Highly diastereoselective, biogenetically patterned synthesis of (+)- (1S,6R)-volvatellin (= (+)-(4R,5S)-5-hydroxy-4-(5-methyl-1-methylenehex-4-en- 2-ynyl)cyclohex-1-ene-1-carbaldehyde)

Article abstract of DOI:10.1002/(SICI)1522-2675(20000412)83:4<694::AID-HLCA694>3.0.CO;2-3

The synthesis of volvatellin (4a), previously isolated from a herbivorous marine mollusk, was achieved with high diastereoselectivity from putative dietary oxytoxin-1 (2). A biogenetically patterned carbonyl-ene route was chosen, proceeding from 2 predominantly via the trans cyclization product 3 without the use of enzymes. This challenges the involvement of enzymes in the formation of 4a in nature. The optical purity and absolute configuration (IS,4S,6R), assigned to 3 from high-field ¹H-NMR examination of its Mosher (MTPA) esters 6, was retained on its chemical conversion to (+)-(1S,6R)-configured 4a and is consistent with the (4S) configuration previously established for caulerpenyne (1).

Full text of DOI:10.1002/(SICI)1522-2675(20000412)83:4<694::AID-HLCA694>3.0.CO;2-3

694
Helvetica Chimica Acta ± Vol. 83 (2000)
Highly Diastereoselective, Biogenetically Patterned Synthesis
of (‡)-(1S,6R)-Volvatellin (ˆ(‡)-(4R,5S)-5-Hydroxy-4-(5-methyl-1-
methylenehex-4-en-2-ynyl)cyclohex-1-ene-1-carbaldehyde)
by Ines Mancinia)¹ Graziano Guellaa), and Francesco Pietraa)b)*
a
Á
) Laboratorio di Chimica Bioorganica, Universita di Trento, I-38050 Povo-Trento
b) Centro Linceo Interdisciplinare Beniamino Segre, via della Lungara 10, I-00165 Roma
The synthesis of volvatellin (4a), previously isolated from a herbivorous marine mollusk, was achieved with
high diastereoselectivity from putative dietary oxytoxin-1 (2). A biogenetically patterned carbonyl-ene route
was chosen, proceeding from 2 predominantly via the trans cyclization product 3 without the use of enzymes.
This challenges the involvement of enzymes in the formation of 4a in nature. The optical purity and absolute
1
configuration (1S,4S,6R), assigned to 3 from high-field H-NMR examination of its Mosher (MTPA) esters 6,
was retained on its chemical conversion to (‡)-(1S,6R)-configured 4a and is consistent with the (4S)
configuration previously established for caulerpenyne (1).
1. Introduction. ± Coevolution of seaweeds which produce distasteful metabolites
as a protection against herbivorous grazers, and shell-less opisthobranch mollusks
which use such algal products for their own defense, finds experimental support [1].
However, apart from clear-cut cases of uptake of algal metabolites, the role of these
mollusks as chemical machineries, either in transforming the algal products to tune
up their effects, or in de novo synthesis of distasteful metabolites, is far from being
clear [2].
Our work presented here challenges the hypothesis of the evolution of the enzyme
system of these mollusks in a specific, representative instance. This concerns the
sesquiterpenoid volvatellin (4a), which co-occur with caulerpenyne (1), which is
presumably taken up by the opisthobranch mollusk Volvatella sp. from the Indian coast
of the Bay of Bengal from the algal diet, and which was proposed to be formed
enzymatically from caulerpenyne-derived oxytoxin-1 (2) via the elusive intermediate 3
(Scheme 1) [3].
2. Results and Discussion. ± For the projected synthesis of volvatellin (4a), we had
at hand both caulerpenyne (1) and oxytoxin-1 (2), recently isolated alongside
glycoglycerolipids, and a-keto esters from a Mediterranean sample of the green alga
Caulerpa taxifolia [4a]. On the other hand, it has already been shown that 2 may be
obtained by the nonenzymatic transformation of 1 [4b]. Thus, oxytoxin-1 (2) was
treated with Lewis acids to give the trans-cyclization product 3 and its cis-
diastereoisomer 5 (Scheme 2) in good overall yields (ca. 70%, based on isolated
products after HPLC purification) and relative ratios that depended on the acidic
system used. In any case, the 3/5 ratio was always in favor of the trans product, as was
preliminarily established by integration of the H C(1) NMR signals of the crude

