Synthesis of 6,6′-binaphthopyran-2-one natural products: Pigmentosin A, talaroderxines A and B

Article abstract of DOI:10.1021/ol301743t

Efficient and stereoselective syntheses of pigmentosin A, talaroderxine A, and its diastereomer talaroderxine B are reported. The binaphthyl ring system is assembled by vanadium-catalyzed phenolic coupling of tricyclic precursors. These key intermediates were prepared by Michael-Dieckmann annulation of a protected orsellinate ester, with the requisite pyranones accessed by a new variant of Ghosez's sulfone-epoxide annulation. Preliminary biological experiments are reported for pigmentosin.

Full text of DOI:10.1021/ol301743t

ORGANIC
LETTERS
Synthesis of 6,60-Binaphthopyran-2-one
Natural Products: Pigmentosin A,
Talaroderxines A and B
2012
Vol. 14, No. 17
4338–4341
Charles I. Grove, Michael J. Di Maso, Firoz A. Jaipuri, Michelle B. Kim, and
Jared T. Shaw*
Department of Chemistry, One Shields Avenue, University of California, Davis,
California 95616, United States
jtshaw@ucdavis.edu
Received June 25, 2012
ABSTRACT
Efficient and stereoselective syntheses of pigmentosin A, talaroderxine A, and its diastereomer talaroderxine B are reported. The binaphthyl ring
system is assembled by vanadium-catalyzed phenolic coupling of tricyclic precursors. These key intermediates were prepared by
MichaelꢀDieckmann annulation of a protected orsellinate ester, with the requisite pyranones accessed by a new variant of Ghosez’s sulfone-
epoxide annulation. Preliminary biological experiments are reported for pigmentosin.
6,60-Binaphthopyranones form a small class of biologi-
cally interesting secondary metabolites. Pigmentosin A
and talaroderxines A and B (Figure 1) are biosyntheti-
cally related natural products isolated from the lichen
Hypotrachyna immaculata1 and fungus Talaromyces derxii,2
respectively. These compounds, along with viriditoxin3
and asteromine,4 likely result from the oxidative dimeriza-
tion of condensed polyketide precursors. Our interest in
developing general synthetic routes to these compounds
stems from their interesting biological activity relevant to
fighting bacterial diseases. Specifically, viriditoxin is re-
ported to inhibit the bacterial cell division protein FtsZ,5
and the talaroderxines were recently reported to inhibit
botulinum neurotoxin serotype A (BoNT/A), one of the
causative agents in botulism paralysis.6
We recently devised an efficient strategy for assembling
6,60-binaphthopyranones that culminated in the stereose-
lective synthesis of viriditoxin.7 The key biaryl bond was
formed by an oxidative coupling of an orthogonally
protected 7-hydroxy naphthopyranone.8 The stereochemi-
cal outcome of this reaction was subject to complete
control by choice of the vanadium catalyst that was em-
ployed.9 We sought to apply this strategy to the synthesis
(6) Cardellina, J. H., II; Roxas-Duncan, V. I.; Montgomery, V.;
Eccard, V.; Campbell, Y.; Hu, X.; Khavrutskii, I.; Tawa, G. J.; Wallqvist,
A.; Gloer, J. B.; Phatak, N. L.; Holler, U.; Soman, A. G.; Joshi, B. K.; Hein,
S. M.; Wicklow, D. T.; Smith, L. A. ACS Med. Chem. Lett. 2012, 3,
387–391.
(1) Elix, J. A.; Wardlaw, J. H. Aust. J. Chem. 2004, 57, 681–683.
(2) Suzuki, K.; Nozawa, K.; Nakajima, S.; Udagawa, S.; Kawai, K.
Chem. Pharm. Bull. 1992, 40, 1116–1119.
ꢀ
ꢀ
(3) (a) Weisleder, D.; Lillehoj, E. B. Tetrahedron Lett. 1971, 12, 4705–
4706. (b) Suzuki, K.; Nozawa, K.; Nakajima, S.; Kawai, K. Chem.
Pharm. Bull. 1990, 38, 3180–3181.
(4) Arone, A.; Assante, G.; Montorsi, M.; Nasini, G. Phytochemistry
1995, 38, 595–597.
(5) Wang, J.; Galgoci, A.; Kodali, S.; Herath, K. B.; Jayasuriya, H.;
Dorso, K.; Vicente, F.; Gonzalez, A.; Cully, D.; Bramhill, D.; Singh, S.
J. Biol. Chem. 2003, 278, 44424–44428.
(7) (a) Park, Y. S.; Grove, C. I.; Gonzalez-Lopez, M.; Urgaonkar, S.;
Fettinger, J. C.; Shaw, J. T. Angew. Chem., Int. Ed. 2011, 50, 3730–3733.
(b) Grove, C. I.; Fettinger, J. C.; Shaw, J. T. Synthesis 2012, 44, 362–371.
(8) For a comprehensive review on strategies for the synthesis of
axially chiral compounds, see:Bringmann, G.; Gulder, T.; Gulder,
T. A. M.; Breuning, M. Chem. Rev. 2011, 111, 563–639.
(9) Guo, Q.-X.; Wu, Z.-J.; Luo, Z.-B.; Liu, Q.-Z.; Ye, J.-L.; Luo, S.-W.;
Cun, L.-F.; Gong, L.-Z. J. Am. Chem. Soc. 2007, 129, 13927–13938.
r
10.1021/ol301743t
Published on Web 08/13/2012
2012 American Chemical Society

