Identification of a precursor to naturally occurring β-damascenone

Article abstract of DOI:10.1016/S0040-4039(01)01411-3

9-Hydroxymegastigma-3,5-dien-7-yne 8a was synthesised and shown to be identical to an intermediate found in the acid-catalysed conversion of 3,5,9-trihydroxymegastigma-6,7-diene 4 to β-damascenone 1, 3-hydroxydamascone 5 and megastigma-5-en-7-yne-3,9-diol 6. When subjected to acid hydrolysis, 8a produced β-damascenone 1, in high yield. Importantly, the hydrolysate was completely free of 3-hydroxydamascone 5.

Full text of DOI:10.1016/S0040-4039(01)01411-3

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 6937–6939
Identification of a precursor to naturally occurring
b-damascenone†
Carolyn J. Puglisi,a,b Gordon M. Elsey,a,* Rolf H. Prager,b George K. Skouroumounisa and
Mark A. Seftona
aThe Australian Wine Research Institute, PO Box 197, Glen Osmond, South Australia 5064, Australia
bSchool of Chemistry, Physics and Earth Sciences, Flinders University, PO Box 2100, Adelaide, South Australia 5001, Australia
Received 6 June 2001; accepted 27 July 2001
Abstract—9-Hydroxymegastigma-3,5-dien-7-yne 8a was synthesised and shown to be identical to an intermediate found in the
acid-catalysed conversion of 3,5,9-trihydroxymegastigma-6,7-diene
megastigma-5-en-7-yne-3,9-diol 6. When subjected to acid hydrolysis, 8a produced b-damascenone 1, in high yield. Importantly,
the hydrolysate was completely free of 3-hydroxydamascone 5. © 2001 Elsevier Science Ltd. All rights reserved.
4 to b-damascenone 1, 3-hydroxydamascone 5 and
1. Results and discussion
that b-damascenone 1 can be formed in vivo, as shown
in Scheme 1.
b-Damascenone 1 is one of the most important flavour
compounds known to science;1 it is found in many fruit
and vegetable products, and tonne quantities are pro-
duced commercially for the perfume and flavouring
industries. It is believed to be formed in nature from the
OR
OH
•
HO
hydrolytic
breakdown
of
complex
secondary
metabolites2–6 derived from carotenoids such as neox-
7
8a, R=H
8b, R=Ac
anthin 2. Recent studies of wine grapes have indicated
O
•
•
O
OH
HO
OH
HO
OH
2
3
OH
O
O
OH
•
+
+
OH
HO
HO
HO
6
1
5
4
Scheme 1.
Keywords: b-damascenone; flavour precursors; wine; carotenoid metabolites.
* Corresponding author. E-mail: gelsey@awri.adelaide.edu.au
† All authors are members of the Cooperative Research Centre for Viticulture, PO Box 154, Glen Osmond, South Australia 5064.
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01411-3