Helvetica Chimica Acta ± Vol. 83 (2000)
695
Scheme 1. Hypothetical Biosynthesis of Volvatellin (4a) from Dietary Caulerpenyne (1) in a Herbivorous
Mollusk
mixture and then confirmed by analysis of the isolated products. Thus, a 3/5 selectivity
ratio of 85 :15 was obtained with either ZnBr2 at room temperature or EtAlCl2 at
788. That the latter reaction is kinetically controlled was shown by comparative data
obtained at room temperature (see Exper. Part)1); in agreement, 5 remained
unchanged in the presence of ZnBr2 under stirring at r.t. for 8 h.
trans-Configuration of the substituents at the newly formed C(1) C(6) bond of 3,
and cis-configuration for 5, as well as the axial position for AcO C(4), firmly rest on
the J values for H C(1), H C(6), and H C(4) (see Exper. Part)2). The (Z)-
configuration at the enol acetate double bond was established by a NOE enhancement
observed between H C(3') and Heq C(2). The side-chain conformation (see
derivative 6 in Scheme 2) is supported by a NOE enhancement observed between
the signals at 5.40 and 2.55 ppm (H C(7') and H C(6), resp.; see Exper. Part). MM
Calculations correctly simulated this situation.
While volvatellin (4a) proved unstable [3], compound 3 ± being stable and in the
middle of the synthetic sequence (vide infra) ± looked promising for assignment of the
absolute configuration for all the compounds involved. Thus, the a-methoxy-a-
(trifluoromethyl)benzeneacetates (MTPA) 6a and 6b were prepared from 3 and either
( )-(R)-MTPA-Cl or (‡)-(S)-MTPA-Cl, respectively, as single diastereoisomers
(Scheme 2 and Exper. Part), thus establishing the enantiomer purity of 3. Examination
1
of 6a and 6b according to the high-field H-NMR Mosher method [5] allowed us to
1
1
2
)
Using the protic acid p-toluenesulfonic acid monohydrate, the selectivity ratio 3/5 was 55 :45 after ꢀ2 h at
room temperature.
Arbitrary numbering; for systematic names, see Exper. Part.
)

696
Helvetica Chimica Acta ± Vol. 83 (2000)
Scheme 2. Carbonyl-Ene Reaction of Oxytoxin-1 (2) to 3 and 5
a) ZnBr2, CHCl3, r.t., 8 h; 70%, 3/5 85 :15. b) EtAlCl2, CH2Cl2, 788, 1 h; 74%, 3/5 85 :15. c) EtAlCl2, CH2Cl2,
r.t., 5 min, 72%, 3/5 7:3. d) K2CO3, MeOH, r.t., 5 min; 90%. e) ( )-(R)-MTPA-Cl or (‡)-(S)-MTPACl, DMAP,
pyridine, r.t., overnight; 88%.
assign the absolute configuration (1S,4S,6R)2) to 33), in agreement with the
configuration (4S) attributed to caulerpenyne (1) [6].
Formation of 3 and 5 only, out of four possible diastereoisomers in the carbonyl-ene
reactions [7] of 2 (Scheme 3), with a bias towards 3, warrants some comment. The
ꢁaxialꢂ position of AcO C(4)2) in folded conformations4) of 2 ± which alleviates the
1,3-allylic strain [8][9] of the ꢁexoꢂ-methylene and acetoxy groups ± leads to four
unequally populated conformations 2a ± d. Treating the carbonyl-ene reaction as a
concerted process [7], it is only from 2a and 2b that formation of an incipient chair-like
cyclohexene ring can be imagined (rate-limiting formation [7] of transition states 2a=
3
)
A conformational analysis with the aid of MM calculations was carried out for the conformers derived by
rotation around the relevant bonds. Thus, in the least-strain-energy conformation of 6a, approximately
depicted in Scheme 2, the CˆO and C CF3 bonds deviate by ‡ 12 and 648, respectively, from the ideal
Mosher plane that should align them with the H C(1) bond [5]. For 6b, corresponding values of 168 and
‡ 608 were calculated. Although noticeable, this deviation of the C CF3 bond from the ideal plane does
not invalidate the Mosher analysis which, according to widely established models [5], could be carried out
1
with confidence in this case. It should also be mentioned that no splitting or broadening of the H-NMR
signals of the diastereoisomer 5 could be noticed on addition of the chiral shift reagent [Yb(hfbc)3] (see
Exper. Part). Even though not a proof per se, this is in line with the conclusions from the Mosher experiment
as to the enantiomer purity of 3.
MM Calculations suggested that elongated conformations entail less strain energy than folded
conformations, 2c and 2d turning out to be less strained than 2a and 2b. This implies that the product
distribution is not controlled by the relative strain of the least-strained conformers and, therefore, 2a and 2b
may be termed reactive conformations.
4
)