of pigmentosin A and the related talaroderxines with the
ultimate goal of accessing natural and synthetic com-
pounds that might also inhibit the bacterial cell division
protein FtsZ. Herein we describe a new reagent for the
rapid assembly of pyranones related to 6 and their sub-
sequent use in the first syntheses of pigmentosin A, talar-
oderxine A, and talaroderxine B.
Synthesis of the two tricyclic cores required two key
fragments (Figure 1). The orthogonally protected orselli-
nic methyl ester 5 was used in the viriditoxin synthesis and
is available in three steps from methyl acetoacetate. The
second fragment is a C-3 chiral R,β-unsaturated lactone,
which could be accesseda varietyofways. For pigmentosin
A, 6a could be made by elongation of commercially
available hydroxy acid 8 and cyclization. A similar route
was not available for the talaroderxines. We anticipated
employing a sequence that would start with alcohol 10,
available from enantioselective allylation of n-butyraldehyde.
An attractive alternative that would provide the most
general access to the requisite pyranones involved using
acetate,10 then reduced, and cyclized with concomitant
dehydration by acid catalysis (Figure 2). Initially 6b
was accessed via an enantioselective Keck allylation11 of
n-butyraldehyde to provide homoallylic alcohol 10 followed
by acryloylation and ring closure using Grubbs’ second
generation catalyst. Although the synthesis of 6b was
concise and stereoselective, we were interested in a se-
quenceavoiding alkyltinreagentsthatcould be generalized
for the synthesis of analogs.
Figure 2. Synthesis of lactones 6a and 6b.
We envisioned using trimethyl-3-phenylsulfonyl ortho-
propionate 11employed by Ghoseztoaccesslactones from
enantiopureepoxides.12 This attractive route can, in theory,
provide a one-pot synthesis of unsaturated lactones related
to 6 from any epoxide. That said, we experienced extreme
difficulty in the preparation of sulfone 11, consistent with
other studies of this compound subsequent to the original
report.13 In order to circumvent the liabilities of the
original route, we designed a new reagent, namely 2,6,7-
trioxabicyclo[2.2.2]octane (OBO) sulfone ester 15. Treat-
ment of maleic anhydride with benzene sulfinic acid so-
dium salt in refluxing aqueous acetic acid provided 13 in
good yield.14 Esterification of the free acid with 16 fol-
lowed by Lewis acid catalyzed rearrangement provided 15
(Scheme 1A) ingood yield. Sulfone15isa white, crystalline
Scheme 1. Synthesis of OBO-Sulfone 14 and Assembly of 6b
Figure 1. Retrosynthesis of pigmentosin A shown. Compounds
2 and 3 were omitted for clarity.
sulfone 11 in an annulation route reported by Ghosez. We
ultimately examined both routes and discovered a general
solution to the drawbacks associated with the preparation
and handling of 11.
Lactones 6a and 6b were prepared in short sequences.
Commercially available methyl-(R)-3-hydroxybutyrate
8 was elongated with the lithium enolate of tert-butyl
(10) Drochner, D.; Muller, M. Eur. J. Org. Chem. 2001, 211–215.
Org. Lett., Vol. 14, No. 17, 2012
4339

