New methods for the synthesis of carotenoid glycosides

Article abstract of DOI:10.1016/j.tetlet.2014.04.114

We describe our studies on the synthesis of carotenoid glucosides and deoxyglucosides using the acetimidate method and the Ferrier rearrangement, respectively. In both cases the reaction conditions were optimized until the yields were superior to those of

Full text of DOI:10.1016/j.tetlet.2014.04.114

Accepted Manuscript
New methods for the synthesis of carotenoid glycosides
Miki Hanaura, Attila Agócs, Katalin Böddi, József Deli, Veronika Nagy
PII:
S0040-4039(14)00751-5
DOI:
http://dx.doi.org/10.1016/j.tetlet.2014.04.114
TETL 44576
Reference:
Tetrahedron Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
25 March 2014
10 April 2014
30 April 2014
Please cite this article as: Hanaura, M., Agócs, A., Böddi, K., Deli, J., Nagy, V., New methods for the synthesis of
carotenoid glycosides, Tetrahedron Letters (2014), doi: http://dx.doi.org/10.1016/j.tetlet.2014.04.114
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Graphical abstract
New methods for the synthesis of carotenoid glycosides
Miki Hanaura, Attila Agócs*, Katalin Böddi, József Deli, Veronika Nagy
We describe our studies on the synthesis of carotenoid glucosides and deoxyglucosides using the
acetimidate method and the Ferrier rearrangement, respectively. In both cases the reaction conditions
were optimized until the yields were superior to those of previously published glycosylations.

New methods for the synthesis of carotenoid glycosides
Miki Hanaura,a Attila Agócs,a,* Katalin Böddi,a József Deli,b Veronika Nagya
aDepartment of Biochemistry and Medical Chemistry, University of Pécs, Medical School,
Szigeti u. 12, H-7624 Pécs, Hungary
bDepartment of Pharmacognosy, University of Pécs, Medical School,
Rókus u. 2, H-7624 Pécs, Hungary
*Corresponding author: Tel.: +36-72-536001/ext. 31864; fax:+36-72-536225; e-mail:
attila.agocs@aok.pte.hu
Abstract
We describe our studies on the synthesis of carotenoid glucosides and deoxyglucosides using
the acetimidate method and the Ferrier rearrangement, respectively. In both cases the reaction
conditions were optimized until the yields were superior to those of previously published
glycosylations.
Keywords: carotenoid; glycoside; acetimidate; Ferrier rearrangement.
Thermoxanthins, i.e., C40 carotenoid glycosides, have been described as constituents
of the cell membrane of certain Thermus bacteria, and are believed to be partially responsible
for the heat resistance of the Thermus species.1 Carotenoids are also good anti-oxidants,
whilst hydrophilic carotenoids are even more so,2 and therefore thermoxanthins in the cell
membrane can play a role in the inhibition of oxidative stress. Since carotenoid glycosides are
of limited availability from natural sources,3 their effect on oxidative stress has not been
studied as yet.
Carotenoid glycosides were previously synthesized by direct glycosylation of
carotenoid alcohols using the classical Königs-Knorr procedure,4 and by total synthesis
starting from 3-hydroxy-β-ionone.5 However, these methods gave glycosides with rather low
(~3-8%) overall yields. Metabolic engineering using recombinant DNA techniques was also
applied for the synthesis of carotenoid glycosides.6 A new tentative approach for the
preparation of thioglycosides as thermoxanthin mimetics was reported in one of our previous
papers.7,8

Here we describe our results on using modern glycosylation methods for the synthesis of
carotenoid glycosides starting from some accessible hydroxy-carotenoids (Figure 1).
1
9
13
6
OH
9'
13'
3
5
8'-apo- -carotenol
5'
6'
3'
1
9
13
6
9'
9'
13'
3
1'
HO
HO
5
-cryptoxanthin
5'
6'
OH
3'
1
9
13
6
13'
3
1'
5
zeaxanthin
5'
6'
OH
3'
1
1
9
9
13
6
9'
9'
13'
13'
3
3
1'
HO
HO
5
lutein
O
13
6'
6
3'
OH
5
capsanthin
Figure. 1. Hydroxy-carotenoids used in the glycosylation reactions
Initially, the Königs-Knorr reaction starting from β-cryptoxanthin and tetra-O-acetyl-
β-D-glucopyranosyl bromide promoted by silver triflate, was reproduced and gave the target
monoglycoside in ~8% yield. However this does not represent improvement over the previous
reactions.
In our first attempts to improve the above yield we reacted the symmetrical diol
zeaxanthin, with tetra-O-acetyl-α-D-glucopyranosyl trichloroacetimidate 1 (synthesized in
three steps from D-glucose),9 which is a common, active, generally used glycosylating agent.
Different Brønsted and Lewis acids were applied (5-10 mol%) as catalysts in dry
dichloromethane. The results showed decomposition in almost all of the cases and the yields
were rather low except with p-toluenesulfonic acid (p-TsOH), where the monoglycoside 2
was obtained in an excellent 77% yield (Scheme 1, Table 1).