Helvetica Chimica Acta ± Vol. 83 (2000)
697
Scheme 3. Suggested Pathways for the Carbonyl-Ene Reaction of Oxytoxin-1 (2)
and 2b=). The corresponding transition states from 2c and 2d (2c= and 2d=) are in boat-
like incipient cyclohexene forms, which may explain why these two routes are
disfavored5). That formation of 3 is favored with respect to 5 is compatible with the
formal boat-like H-transfer in 2a= and chair-like-transfer in 2b=: the cases may be
reconciled by viewing the incipient carbocyclic ring as intermediate between a
cyclohexene and a cyclohexadiene form, where the classical energy values for chair/
Chair-like incipient cyclohexene rings for transition states 2c= and 2d= could also be drawn. However, they
would be very highly strained because of 1,3 repulsive interactions between the Me C(7) and pseudo-axial
AcO C(4) groups.
5
)

698
Helvetica Chimica Acta ± Vol. 83 (2000)
boat situations largely lack validity. Moreover, the route from 2a goes through a trans-
decaline structure, which is favored with respect to the cis-decaline structure of the
route from 2b6), in agreement with a) the preference for trans products in ab initio
calculations for the carbonyl-ene reaction of formaldehyde with propene [7c] and b)
observations for similar disubstituted (E)-C(6)ˆC(7) substrates [7d]7).
On treatment with K2CO3 in MeOH at room temperature, 3 gave immediately
dextrorotatory 4a, which is thus assigned the absolute configuration (1S,6R)2).
1H-NMR Spectra that are nicely superimposable with those of the natural volvatellin
[3] were observed for the semisynthetic 4a. The transformation 3 ! 4a was amazingly
clean8), provided that evaporation to dryness was avoided in the presence of K2CO3,
which would have triggered degradation of 4a.
It should be noted that optical rotations previously reported for both natural
volvatellin (4a) and its derivative 4b [3] are opposite in sign and larger in absolute value
with respect to our values (see Exper. Part). Our products are enantiomerically pure,
and the discordance of the chiroptical data for 4a can not be imputed to the different
solvents used since the discordance of chiroptical data is similar for 4b, where the same
solvent was used. Moreover, our data relate to a sequence of products that, starting
from (4S)-caulerpenyne (1), yielded pure 3, the absolute configuration of which was
independently determined. Therefore, we suggest that previous chiroptical data [3]
might have suffered from difficulties in the separation of the natural product from
highly optically active contaminants and in weighing the noncrystalline compounds
available in minute amounts. In any event, no matter what we have obtained ± either
naturally occurring volvatellin or its enantiomer, a question that the published data [3]
do not answer ± our conclusion remains that 4a is formed from 2 via a highly
6
)
)
MM Calculations indicated that 3 is definitely less strained (by 0.8 or 1.5 kcal mol 1 according to the force
field used, MM2 or MM3, resp.) than 5.
7
Parallel experiments were devised to become experimentally acquainted with the carbonyl-ene reaction.
We chose the reaction of (‡)-(R)-citronellal (8) since its carbonyl-ene conversion to ( )-isopulegol (9) is at
the basis of the industrial preparation of ( )-menthol, and thus, data for various Lewis and protic acids [10]
as well as superacids [11] as catalysts are available. Albeit more reactive than 2 in giving 3 and 5, (‡)-(R)-
citronellal (8) showed, under similar conditions [7d][10], much the same diastereoselectivity of ring closure
to 9 and 10 (Scheme 4; see Exper. Part). However, the similar behavior of 8 and 2 may be fortuitous since in
8, the H-transfer takes place predominantly from the Me group in trans position to the C-chain.
Scheme 4. Carbonyl-Ene Reaction of (‡)-(R)-Citronellal (8). Conditions similar to those in Scheme 2.
8
)
Addition of solid K2CO3 to a solution of either 3 or 5 in CD3OD in a NMR tube allowed the observation,
within a few minutes, of the 1-H-NMR spectrum of 4a or of the diastereoisomeric aldehyde 7, respectively,
each one not contaminated by the other one arising from complete hydrolysis of the corresponding
precursor (see Exper. Part). This further established the enantiomer purity of the compounds used and
obtained here, in agreement with the text and Footnote 3.