solid that is bench-stable and easily prepared in multigram
quantities. Treatment of epoxide 1215,16 with the lithium
anion of 15 produced 6b (Scheme 1B) in 60% overall yield
after hydrolysis, lactonization, and elimination. Under opti-
mal conditions, 1.3 equiv of 15 was employed whereas 3
equiv of 11 were often employed in related transformations.
The naphthopyranones needed for the key phenolic
coupling reaction were prepared in short sequences from
6a and 6b (Scheme 2). MichaelꢀDieckmann annulation
with the lithium enolate of 57a proceeded smoothly with 6a
4b with the optimal catalyst for 4a resulted in a disappoint-
ingly low ratio of 76:24 selectivity (entry 6) favoring (Ra)-19,
i.e. the isomer that would eventually lead to talaroderxine
B. Changing the catalyst substituent from cyclo-hexyl to
tert-butyl boosted the selectivity up to 86:14. A similar
trend was observed for the production of (Sa)-19, with the
enantiomeric catalyst (Ra,S)-20d producing a nearly com-
plete reversal of selectivity (14:86). The configuration of the
products is assigned based on analogy to our previous studies.
Table 1. Oxidative Coupling of Naphthopyranones Stereo-
chemistry at C-3 Shown as R for Clarity
Scheme 2. Synthesis of Naphthopyranones 4a and 4b
and 6b to furnish 17a and 17b, respectively after oxidation.
We have previously observed higher yields from this two-
step process when compared to the one-step Stauntonꢀ
Weinreb conditions employing β-alkoxy pyranones.17
Naphthopyranones 4a and 4b underwent atrop-selective
phenolic coupling reactions that were controlled by the
structure of the catalyst. VO(acac)2 catalyzed coupling of
4a proceeded in good yield and with little diastereoselec-
tivity. Use of Gong-type9 catalysts produced (Ra)-19 with
high atrop-selectivity that topped out at 94:6 with catalyst
(Sa,R)-20c, which is derived from cyclohexylglycine. The
couplings of 4b were generally less selective. Treatment of
a Reactions conducted in DCM at 0.03 M for 16 h. b Diasteromeric
ratios were determined by HPLC, and the product configuration
(Ra or Sa) is based on previous studies and assigned by CD. c Yields of
isolated product.
Completion of the syntheses of 1ꢀ3 from 18 and 19 was
straightforward. Methylation of 18 followed by global
deprotection with BCl3 produced pigmentosin A (1) in
48% yield over two steps (Scheme 3). Talaroderxines A
and B were completed by a global deprotection of (Ra)-19
and (Sa)-19, which provided the two atropisomers in 34%
and 14% yield, respectively, after HPLC separation.
Synthetic pigmentosin A exhibits physical properties
that are in good agreement when compared to a natural
sample. The 1H and 13C NMR spectra for synthetic
pigmentosin A were identical to what was reported for
the natural sample. The optical rotation values for natural
pigmentosin A is surprisingly small at ꢀ7.1, and the value
recorded for the synthetic sample ꢀ20.8 compared favor-
ably. Finally, a sample was sent to Prof. Elix who demon-
strated that coinjected samples of natural extract and
synthetic pigmentosin A had identical retention times by
HPLC analysis.18
(11) Keck, G. E.; Krishnamurthy, D. Org. Synth. 1998, 75, 12–18.
(12) De Lombaert, S.; Nemery, I.; Roekens, B.; Carretero, J. C.;
Kimmel, T.; Ghosez, L. Tetrahedron Lett. 1986, 27, 5099–5102.
(13) Craig, D.; Lu, P.; Mathie, T.; Tholen, N. T. H. Tetrahedron 2010,
66, 6376–6382.
(14) This reaction was observed during the study of conjugate addi-
tion reactions of maleic anhydrides for a different project in the
laboratory. For a related base-mediated process, see: Wunderlich, K.;
Harms, W. ; Herd, K. J.; Jager, H. Triphendioxazine Dyestuffs. U.S.
Patent 4,665,179, May 12, 1987.
(15) Prepared in three steps from D-norvaline. Margot, C.; Simmons,
D. P.; Reichlin, D.; Skuy, D. Helv. Chim. Acta 2004, 87, 2662–2684.
(16) Epoxide 12 can also be prepared from (S)-epichlorohydrin. See:
Holub, N.; Neidhoefer, J.; Blechert, S. Org. Lett. 2005, 7, 1227–1229.
(17) (a) Evans, G. E.; Leeper, F. J.; Murphy, J. A.; Staunton, J. Chem.
Commun. 1979, 205–206. (b) Dodd, J. H.; Weinreb, S. M. Tetrahedron
Lett. 1979, 20, 3593–3596. (c) Tan, N. P. H.; Donner, C. D. Tetrahedron
2009, 65, 4007–4012.
(18) See Supporting Information.
4340
Org. Lett., Vol. 14, No. 17, 2012

