.
[19] CCDC 781109 (rac-13a), CCDC 781110
(12), and CCDC 781111 (19) contain the supplementary crys-
tallographic data for this paper. These data can be obtained free
of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
[4] The synthesis of highly functionalized tetra-ortho-substituted
benzophenones is challenging. A still unsolved problem in this
Angew. Chem. Int. Ed. 2010, 49, 7588 –7591
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7591