Olefin metathesis in carotenoid synthesis

Article abstract of DOI:10.1039/b915390j

Olefin metathesis is a powerful and widely applicable synthetic method for carbon-carbon double bond formation. However, its application to the synthesis of conjugating polyene chains has been very limited because of possible undesired side reactions. We attempted to apply this method to the synthesis of symmetrical carotenoids. In this paper, the syntheses of violaxanthin and mimulaxanthin are described using the olefin metathesis protocol.

Full text of DOI:10.1039/b915390j

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/obc | Organic & Biomolecular Chemistry
Olefin metathesis in carotenoid synthesis†
Takayuki Kajikawa, Naoko Iguchi and Shigeo Katsumura*
Received 29th July 2009, Accepted 8th September 2009
First published as an Advance Article on the web 21st September 2009
DOI: 10.1039/b915390j
Olefin metathesis is a powerful and widely applicable syn-
thetic method for carbon–carbon double bond formation.
However, its application to the synthesis of conjugating
polyene chains has been very limited because of possible
undesired side reactions. We attempted to apply this method
to the synthesis of symmetrical carotenoids. In this paper, the
syntheses of violaxanthin and mimulaxanthin are described
using the olefin metathesis protocol.
polyene chains is very limited because of possible undesired side
reactions.4
Carotenoids are ubiquitous natural pigments distributed widely,
and their isolation from a variety of species such as bacteria, yeast,
algae, plants, animals and even humans has been documented.5
They generally possess 40 carbon atoms and contain a long con-
jugated polyene chain. The representative coupling methods for
the functionalized polyene skeleton construction of carotenoids
are the utilization of the Wittig reaction,6 Stille coupling,7 and
modified Julia olefination.8 In these methods, characteristic func-
tional groups are needed. We then attempted to apply the olefin
metathesis reaction to the synthesis of symmetrical carotenoids
such as violaxanthin (1) and mimulaxanthin (2) (Fig. 1).9 In these
syntheses, we need only one kind of synthon for the metathesis, and
hence their synthesis is rather simple and convenient. In addition,
we anticipated that the desired all-trans compounds would be
obtained based on the thermodynamic stability of the products. In
this paper, the syntheses of violaxanthin (1) and mimulaxanthin
(2) using the olefin metathesis protocol are described, as shown in
Fig. 1.
Over the past decade, olefin metathesis has emerged as a
powerful and widely applicable olefin forming reaction,1 and
been utilized for many complicated natural product syntheses.
Obviously, this reaction has been accepted as one of the most
general and convenient protocols for C–C double bond formation.
Although this reaction is also used for the ene–yne metathesis2
and polymerization,3 application to the synthesis of conjugating
Department of Chemistry and Open Research Center on Organic
Tool Molecules, School of Science and Technology, Kwansei Gakuin
University, Gakuen 2-1, Sanda, Hyogo, 669-1337, Japan. E-mail:
katsumura@kwansei.ac.jp; Fax: +81-79-565-9077; Tel: +81-79-565-8314
† Electronic supplementary information (ESI) available: Experimental
procedures and characterization data; copies of NMR spectra. See DOI:
10.1039/b915390j
Violaxanthin (1) was isolated from Viola tricolor by Kuhn and
Winterstein,10 and its structure was determined by synthesis, utiliz-
ing the Wittig and Horner-Wadsworth-Emmons (HWE) reactions,
Fig. 1 Synthons for olefin metathesis.
This journal is
4586 | Org. Biomol. Chem., 2009, 7, 4586–4589
The Royal Society of Chemistry 2009
©

View Article Online
by Eugster’s group.11 Prior to the synthesis of violaxanthin (1), we
targeted the shorter C30 violaxanthin analogue 3 to understand
the applicability of the olefin metathesis reaction to the polyene
synthesis (Fig. 1). Thus, vinyl stannane 8, which was prepared
starting from the known (-)-epoxyaldehyde 712 according to the
reported procedure,7,8 was transformed into the corresponding
vinyl iodide 9 in excellent yield (Scheme 1). The Stille cross-
coupling reaction of iodide 9 with the known vinyl stannane
1013 afforded the desired alcohol 11, which was transformed in
good yield into triene 6, the synthon for the metathesis, by MnO2
oxidation followed by the Wittig reaction. Fortunately, the homo-
coupling of the obtained 6 in the presence of Grubbs second-
generation catalyst (5 mol%) at 45 ◦C produced the expected
product 3 in 53% yield. In this case, it was estimated to be
92% of the desired all-trans 3, based on HPLC and 400 MHz
NMR analyses of the product.14 Thus, the pentaene analogue 3 of
violaxanthin was rapidly synthesized.