that rotation values for the talaroderxines varied signifi-
cantly in magnitude, but not in sign, based on the source
of the methanol used to prepare the solutions. The con-
vincing NMR, CD, and HPLC data all support the
identical structures of natural and synthetic talaroderxines
A and B.18
Scheme 3. Completion of Natural Products
Preliminary biological studies of pigmentosin A are
encouraging. This compound inhibits the growth of
Bacillus subtilis with an MIC of 20 μM, which is slightly
higher than that of a 1:1 mixture of talaroderxines A and B,
as reported by Suzuki.3b Given that our original interest in
these molecules stemmed from its structural similarity to
viriditoxin, the mammalian toxicity was evaluated. Viridi-
toxin, whose structure differs from pigmentosin only by
replacement of the methyl groups on the heterocycle with
CH2CO2CH3, exhibits a low LD50 of 2.8 mg/kg. Pigmen-
tosin was administered to mice in increasing doses with
no significant difference from control groups based on
body weight and behavior up to 30 mg/kg. Although a true
LD50 was not measured for pigmentosin A, this compound
is safely described as significantly less toxic to mice than
viriditoxin.
In summary we have completed the first syntheses of
pigmentosin A, talaroderxine A, and talaroderxine B.
Central to the success of the latter two compounds was
the development of a new annulation reagent that enables
rapid assembly of unsaturated pyranones from epoxides.
Preliminary biological data are encouraging in that the low
mammalian toxicity of pigmentosin A suggests that these
structures may warrant further investigation as biological
probes.
Synthetic samples of talaroderxines A and B agreed with
1
most of the analytical data for the natural materials. H
and 13C NMR spectra (CDCl3) matched closely with those
reported by Suzuki (d6-DMSO) and with spectra from a
more recent isolation by the Gloer group, also recorded in
CDCl3.6 That said, the chemical shifts in both experiments
are very similar for the diastereomeric natural products. A
mixed sample of synthetic and natural talaroderxine B
produced no additional peaks among the keyaromatic and
hydroxyl signals in the 1H NMR spectrum. Furthermore,
HPLC analysis showed that the retention times were
identical when injected separately or analyzed as a coin-
jection. CD spectra are nearly equal and opposite in their
shape and compare favorably with data reported by
Suzuki. The only discrepancy arises in the optical rotation
values. Suzuki reports optical rotations of ꢀ75.4
and ꢀ86.8 for talaroderxines A and B, respectively. Synthetic
samples showed þ67.1 and ꢀ71.8, which is consistent with
many examples of atrop-diastereomers that have opposite
axial chirality and the same configuration of a distal
stereogenic center.19 Following this trend, freshly isolated
talaroderxines A and B were found to have rotations
of þ117.8 and ꢀ215.0, suggesting that Suzuki’s material
was not isomerically pure or that the sign of the rotation of
talaroderxineA wasreported incorrectly.20 Wedidobserve
Acknowledgment. This work was supported by the
NIH/NIAID (R01AI080931-01) and by startup funds
from the University of California, Davis. Prof. James
Gloer (University of Iowa) is acknowledged for helpful
discussions and for providing the extract containing talar-
oderxines A and B. Prof. John A. Elix (Australia National
University) is acknowledged for conducting HPLC ana-
lyses to compare natural and synthetic pigmentosin A.
Prof. Alan R. Buckpitt (University of California, Davis) is
acknowledged for conducting mouse toxicity studies. Prof.
Douglas Weibel and Ms. Marie Foss (University of
Wisconsin) are acknowledged for assistance with the
ꢀ
ꢀ
MIC measurements. Dr. Marcos Gonzalez-Lopez (The
Scripps Research Institute) is acknowledged for helpful
discussions at the inception of this project.
Supporting Information Available. Characterization of
all new compounds, 1H and 13C NMR spectra, and
comparative analytical data. This material is available
free of charge via the Internet at http://pubs.acs.org.
(19) (a) Kuyama, S.; Tamura, T. J. Am. Chem. Soc. 1957, 79, 5725–5726.
€
(b) Bringmann, G.; Gunther, C.; Saeb, W.; Mies, J.; Wickramasinghe, A.
Mudogo, V.; Brun, R. J. Nat. Prod. 2000, 63, 1333–1337. (c) Bringmann,
G.; Messer, K.; Brun, R.; Mudogo, V. J. Nat. Prod. 2002, 65, 1096–1101.
(d) Mitchell, T. K.; Alejos-Gonzalez, F.; Gracz, H. S.; Danehower,
D. A.; Daub, M. E.; Chilton, W. S. Phytochemistry 2003, 62, 723–732. (e)
Watanabe, T.; Tanaka, Y.; Shoda, R.; Sakamoto, R.; Kamikawa, K.;
Uemura, M. J. Org. Chem. 2004, 69, 4152–4158. (f) Talele, H. R.; Sahoo,
S.; Bedekar, A. V. Org. Lett. 2012, 14, 3166–3169.
(20) Extract from a liquid fungal culture (Delitschia sp.) containing
talaroderxines A and B was prepared by the Gloer group (U. Iowa)
according to ref 2 and transferred to UC Davis for HPLC purification
and analysis. See Supporting Information.
Note Added after ASAP Publication. This manuscript
was published ASAP on August 13, 2012. Scheme 2 has
been updated. The corrected version was reposted on
August 15, 2012.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 17, 2012
4341