Stereocontrolled first total syntheses of amarouciaxanthin A and B

Article abstract of DOI:10.1021/ol402540g

The first total syntheses of amarouciaxanthin A and B (C₄₀) via the stereoselective Wittig reaction of C₁₅-allenic and C ₁₅-acetylenic tri-n-butylphosphonium salts with the unprecedented C₂₅-3,8-dihydroxy-5,6-epoxyapocarotenal have been completed. Oxidation of the two hydroxyl groups in the left part of the resulting condensation products followed by regioselective oxirane ring opening gave the target carotenoids.

Full text of DOI:10.1021/ol402540g

ORGANIC
LETTERS
2013
Vol. 15, No. 20
5310–5313
Stereocontrolled First Total Syntheses
of Amarouciaxanthin A and B
Yumiko Yamano,* Mahankhali Venu Chary, and Akimori Wada
Department of Organic Chemistry for Life Science, Kobe Pharmaceutical University,
4-19-1, Motoyamakita-machi, Higashinada-ku, Kobe 658-8558, Japan
y-yamano@kobepharma-u.ac.jp
Received September 5, 2013
ABSTRACT
The first total syntheses of amarouciaxanthin A and B (C40) via the stereoselective Wittig reaction of C15-allenic and C15-acetylenic tri-n-
butylphosphonium salts with the unprecedented C25-3,8-dihydroxy-5,6-epoxyapocarotenal have been completed. Oxidation of the two hydroxyl
groups in the left part of the resulting condensation products followed by regioselective oxirane ring opening gave the target carotenoids.
Amarouciaxanthin A (1) and B (2) (Figure 1), which
have a novel γ-hydroxy cyclohexenone moiety, were first
isolated from the tunicate Amaroucium pliciferum and are
thought to be metabolites of fucoxanthin (3) via oxirane
ring opening.1 Biotransformation of fucoxanthinol (4), a
deacetylated metabolite of 3, into 1 was indeed confirmed
by incubation of 4 with mouse liver homogenate.2 Fucox-
anthin (3) is widely distributed in brown algae, most of
which are edible, and has various physiological functions
including anticarcinogenic3 and antiobese activities.4 Re-
cently, amarouciaxanthin A (1) was reported to signifi-
cantly suppress adipocyte differentiation in comparison
with its metabolite precursor 4.4a Thus, both 1 and 2 are
expected to show strong or specific effects in various
functions. However, because of their limited availability
from natural sources, their properties and functions are
not yet well understood. Moreover, the absolute config-
uration at C6 in these compounds has not been unequi-
vocally established; the configuration is presumed to be S
basedonasimplecomparisonoftheirCDspectrawiththat
of (S)-abscisic acid (5).1 Thus, interest in their function and
structure prompted us to undertake the first total synthesis
of 1 and 2.
(1) Matsuno, T.; Ookubo, M.; Komori, T. J. Nat. Prod. 1985, 48, 606.
(2) Asai, A.; Sugawara, T.; Ono, H.; Nagao, A. Drug Metab. Dispos.
2004, 32, 205.
(3) (a) Okuzumi, J.; Takahashi, T.; Yamane, T.; Kitao, Y.; M.
Inagake, M.; Ohya, K.; Nishino, H.; Tanaka, Y. Cancer Lett. 1993,
68, 159. (b) Kotake-Nara, E.; Kushiro, M.; Zhang, H.; Sugawara, T.;
Miyashita, K.; Nagao, A. J. Nutr. 2001, 131, 3303. (c) Hosokawa, M.;
Kudo, M.; Maeda, H.; Kohno, H.; Tanaka, T.; Miyashita, K. Biochim.
Biophys. Acta 2004, 1675, 113. (d) Das, S. K.; Hashimoto, T.; Kanazawa,
K. Biochim. Biophys. Acta 2008, 1780, 743. (e) Yu, R.; Hu, X.; Xu, S.;
Jiang, Z.; Yang, W. Eur. J. Pharmacol. 2011, 657, 10.
Scheme 1 shows our retrosynthetic analyses. It is becom-
ing common to utilize metal-catalyzed coupling reac-
tions in polyene synthesis of carotenoids;5 however, these
(4) (a) Yim, M.-J.; Hosokawa, M.; Mizushina, Y.; Yoshida, H.;
Saito, Y.; Miyashita, K. J. Agric. Food Chem. 2011, 59, 1646. (b) Tsukui,
T.; Konno, K.; Hosokawa, M.; Maeda, H.; Sashima, T.; Miyashita, K.
J. Agric. Food Chem. 2007, 55, 5025. (c) Maeda, H.; Hosokawa, M.;
Sashima, T.; Funayama, K.; Miyashita, K. Biochem. Biophys. Res.
Commun. 2005, 332, 392.
(5) (a) Zeng, F.; Negishi, E. Org. Lett. 2001, 3, 719. (b) Vaz, B.;
Alvarez, R.; Bruckner, R.; de Lera, A. R. Org. Lett. 2005, 7, 545. (c) Vaz,
B.; Domınguez, M.; Alvarez, R.; de Lera, A. R. J. Org. Chem. 2006, 71,
5914. (d) Vaz, B.; Domınguez, M.; Alvarez, R.; de Lera, A. R. Chem.;
€
€
Eur. J. 2007, 13, 1273. (e) Burghart, J.; Bruckner, R. Angew. Chem., Int.
Ed. 2008, 47, 7664.
r
10.1021/ol402540g
Published on Web 10/03/2013
2013 American Chemical Society