Table 1 Synthesis of violaxanthin (1)
yield of crude
entry cat. (mol%)
temp. (◦C) time (min) products (%)
1
2
3
4
5
6
7
8
Grubbs 2nd
(5) rt
(5) 45
(5) 45
(10) 45
(10) 60
(5) 80
20
20
120
20
10
10
trace
37
trace
56
67
13
Hoveyda–Grubbs 2nd (10) 60
10
10
23
10
Grubbs 1st
(10) 60
Scheme 2 Synthesis of pentaene 4. Reagents and Conditions: (a) MnO2,
THF, r.t., 2 h; (b) triethyl 3-methyl 4-phosphonocrotonate, nBuLi, DMPU,
THF, 0 ◦C, 10 min, 85% for 2 steps; (c) DIBAL, CH2Cl2, -78 ◦C, 10 min,
84%; (d) 9, Pd(PPh3)4, LiCl, iPrNEt, DMF, 65 ◦C, 30 min, 64%; (e) MnO2,
THF, r.t., 50 min; (f) methyltriphenylphosphonium bromide, NaHMDS,
THF, -20 ◦C, 5 min, 70% for 2 steps.
Scheme 1 Synthesis of C30 violaxanthin analogue 3. Reagents and
Conditions: (a) I2, Na2CO3, CH2Cl2, 0 ◦C, 15 min, 98%; (b) 10,
Pd(CH3CN)2Cl2, LiCl, DMF, r.t., 20 min, 86%; (c) MnO2, THF, r.t.,
20 min; (d) methyltriphenylphosphonium bromide, NaHMDS, THF, r.t.,
5 min, 72% for 2 steps; (e) Grubbs 2nd generation cat., toluene, 45 ◦C,
20 min, 53%.
in toluene at 45 ◦C violaxanthin (1) was obtained in about
35% yield. Next, we surveyed the equivalents of the catalyst
used, temperature, and the reaction time using Grubbs second-
generation catalyst in toluene (Table 1). The reaction of 5 mol% of
the catalyst in 45 ◦C gave the coupling products in 37% yield, but
the products gradually decomposed with a longer reaction time
(entries 2 and 3). The use of 10 mol% catalyst improved the yield
(56%), and the coupling products were obtained in about 67%
amount at 60 ◦C (entries 4 and 5). A higher reaction temperature,
80 ◦C, led to a lower yield (entry 6). On the other hand, the
Hoveyda–Grubbs catalyst and Grubbs first-generation catalyst
gave undesirable results (entries 7 and 8).
The detailed analysis of the coupling products obtained in
entry 5 by mobile-phase HPLC showed two major peaks (A–C
of Fig. 2). As shown in (A) of Fig. 2, the HPLC detection at a
470 nm wavelength, which indicates the C40 violaxanthin deriva-
tives, was different from that detected at a 423 nm wavelength,
which was mainly ascribed to the C35 violaxanthin analogues, as
shown in (B) of Fig. 2. We then isolated both compounds and
elucidated their structures based on NMR, mass and UV spectra.
Thus, we clarified that peak 1 was the all-trans violaxanthin (1)
Next, we tried to synthesize violaxanthin (1) using the olefin
metathesis. The stereocontrolled synthesis of precursor 4 is shown
in Scheme 2. Vinyl stannane 13 was prepared in good yield from 10
by the sequence of MnO2 oxidation, HWE reaction of the resulting
aldehyde,15 and then DIBAL reduction. The Stille cross-coupling
reaction of vinyl iodide 9 with vinyl stannane 13 in the presence of
i
Pd(PPh3)4, LiCl and Pr2NEt gave the desired alcohol 14 in 64%
yield as a single isomer. The obtained alcohol 14 was transformed
into the pentaene 4 by MnO2 oxidation followed by the Wittig
reaction. The product of the Wittig reaction at -20 ◦C was
mainly the 13E-isomer (13E/13Z = 7/1). The reaction at -78 ◦C
produced a higher stereoselectivity (13E/13Z = 15/1), but in low
yield (32%), while the reaction at a higher temperature (0 ◦C)
gave a lower stereoselectivity (13E/13Z = 3/1). We then used the
mixture of stereoisomers of 4 (13E/13Z = 7/1) for the coupling
reaction.
The results of the metathesis of 4 are shown in Table 1.