Scheme 1. Mono and diglucosides of zeaxanthin
In all the reactions, at least three equivalents of the sugar were applied and it was
added to the reaction mixture in two or three portions. Surprisingly, despite the excess of the
sugar donor, only a small amount (6%) of diglucoside 3 formed.
Table 1. Glycosylation experiments of zeaxanthin with tetra-O-acetyl-α-D-glucopyranosyl
trichloroacetimidate (1)
Entry
1
Catalyst/promoter
p-TsOH
Temp.
r.t.
Reaction time
overnight
Products and yields
2 (77%)
3 (6 %)
2 (10 %)
2
3
4
5
CSA
r.t.
overnight
30 min
20 min
24 h
BF3·OEt2
TMSOTf
CCl3-CHO
0 °C
0 °C
r.t.
2 (12%)
decomposed
no product observed
The Ferrier rearrangement of tri-O-acetyl-D-glucal (4) with zeaxanthin, using boron
trifluoride as the catalyst, delivered the deoxydiglycoside 5 exclusively. Other catalysts gave
only decomposition products (Scheme 2, Table 2.).

Scheme 2. Ferrier reaction with zeaxanthin
Table 2. Glycosylation experiments of zeaxanthin with tri-O-acetyl-D-glucal
Entry
Catalyst/promoter
ZnCl2
BF3·OEt2
Temp. Reaction time Products and yields
1
2
r.t.
overnight
30 min
decomposed
0 °C
5 (55%)
The above results with zeaxanthin showed that the best promoter for the glycosylation
with the trichloroacetimidate donor was p-TsOH, and for the Ferrier reaction was boron
trifluoride-diethyl etherate. These conditions were applied for other hydroxy-carotenoids with
(8'-apocarotenol and lutein) and without (β-cryptoxanthin and capsanthin) an allylic hydroxy
group.
Carotenoids with allylic hydroxy groups are prone to form carbocations especially with
Brønsted acids,10 so some by-products were to be expected. Unfortunately, in these
experiments the only reaction that occured was decomposition/dehydration of the carotenoid
(Table 3).
Table 3. Glycosylation experiments with other hydroxy-carotenoids
Carotenoid
Sugar
Catalyst/
promoter
Temp.
Reaction
time
Yield and products
(monoglycoside=MG
diglycoside=DG)
p-TsOH
p-TsOH
p-TsOH
p-TsOH
r.t.
r.t.
r.t.
r.t.
overnight 6 MG (8%)
+ decomposed
1
1
1
1
4
4
β-cryptoxanthin
capsanthin
overnight 7 MG (41%)
8 DG (22%)
overnight decomposed
8'-β-apocarotenol
lutein
overnight decomposed
BF3·OEt2 0 °C
BF3·OEt2 0 °C
30 min
30 min
9 MG in trace
amounts (<5%)
10 DG (32%)
β-cryptoxanthin
capsanthin