previously reported10 C10-phosphonate 11. The challeng-
ing transformation of C10-epoxyalcohol 12, whose highly
stereoselective preparation by Sharpless asymmetric epox-
idation has been documented,11 into C11-epoxyaldehyde 9
would be achieved by C1-homologation.
Scheme 1. Retrosynthetic Analyses
Figure 1. Structure of target carotenoids and related compounds.
reactions may give unsatisfactory yields and produce
hazardous waste. In the synthesis of acetylenic carotenoids
in particular, isomerization of the CdC of the enyne
moiety under the coupling conditions is a major flaw.5c,e
We previously reported6 that C15-acetylenic tri-n-butyl-
phosphonium salt 8 is a useful and versatile tool for
stereoselective synthesis of acetylenic carotenoids. Thus,
we planned to construct the polyene chain of 1 and 2
stereoselectively by using 8 and the corresponding allenic
phosphonium salt 7, which could be prepared from
(ꢀ)-actinol (13).7 The fact1 that 1 and 2 are decomposed
into methyl ketones, namely, paracentrone and triophax-
anthin, via a retro-aldol reaction upon alkaline treatment
prompted us to construct the alkali-labile β-hydroxy keto
moiety of these carotenoids in the final step. We envisioned
using the C25-3,8-dihydroxy-5,6-epoxyapocarotenal8 6 as
a condensation partner for the ylides derived from phos-
phonium salts 7 and 8.6 As a key step in constructing the
left part of 1 and 2, oxidation of both C3- and C8-hydroxyl
groups of the condensation products was designed to
obtain diketo compounds, which we envisioned would
undergo regioselective oxirane ring opening via deproto-
nation of the more acidic C4 proton to afford our targets.
The C25-apocarotenal 6 was expected to be prepared by a
three-component connection, which involves the addition
of the C4-alkenyllithium reagent 109 to the C11-epoxyalde-
hyde 9 and the HornerꢀEmmons condensation with the
First, the C15-allenic phosphonium salt 7 was prepared
as shown in Scheme 2. After protecting the hydroxyl group
in 13 with triethylsilyl (OTES) and subsequent triflation,
the resulting vinyltriflate 15 was transformed into allylic
alcohol 17 by palladium-catalyzed methoxycarbonylation11,12
Scheme 2. Synthesis of C15-Allenic Phosphonium Salt 7
(6) Yamano, Y.; Chary, M. V.; Wada, A. Org. Biomol. Chem. 2012,
10, 4103.
(7) Leuenberger, G. W.; Bouguth, W.; Widmer, E.; Zell, R. Helv.
Chim. Acta 1976, 59, 1832.
(8) We employed the numbering system used in carotenoids.
(9) (a) Tode, C.; Yamano, Y.; Ito, M. Chem. Pharm. Bull. 2000, 48,
1833. (b) Tode, C.; Yamano, Y.; Ito, M. J. Chem. Soc., Perkin Trans. 1
2002, 1581.
(10) Yamano, Y.; Yoshizawa, M.; Ito, M. J. Nutr. Sci. Vitaminol.
1999, 45, 49.
Org. Lett., Vol. 15, No. 20, 2013
5311