The reaction did not proceed in dichloromethane at 40 ◦C, but
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4586–4589 | 4587
©

View Article Online
10 min under the same conditions as those for the violaxanthin
synthesis, and produced a mixture of products in a 34% yield.
Meanwhile, the addition of Grubbs second-generation catalyst
(5 mol%) for four times at 5 min intervals provided a 56% yield.
The detailed analysis of the coupling products by HPLC detected
at a 470 nm wavelength showed one major peak and some minor
peaks as shown in (D) of Fig. 2. In this case, the major peak
1 comprised 52% of the reaction products. We then isolated the
major peak 1 and elucidated its structure based on NMR, mass
and UV spectra. Fortunately, peak 1 was the desired all-trans
mimulaxanthin (2) (lmax= 470 nm).18 The spectral data of the
isolated all-trans mimulaxanthin (2) were in good agreement with
those already reported.17
In summary, we demonstrated that the olefin metathesis pro-
tocol could be applied to the synthesis of conjugated polyene
compounds, in particular to symmetrical carotenoid synthesis.
Violaxanthin (1) possessing a conjugating nonaene system and
mimulaxanthin (2) possessing a heptaene system conjugating to
two allenic functions were synthesized by utilizing the olefin
metathesis as a key reaction at the final step. In carotenoid
synthesis, the isolation of the desired compound from a mixture of
stereoisomers is usually troublesome, and for the product obtained
by the olefin metathesis strategy this is also unavoidable. The above
results, however, would provide a new strategy for the synthesis of
symmetrical carotenoids.
Fig. 2 Results of HPLC analysis of violaxanthin (A–C) and mimulaxan-
thin (D).
(lmax= 470 nm). Meanwhile, we estimated that peak 2 was the C35
violaxanthin analogue (lmax= 423 nm) by mass, UV and NMR
spectra. The spectral data of the isolated all-trans violaxanthin
(1) were in good agreement with those already reported.11 Next,
in order to determine the ratio of these compounds, we changed
the wavelength for the detection from 423 nm to 470 nm after
30 minutes. As shown in (C) of Fig. 2, the ratio between the
compounds of peak 1 and peak 2 was approximately 5 to 1.
Analysis of the compounds detected by 470 nm revealed that this
mixture consisted of the desired all-trans violaxanthin (1) (peak
1; 84%), its three stereoisomers (7, 5 and 2%) and others (2%) by
HPLC and mass spectroscopy. Thus, the all-trans violaxanthin (1)
was rapidly synthesized by the olefin metathesis protocol in 49%
estimated yield .
Next was the synthesis of mimulaxanthin (2), which was isolated
from the flowers of Mimulus guttatus and Lamium montanum16
and its structure determined by synthesis, utilizing the Wittig and
HWE reactions, by Eugster and Buchecker.17 The C20-allenic triol
15 (Scheme 3), which was previously synthesized by us, was a
synthetic intermediate and was prepared as a E/Z mixture at the
C13 position (13E/13Z =10/1).6c The coupling precursor, allenic
tetraene 5, was prepared from 15 by MnO2 oxidation followed
by the Wittig reaction as in the case of violaxanthin (1). Olefin
metathesis of the obtained allenic tetraene 5 proceeded within
Acknowledgements
We thank Dr Thomas Netscher of DSM Nutritional Products,
Ltd., for the donation of (-)-actinol. This work was supported by
a Grant-in-Aid for Science Research on Priority Areas 16073222
from the Ministry of Education, Culture, Sports, Science and
Technology, Japan. This work was also supported by a Matching
Fund Subsidy for a Private University.
Notes and references
1 For recent reviews on olefin metathesis, see: R. H. Grubbs, Ed.
Handbook of Metathesis, Wiley-VCH, Weinheim, Germany, 2003; R.
H. Grubbs and T. M. Trnka in Ruthenium in Organic Synthesis, S.–I.
Murahashi, Ed.; Wiley-VCH, Weinheim, 2004, chapter 6.; A. Deiters
and S. F. Martin, Chem. Rev., 2004, 104, 2199; T. Katz, Angew. Chem.,
Int. Ed., 2005, 44, 3010.
2 For reviews on enyne metathesis, see: C. Aubert, O. Buisine and M.
Malacria, Chem. Rev., 2002, 102, 813; A. J. Giessert and S. T. Diver,
Chem. Rev., 2004, 104, 1317.
3 For recent reports on polymerization, see: A. Hejl, M. W. Day and
R. H. Grubbs, Organometallics, 2006, 25, 6149; I. A. Gorodetskaya,
T. Choi and R. H. Grubbs, J. Am. Chem. Soc., 2007, 129, 12672; I.
A. Gorodeskayam, A. A. Gorodetsky, E. V. Vinogradova and R. H.
Grubbs, Macromolecules, 2009, 42, 2895.