Helvetica Chimica Acta ± Vol. 83 (2000)
699
diastereoselective intramolecular carbonyl-ene cyclization. This challenges in this case
the hypothesis of the evolution of the enzyme system to transform dietary products in
the mollusk [3], where the presence of biological membranes and interfaces makes it
difficult to imagine and to simulate the state of the acids presumptively involved in the
carbonyl-ene reaction described here.
Experimental Part
1. General. Yields are given in terms of substrates reacted. All evaporations were carried out under reduced
pressure. Flash chromatography (FC): Merck Si-60 (15 ± 25 mm). TLC: Merck silica gel 60 PF254 and Merck RP-
18 F254. Prep. HPLC: Merck LiChrospher Si60, 25 Â 1 cm columns; UV monitoring at l 254 nm and solvent flux
5 ml min 1, if not otherwise stated; tR in min. Optical rotation: JASCO-DP-181 polarimeter, the integrity of
which was positively checked with known optically active compounds of high purity; [a] in deg cm2 g 1. NMR
Spectra: Varian-XL-300 spectrometer, 1H at 299.94 and 13C at 75.43 MHz; in CDCl3, if not otherwise stated; d in
ppm rel. to SiMe4 (ˆ0 ppm) and J in Hz; multiplicities from DEPT; 1H,1H correlations from COSY60 and
selective decoupling irradiations; 1H,13C assignments from 1H,13C-COSY; NOE (ˆ1D differential NOE)
reported as ꢁirradiated proton ! NOE observed on the proton(s)ꢂ. EI-MS: Kratos-MS80 mass spectrometer
equipped with a home-built computerized data system. MM Calculations: programs PCMODEL 4.0 from
Serena Software, Bloomingtone, Indiana, and MM3(96) from QCPE, Indiana University.
2. Isolation of Compounds. Oxytoxin-1 (2) was obtained by HPLC purification of a 2 :1 mixture with
caulerpenyne (1), as contained in Fr. 11 from FC (hexane/AcOEt/MeOH gradient) of the residue from
evaporation of the EtOH extracts of Caulerpa taxifolia from Elba Island [4a]).
3. Cyclizations of Oxytoxin-1 (2). 3.1. With ZnBr2. To a soln. of 2 [4a] (20.0 mg, 0.06 mmol) in either CHCl3
or CH2Cl2 (3 ml), ZnBr2 (1 equiv.) was added at r.t., and stirring was continued for 8 h until all 2 had
disappeared (TLC monitoring). The mixture was then evaporated and the residue submitted to FC (SiO2,
CHCl3/Et2O 1:1; TLC monitoring). The eluate of interest was subjected to HPLC purification (hexane/AcOEt
64 :36): 3 (tR 8.3; 11.9 mg, 60%) and 5 (tR 6.6; 2.1 mg, 10%).
3.2. With EtAlCl2. The reagents were mixed as described in 3.1, except that EtAlCl2 was used as Lewis acid
and CH2Cl2 as solvent and stirring at 788 for 1 h. The crude mixture was subjected to HPLC to give 3 (12.6 mg,