followed by DIBALH reduction. Sharpless oxidation of 17
assisted by (ꢀ)-diethy-D-tartrate (D-DET) provided diaster-
eomerically pure anti(R)-epoxide 18, which was oxidized by
using Dess-Martin periodinane (DMP) to afford aldehyde 19.
C1-extension of 19 by the Colvin/Shioiri protocol13 gave the
epoxyalkyne 20 similarly as described for closely related
substrates.5c,14 By using Pd(PPh3)4 and CuI as catalysts in
diisopropylamine and by degassing the reaction mixture,
Sonogashira cross-coupling between 20 and vinyl iodide
2111b gave the desired epoxyalcohol 22 in high yield. The
stereospecific SN20 hydride reduction of 22 with DIBALH
produced the known allenic diol 23,15 which was converted to
tri-n-butylphosphonium salt 7 via the corresponding allylic
chloride. The total yield of the phosphonium salt 7 from
(ꢀ)-actinol (13) was 34% over 11 steps.
provided the desired saturated ester 28 in 54% yield,
accompanied by some products of oxirane ring opening.16
Ester 28 was transformed into the C1-shortened aldehyde 9
in three steps. Treatment of 28 with commercially available
oxaziridine 2917 in the presence of LDA yielded R-hydro-
xylated 30 as a single diastereomer; this was reduced with
LAH, and the resulting glycol was cleaved with NaIO4 to
afford aldehyde 9 in high yield. The total yield of 9 from
(ꢀ)-actinol (13) was 30% over 11 steps.
C11-Epoxyaldehyde 9 was then converted into C25-
apocarotenal 6 as shown in Scheme 4. Aldehyde 9 was
treated with a reagent obtained from alkenyl bromide
319,18 and t-BuLi to give the diastereomeric alcohols 32
as an unassigned 2:1 mixture; without separating the
diastereomers, 32 was acetylated and then desilylated to
yield diol 33. MnO2 oxidation of 33 and protection of the
C3-hydroxyl with TES afforded aldehyde 34. The C8-
acetoxy group on compound 34 is ultimately transformed
into a carbonyl group, which destroys the stereogenic
center and makes the separation of the diastereomers
unnecessary. However, we separated the diastereomers in
this step for the sake of convenience in spectral analyses of
subsequent compounds. The major diastereomer of 34 was
condensed with the previously reported10 C10-phospho-
nate 11, and the resulting pentaenoate was subjected to
LAH reduction followed by MnO2 oxidation and subse-
quent deprotection to provide all-E-apocarotenal 6 and its
13Z-isomer in 50% and 19% yield from 34, respectively.
The 13Z-isomer was converted into the desired all-E-
isomer (73%: HPLC yield) by isomerization19 using a
palladium catalyst. The minor diastereomer of 34 was also
converted into the corresponding apocarotenal.
Scheme 3. Synthesis of C11-Epoxyaldehyde 9
Scheme 4. Synthesis of C25-Apocarotenal 6
Next, C11-epoxyaldehyde 9, the precursor of apocarote-
nal 6, was prepared as shown in Scheme 3. Previously
reported9 epoxyacetate 25a was easily converted into
epoxyaldehyde 9 by LAH reduction followed by DMP
oxidation, but the overall yield of 9 from 13 was as low as
19%, mainly due to the poor diastereoselectivity of epox-
idizing alkene 24. Thus, weinvestigatedthe transformation
of C10-epoxyalcohol 12, which has been prepared11 in a
highly stereoselective manner, into C11-compound 9. Be-
cause several trials of the direct C1-extension using alcohol
12 and corresponding aldehyde 26 were unsuccessful,
aldehyde 26 was converted into a C2-elongated conjugated
ester 27 by the HornerꢀEmmons reaction. After a detailed
investigation of reduction reagents and catalysts, it was
found that hydrogenation of 27 by using Pd/C as a catalyst
The next step toward amarouciaxanthin A and B was
Wittig condensation of C25-apocarotenal 6 with C15-tri-n-
butylphosphonium salts 7and 8(Scheme5). Weinvestigated
the further transformation by using the major diastereomer
(16) The data of oxirane ring opening products are provided in the
Supporting Information.
(11) (a) Furuichi, N.; Hara, H.; Osaki, T.; Mori, H.; Katsumura, S.
Angew. Chem., Int. Ed. 2002, 41, 1023. (b) Furuichi, N.; Hara, H.; Osaki,
T.; Nakano, M.; Mori, H.; Katsumura, S. J. Org. Chem. 2004, 69, 7949.
(12) Stille, J. K.; Wong, P. K. J. Org. Chem. 1975, 40, 532.
(17) Davis, F. A.; Reddy, G. V.; Chen, B.-C.; Kumar, A.; Haque,
M. S. J. Org. Chem. 1995, 60, 6148.
ꢀ
(18) de Lera, A. R.; Iglesias, B.; Rodriguez, J.; Alvarez, R.; Lopez, S.;
ꢀ
Villanueva, X.; Padros, E. J. Am. Chem. Soc. 1995, 117, 8220.
(13) Miwa, K.; Aoyama, T.; Shioiri, T. Synlett 1994, 107.
(19) (a) Fischli, A.; Mayer, H.; Simon, W.; Stoller, H.-J. Helv. Chim.
Acta 1976, 59, 397. (b) Yamano, Y.; Ito, M. Org. Biomol. Chem. 2007, 5,
3207.
€
(14) Olpp, T.; Bruckner, R. Angew. Chem., Int. Ed. 2006, 45, 4023.
(15) Baumeler, A.; Eugster, C. H. Helv. Chim. Acta 1991, 74, 469.
5312
Org. Lett., Vol. 15, No. 20, 2013