BF3 OEt2 0 °C
BF3·OEt2 0 °C
20 min
30 min
decomposed
decomposed
4
4
8'-β-apocarotenol
lutein
Capsanthin gave the monoglycoside (7) and the diglycoside (8)11 and also the deoxy
diglycoside (10) in moderate yields, whereas much to our surprise, in the case of β-
cryptoxanthin which bears only one hydroxy group, almost no product formation was
observed. It seems that the rate of cation formation and rearrangement of the non-
hydroxylated β-end group under the action of acids is comparable with that of the
glycosylation.
Based on our results, the dihydroxy carotenoids, zeaxanthin and capsanthin, are the
ideal starting materials for acid-promoted glycosylations to give the target glycosides in
moderate to good yields in a single step. Whereas natural and synthetic zeaxanthin glucosides
are already known, capsanthin glycosides or any κ-carotenoid glycosides have not been
described in Nature as yet. Considering the exceptional antioxidant properties of compounds
with a κ-end group among the carotenoids,12 it will be interesting to study how these
properties changed in the new derivatives.
Deprotection of the sugars delivers water-soluble carotenoid glycosides.
Unfortunately, this has not been achieved yet with acceptable yield. Hence a mild but
effective saponification method is still required, especially for the acid-labile
deoxyglycosides.
Our results show that modern glycosylation methods can deliver carotenoid glycosides
with moderate to good yields, but because of the acidic activation of the donor sugars these
methods are restricted to carotenoids bearing non-conjugated hydroxy group(s).
Acknowledgements
We thank Mrs. Judit Rigó and Mr. Roland Lukács for their assistance, Mr. Gergely
Gulyás-Fekete for the NMR spectra and Ms. Zsuzsanna Götz for HPLC measurements. This
study was supported by the grant OTKA K 83898 (Hungarian National Research Foundation).
References
1. Yokoyama, A.; Sandmann, G.; Hoshino, T.; Adachi, K.; Sakai, M.; Shizuri, Y. Tetrahedron
Lett. 1995, 36, 4901-4904.

2. Sliwka, H.R.; Melø, T.B.; Foss, B.J.; Abdel-Hafez, S.H.; Partali, V.; Nadolski, G.; Jackson,
H.; Lockwood, S.E. Chem. Eur. J. 2007, 13, 4458--4466.
3. Choi, S.K.; Osawa, A.; Maoka, T.; Hattan, J.; Ito, K.; Uchiyama, A.; Suzuki, M.; Shindo,
K.; Misawa, N.; Appl. Microbiol. Biotechnol. 2013, 97, 8479–8486.
4. Pfander, H., Hodler, M. Helv. Chim. Acta 1974, 57, 1641-1651.
5. Yamano, Y.; Sakai, Y.; Hara, M.; Ito, M. J. Chem. Soc., Perkin Trans. 1 2002, 2006-2013.
6. Yokoyama, A.; Shizuri, Y; Misawa, N. Tetrahedron Lett. 1998, 39, 3709-3712.
7. Nagy, V.; Agócs, A.; Turcsi, E.; Deli, J. Tetrahedron Lett. 2010, 51, 2020–2022.
8. Háda, M.; Nagy, V.; Deli, J.; Agócs, A.; Molecules 2012, 17, 5003-5012 (open access).
9. Schmidt, R.R.; Stump, M. Liebigs Ann. Chem. 1983, 7, 1249-1256.
10. Kildahl-Andersen, G.; Bruås, L.; Lutnaes, B. F. Org. Biomol. Chem. 2004, 2, 2496-2506.
Lutneas, B. J.; Kildahl-Andersen, G.; Krane, J.; Liaaen-Jensen, S. J. Am. Chem. Soc. 2004,
126, 8981–8990.
11. Synthesis of capsanthin glucosides: capsanthin (40 mg, 0.068 mM) was added to
trichloroacetimidate (350 mg, 0.71 mM) dissolved in dry CH2Cl2 (3 mL) at room temperature
under N2. After complete dissolution, dried p-TsOH (8 mg, 0.044 mM) was added. After
stirring for 6 h, trichloroacetimidate (175 mg, 0.35 mM) was added again to the solution
which was stirred overnight. When TLC showed complete conversion of capsanthin, the
mixture was diluted with Et2O (100 mL) and was washed with sat. NaHCO3 (40 mL) and
brine (40 mL). After drying over Na2SO4 and evaporation the two main products were
separated on a silica gel column. In most cases, subsequent purification on preparative TLC
(Merck, Kieselgel 60, eluent n-hexane:acetone 7:3) was necessary to obtain the pure red
products 7 (25 mg, 41%) and 8 (19 mg, 22%).
Capsanthin tetra-O-acetyl-β-D-glucoside (7): UV (λmax nm, EtOH): 476; IR (KBr pellet, cm-
1): 3319 m, 1726 vs (νCO, acetate), 1695 vs (conjugated νCO), 1616 s (νC=C), 1246, 1110 vs
(νC-O-C). MS (MALDI-TOF) m/z = 937.5 (M+Na). 1H-NMR (500 MHz, CDCl3) δ ppm: 0.84
(3H, s, 16'-Me), 1.08 (6H, s, 16,17-Me), 1.21 (3H, s, 17'-Me), 1.37 (3H, s, 18'-Me), 1.74 (3H,
s, 18-Me), 1.96-2.2 (30H, m, 19-Me, 19'-Me, 20-Me, 20'-Me, 2-H, 2'-H, acetyl-Me), 2.35
(1H, dd, J=17 Hz, 6, 4α-H), 3.01 (1H, dd, J=15.5 Hz, 9, 4'α-H), 3.99 (1H, m, 3-H), 4.37 (1H,
m, 3'-H), 4.23 (2H, m, 6''-H), 4.95 (1H, d, J=10 Hz, 4''-H), 5.20-5.32 (3H, m, 2''-H, 3''-H, 5''-
H), 5.74 (1H, d, J=6.0 Hz, 1''-H), 6.13 (2H, s, 7-H, 8-H), 6.16-6.75 (11H, m, olefin-Hs), 7.36
(1H, d, J=14 Hz, 8'-H). 13C-NMR (125 MHz, CDCl3): 12.7-12.85 (m), 21.5, 21.7, 25.1, 25.6,
28.6, 29.9, 37.0 (1-C), 40.8 (4-C), 43.4 (1'-C), 46.5 (4'-C), 49.0 (2-C), 50.6 (2'-C), 58.5 (5'-C),
63.2, 66.7 (3-C), 69.4, 71.9, 73.2, 73.8, 74.3, 96.9 (1''-C), 120.8, 125.9, 126.2, 129.8, 131.3,