4 For recent reports in retinoid and polyene macrolide synthesis, see: A.
Wojtkielewicz, J. Maj and J. W. Morzycki, Tetrahedron Lett., 2009, 50,
4734; M. J. Mitton-Fry, A. J. Cullen and T. Sammakia, Angew. Chem.,
Int. Ed., 2007, 46, 1066.
5 For authorized monographs, see: G. Britton, S. Liaaen-Jensen and H.
Pfander, Eds. Carotenoids. Volume 1 and 2, Birkhauser, Basel, 1995.
6 C. Tode, Y. Yamano and M. Ito, J. Chem. Soc., Perkin Trans. 1, 2002,
1581; Y. Yamano and M. Ito, Chem. Pharm. Bull., 2004, 52, 780; Y.
Murakami, M. Nakano, T. Shimofusa, N. Furuichi and S. Katsumura,
Org. Biomol. Chem., 2005, 3, 1372; F. Khachik and A. Chang, J. Org.
Chem., 2009, 74, 3875.
Scheme 3 Synthesis of mimulaxanthin (2). Reagents and Conditions:
(a) MnO2, AcOEt, r.t., 20 min; (b) methyltriphenylphosphonium bromide,
NaHMDS, THF, 0 ◦C, 10 min, 44% for 2 steps; (c) Grubbs 2nd generation
cat., toluene, 60 ◦C, 20 min, 56%.
7 B. Vaz, R. Alvarez and A. R. de Lera, J. Org. Chem., 2002, 67, 5040;
B. Vaz, M. Dominguez, R. Alvarez and A. R. de Lera, Chem.–Eur. J.,
4588 | Org. Biomol. Chem., 2009, 7, 4586–4589
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
1
2007, 13, 1273; J. Burghart and R. Bru¨ckner, Angew. Chem., Int. Ed.,
2008, 47, 7664.
structure of 3. We then estimated it by referring H NMR spectra of
violaxanthin (1), because the environment around the central olefin of
the both compounds is almost same. Our conclusion agrees well with
the empirical 1H NMR data of carotenoids, see ref. 5.
8 N. Furuichi, H. Hara, T. Osaki, H. Mori and S. Katsumura, Angew.
Chem., Int. Ed., 2002, 41, 1023; N. Furuichi, H. Hara, T. Osaki, M.
Nakano, H. Mori and S. Katsumura, J. Org. Chem., 2004, 69, 7949; T.
Olpp and R. Bru¨ckner, Angew. Chem., Int. Ed., 2006, 45, 4023.
9 For the synthesis of symmetrical carotenoids by using one kind of
synthon, see: J. E. McMurry and M. P. Fleming, J. Am. Chem. Soc.,
1974, 96, 4708.
10 R. Kuhn and A. Winterstein, Chem. Ber., 1931, 64, 326.
11 M. Acemoglu, P. Uebelhardt, M. Rey and C. H. Eugster, Helv. Chim.
Acta, 1988, 71, 931.
15 B. Dominguez, B. Iglesias and A. R. de Lera, Tetrahedron, 1999, 55,
15071.
16 H. Z. Nitsche, Naturforsch, 1973, 28C, 481; R. Buchecker and C. H.
Eugster, Helv. Chim. Acta, 1980, 63, 2531.
17 R. Buchecker and C. H. Eugster, Helv. Chim. Acta, 1991, 74, 469.
18 Other minor peaks of (D) in Fig. 2 were assumed to be stereoisomers
of C40 mimulaxanthin resulting from the mass spectra. In order to
estimate the yield of the obtained mimulaxanthin (2) in a similar
manner to that of violaxanthin (1), we also measured HPLC detected
by 423 nm wavelength that could detect C35 mimulaxanthin analogues.
Unfortunately, C35 mimulaxanthin analogues could not be separated
from the stereoisomers of C40 mimulaxanthin, and the yield of
mimulaxanthin (2) synthesized was not estimated.
12 M. Kuba, N. Furuichi and S. Katsumura, Chem. Lett., 2002, 1248.
13 J. F. Betzer, F. Delaloge, B. Muller, A. Pancrazi and J. Prunet, J. Org.
Chem., 1997, 62, 7768.
14 It was difficult to determine the E-geometry at the C11-C11¢ position
of the product 3 by NOE and J-coupling because of the symmetrical
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4586–4589 | 4589
©