of apocarotenal 6. As expected, the reaction of 6 with the
acetylenic phosphonium salt 8 under previously reported6
conditions (NaOMe in CH2Cl2) stereoselectively pro-
ceeded to afford the all-E condensed C40-epoxydiol 35
in 76% yield. This diol was oxidized with 2-iodoxyben-
zoic acid (IBX) in the presence of Et3N in DMSO/THF to
afford the expected epoxydiketone. The latter gradually
ring-opened to the desired γ-hydroxyenone 36 during
purification using a flash silica gel column. However, this
reaction did not proceed to completion. Therefore, the
mentioned epoxydiketone/γ-hydroxyenone mixture was
treated with a large amount of silica gel in AcOEt over-
night. This provided the γ-hydroxyenone 36 without
isomerization of the enyne moiety.20 Finally, 36 was
treated with pyridinium p-toluenesulfonate (PPTS) in
MeOH to afford amarouciaxanthin B (2) in 89% yield.
Desilylation of 36 with TBAF provided triophaxanthin
via a retro-aldol reaction, whereas combined use of
TBAF and acetic acid (1:1) led to the competing forma-
tion of triketone 37.21 Spectral data for the synthetic
specimen of 2 were in good agreement with the reported
data1 including CD data. This shows that the proposed
configuration at C6 is correct.
Scheme 5. Syntheses of Amarouciaxanthin A and B
By the same approach, amarouciaxanthin A (1) was
effectively synthesized by a sequence of stereoselective
condensation of C25-apocarotenal 6 with C15-allenic phos-
phonium salt 7, oxidation, regioselective oxirane ring
opening, and removal of the TES group. Spectral data
for the synthetic specimen of 1 were in good agreement
with the reported data.1
In summary, we achieved the first total syntheses of
amarouciaxanthin A (1) and B (2) in a highly efficient and
convergent manner. Phosphonium salts 7 and 8 proved
and reproved6 to be, respectively, versatile building blocks
for allenic and acetylenic carotenoids because of their ease
of use and stereocontrol. This method will provide materials
needed for investigating the biological functions of 1 and 2.
(20) Oxidation of 35 with DMP was poorly reproducible and often
resulted in decomposition; oxidation with IBX in the absence of Et3N
was accompanied by the 3,8,30-trioxo compound formed by removal of
the TES group and subsequent oxidation.
(21) Treatment with TBAF in the presence of acetic acid gradually
converted 2 into 37. The chemical structure of 37 was determined by
NMR and MS spectra. The plausible configuration of C5 in 37 was
deduced to be R based on the following semipinacol-type rearrangement
mechanism.
Supporting Information Available. Experimental de-
tails and characterization data on 6, 7, 9, 14ꢀ20, 22, 23,
27ꢀ30, 32ꢀ34, 36, 37, 1, and 2. This material is available
free of charge via Internet at http://pubs.acs.org.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 20, 2013
5313