131.6, 132.4, 133.6, 135.1, 135.7, 137.1, 137.3, 137.4, 138.5, 140.7, 142.0, 146.9, 169-170
(C=O, acetyl), 206.8 (6'-C). Analysis calcd. for C54H74O12: C 70.87, H 8.15, found C 70.73,
H 8.23.
Capsanthin di(tetra-O-acetyl-β-D-glucoside) (8): UV (λmax nm, EtOH): 475; IR (KRS-5
window, cm-1): 1738, 1731vs (νCO, acetate), 1674 s (conjugated νCO), 1599, 1581 s (νC=C),
1
1248, 1124 vs (νC-O-C). MS (MALDI-TOF) m/z=1267.4 (M+Na). H-NMR (500 MHz,
CDCl3) δ ppm: 0.84 (3H, s, 16'-Me), 1.1 (6H, s, 16,17-Me), 1.21 (3H, s, 17'-Me), 1.37 (3H, s,
18'-Me), 1.74 (3H, s, 18-Me), 1.96-2.2 (42H, m, 19-Me, 19'-Me, 20-Me, 20'-Me, 2-H, 2'-H,
acetyl-Me), 2.35 (1H, m, 4α-H), 2.95 (1H, m, 4'α-H), 4.04 (1H, m, 3-H), 4.38 (1H, m, 3'-H),
4.20 (4H, m, 6''-H), 4.95 (2H, m, 4''-H), 5.20-5.45 (6H, m, 2''-H, 3''-H, 5''-H), 5.72 (1H, d,
J=5.3 Hz, 1''-H), 5.75 (1H, d, J=5.5 Hz, 1''-H), 6.15 (2H, s, 7-H, 8-H), 6.16-6.75 (11H, m,
olefin-Hs), 7.35 (1H, d, J=13.5, 8'-H). 13C-NMR (125 MHz, CDCl3): 12.7-12.9 (m), 21.5,
21.6, 25.2, 25.7, 28.6, 29.9, 37.0 (1-C), 40.5 (4-C), 43.4 (1'-C), 46.5 (4'-C), 49.4 (2-C), 50.7
(2'-C), 58.5 (5'-C), 63.3, 67.5 (3-C), 69.7, 71.9, 73.3, 73.8, 74.3, 96.9 (d,1''-C), 120.9, 125.8,
126.1, 129.8, 131.3, 131.6, 132.5, 133.6, 135.1, 135.7, 137.1, 137.2, 137.4, 138.5, 140.8,
142.0, 146.8, 169-170 (C=O, acetyl), 206.7 (6'-C). Analysis calcd. for C68H92O21: C 65.58, H
7.45, found C 65.40, H 7.41.
12. Murillo, E; Nagy, V; Agócs, A; Deli, J: Carotenoids with κ-end group In: Carotenoids:
Food Sources, Production and Health Benefits, Yamaguchi, M.; Nova Science Publishers, Inc.
2013 Chapter 3. pp. 49-78 (open access).