6938
C. J. Puglisi et al. / Tetrahedron Letters 42 (2001) 6937–6939
OR
OR
OH
O
O
a)-b)
c)-d)
e)-f)
11a, R=H
8a, R=H
9
10
11b, R=Ac
8b, R=Ac
Scheme 2. (a) LDA, MeI (84%); (b) i. Br2, ii. pyr, D (79%); (c) LiCꢀCCH(CH3)OLi (99%); (d) Ac2O, THF, D (88%); (e)
P2O5/Celite, PhCH3, D (54%); (f) KOH (78%).
Both the so-called ‘grasshopper ketone’ 37 and the
allenic triol 46 have been identified as components of
wine grape extracts treated with a glycosidase enzyme
preparation. Hydrolytic studies8 have shown that triol 4
is converted rapidly to the products 1, 5 and 6 at room
temperature and pH 3.0 in an aqueous environment.
Furthermore, the ratio of these three products in at
least some grape samples9 was broadly similar to that
found in the hydrolysates of the triol 4. b-Damascenone
1, 3-hydroxydamascone 5 and enyne diol 6 are appar-
ently formed by competing pathways from the triol 4,
as no significant interconversion of these products takes
place at room temperature. The diol 6 (and also its C9
glucopyranoside)‡ will form both 1 and 5 extremely
slowly under these conditions, and therefore may gener-
ate b-damascenone 1 in wine over several years of
bottle ageing, but is not a significant precursor to 1 in
either grapes or young wines.
Hydrolysis of 8a was effected in model wine,15 and
produced b-damascenone 1 as the major product
(>90%) by GC/MS. The absence of 3-hydroxydamas-
cone 5 and the enyne diol 6 in the hydrolysate confirms
the hypothesis that 1, 5 and 6 are formed from the triol
4 by competing pathways.
Compounds 3–6 are known to accumulate in grapes
and other fruits as glycoconjugates. Studies of the
reactivity of allylic and propargyllic alcohols and their
corresponding b-D-glucopyranosides have shown that
the glucopyranosyl moiety retards the acid-catalysed
hydrolysis of these compounds. Thus, in plant products
containing glycoconjugated forms of the allenic triol 4,
the position of glycoconjugation has the capacity to
steer hydrolysis either towards or away from b-damas-
cenone formation by promoting or suppressing forma-
tion of the intermediate 8a in competition with other
products.4,5
In the early stages of hydrolysis, two intermediates in
the conversion of 4 to 1, 5 and 6 were observed by
GC/MS and tentatively identified as 7 and 8a. The
latter has also been reported (without presentation of
evidence) as a constituent of rum.10 We now wish to
report the successful synthesis and characterisation of
8a and its role as an intermediate in the formation of
b-damascenone.
Acknowledgements
We would like to thank the Grape and Wine Research
and Development Corporation and the Cooperative
Research Centre for Viticulture for funding of this
work. Mrs. D. Capone and Dr. A. Pollnitz are
acknowledged for their assistance with the mass spec-
tral work, as are Dr. M. Perkins and Professor P. Høj
for helpful discussions.
The synthesis of 8a was accomplished in six steps
(Scheme 2) from 2,6-dimethylcyclohexanone 9. Methyl-
ation, then bromination followed by dehydrobromina-
tion gave the enone 10, which underwent reaction with
the dilithio derivative11 of 2-hydroxybut-3-yne. Acetyla-
tion of the secondary hydroxy group in 11a gave 11b,
which was dehydrated with P2O5 on Celite.12 Treatment
of 8b13 with potassium hydroxide under mild conditions
produced the free alcohol 8a,14 which was shown by
GC/MS (retention time, fragmentation pattern and co-
injection) to be identical with one of the two com-
pounds observed as intermediates in the hydrolysis of
the triol 4.
References
1. Pickenhagen, W. In Flavor Chemistry—Thirty Years of
Progress; Teranishi, E. L. W.; Hornstein, I., Eds.; Kluwer
Academic/Plenum Publishers: New York, 1999; pp. 75–
87.
2. Ohloff, G.; Rautenstrauch, V.; Schulte-Elte, K. H. Helv.
Chim. Acta 1973, 56, 1503.
3. Isoe, S.; Katsumura, S.; Sakan, T. Helv. Chim. Acta 1973,
56, 1514.
4. Skouroumounis, G. K.; Sefton, M. A. J. Agric. Food
‡ The numbering scheme is that employed for the megastigmane
skeleton.
Chem. 2000, 48, 2033.
5. Skouroumounis, G. K.; Sefton, M. A. In Carotenoid-
derived Aroma Compounds; Winterhalter, P.; Rouseff, R.,
Eds.; ACS Symp. Ser., American Chemical Society:
Washington, DC, 2001, in press.
11
12
8
10
1
4
7
9
6
5
2
3
13
6. Sefton, M. A. Aust. J. Grape Wine Res. 1998, 4, 30.

C. J. Puglisi et al. / Tetrahedron Letters 42 (2001) 6937–6939
6939
7. Meinwald, J.; Erickson, K.; Hartshorn, M.; Meinwald,
Y. C.; Eisner, T. Tetrahedron Lett. 1968, 9, 2959.
8. Skouroumounis, G. K.; Massy-Westropp, R. A.; Sefton,
M. A.; Williams, P. J. Tetrahedron Lett. 1992, 33, 3533.
9. Sefton, M. A.; Francis, I. L.; Williams, P. J. Am. J. Enol.
Vitic. 1993, 44, 359.
10. ter Heide, R.; Schaap, H.; Wobben, H. J.; De Valois, P.
J.; Timmer, R. Qual. Foods Beverages: Chem. Technol.
1981, 1, 183.
Hz, H9); 2.10–2.07 (2H, m, H2); 2.08 (3H, s, OAc); 1.92
(3H, br s, H13); 1.54 (3H, d, J=6.7 Hz, H10); 1.06 (3H, s,
H11); 1.05 (3H, s, H12). 13C NMR (CDCl3, 75.5 MHz),
169.9 (CO); 137.4, 127.6, 126.8, 122.6 (C3, C4, C5, C6);
94.7, 83.2 (C7, C8); 61.3 (C9); 38.2 (C2); 32.7 (C1); 26.9,
26.9 (C11, C12); 21.8 (C10); 21.2 (CH3CO); 20.5 (C13).
14. Selected spectral data for (8a): 5.86–5.76 (2H, m, H3,
H4); 4.75 (1H, q, J=6.6 Hz, H9); 2.11–2.07 (2H, m, H2);
1.93 (3H, br s, H13); 1.52 (3H, d, J=6.6 Hz, H10); 1.06
(6H, s, H11, H12). 13C NMR (CDCl3, 75.5 MHz), 136.5,
127.6, 126.4, 122.9 (C3, C4, C5, C6); 98.5, 82.1 (C7, C8);
58.8 (C9); 38.2 (C2); 32.6 (C1); 26.8, 26.8 (C11, C12); 24.7
(C10); 20.3 (C13).
11. Demole, E.; Enggist, P. Helv. Chim. Acta 1974, 57,
2087.
12. Burton, S. G.; Kaye, P. T. Synth. Commun. 1989, 19,
3331.
13. Selected spectral data for (8b): 1H NMR (CDCl3, 300
MHz), 5.88–5.76 (2H, m, H3, H4); 5.66 (1H, q, J=6.7
15. Model wine: 10% (v/v) EtOH in H2O, buffered at pH
3.0. Hydrolysis was conducted at 45°C overnight.