63%) and 5 (2.2 mg, 11%). The latter reaction was also carried out at r.t. for 5 min, under otherwise identical
conditions, to give 3 (10.1 mg, 51%) and 5 (4.3 mg, 21%).
(1S,2R,4S,5Z)-5-[(Acetyloxy)methylene]-2-(5-methyl-1-methylenehex-4-en-2-ynyl)cyclohexane-1,4-diol 4-
Acetate (3): colorless oil. [a]2D2 ˆ ‡14.6 (c ˆ 0.7, CHCl3). 1H-NMR2): 3.63 (ddd, J ˆ 5.2, 9.9, 11.2, H C(1));
2.31 (ddd, J ˆ 2.2, 11.2, 13.1, Hax C(2)); 2.44 (dd, J ˆ 4.4, 13.1, Heq C(2)); 5.92 (t, J ˆ 2.8, H C(4)); 1.78
(ddd, J ˆ 2.8, 12.9, 15.0, Hax C(5)); 2.00 (ddd, J ˆ 2.8, 3.9, 15.0, Heq C(5)); 2.55 (ddd, J ˆ 3.9, 9.9, 12.6,
H
C(6)); 5.35 (sept., J ˆ 1.3, H C(10)); 1.88 (s, 3 H C(12)); 7.00 (d, J ˆ 2.1, H C(3')); 5.40 (d, J ˆ 1.9,
Ha C(7')); 5.46 (d, J ˆ 1.9, Hb C(7')); 1.81 (s, 3 H C(11')); 2.05 (s, AcO C(4)); 2.16 (s,AcO C(3')).
13C-NMR2): 71.08 (d, C(1); 33.44 or 34.24 (t, C(2)); 118.75 (s, C(3)); 64.97 (d, C(4)); 34.24 or 33.44 (t, C(5));
48.38 (d, C(6)); 131.70 (s, C(7)); 90.03 (s, C(8)); 88.56 (s, C(9)); 104.83 (d, C(10)); 149.70 (s, C(11)); 24.88
(q, C(12)); 131.56 (d, C(3')); 123.14 (t, C(7')); 21.20 (q, C(11')); 169.98 (s, MeCOO C(4)); 167.89
(s, MeCOO C(3')); 20.07 (q, MeCOO C(4)); 21.27 (q, MeCOO C(3')). NOE2): 3.63 (H C(1)) !
Heq C(2), Hax C(5), 5.92 (H C(4)) ! Hax C(5); 5.36 (H C(10)) ! 3 H C(11'); 7.00 (H C(3')) !
.
.
‡
‡
Heq C(2); 5.40 (H C(7')) ! H C(6). MS: 332 (3, M ), 290 (2.6, [M CH2CO] ), 272 (1.8, [M
.
.
‡
‡
CH3COOH] ), 230 (9.3, [M CH2CO CH3COOH] ), 212 (9), 201 (7.2), 183 (10), 169 (8.4), 43 (100).
.
‡
HR-MS: 332.1621 Æ 0.002 (C19H24O5 ; calc. 332.1624).
(1R,2R,4S,5Z)-5-[(Acetyloxy)methylene]-2-(5-methyl-1-methylenehex-4-en-2-ynyl)cyclohexane-1,4-diol 4-
Acetate (5): colorless oil. [a]D22
ˆ
5.3 (c ˆ 0.2, CHCl3). 1H-NMR2): 4.26 (q, J ˆ 2.2, H C(1)); 2.56 (dd, J ˆ 3.3,
14.4, Hax C(2)); 2.28 (dd, J ˆ 3.3, 14.4, Heq C(2)); 6.05 (t, J ˆ 2.9, H C(4)); 2.03 (ddd, J ˆ 2.9, 12.6, 15.4,
Hax C(5)); 1.88 (ddd, J ˆ 2.9, 3.7, 15.1, Heq C(5)); 2.73 (br. ddd, J ˆ 2.2, 3.7, 12.6, H C(6)); 5.36 (sept., J ˆ 1.3,
H
C(10)); 1.90 (s, 3 H C(12)); 6.99 (d, J ˆ 2.2, H C(3')); 5.49 (br. s, Ha C(7')); 5.24 (br. s, Hb C(7')); 1.82
(s, 3 H C(11')); 2.04 (s, AcO C(4)); 2.16 (s, AcO C(3')). 1H-NMR Experiment with [Yb(hfbc)3]: to a soln. of
5 in CDCl3 (0.6 ml) was added 0.042m [Yb(hfbc)3] in CDCl3 in 5-ml portions; at 0.2 mol-equiv. of [Yb(hfbc)3],
the following downfield shifts were observed2): 0.9 ppm for H C(1), 0.6 ppm for H C(4), 0.4 ppm for
Hax C(5), and 0.3 ppm for both 2 H C(2) and H C(6); no doubling or broadening of either these or other

700
Helvetica Chimica Acta ± Vol. 83 (2000)
signals could be noticed. 13C-NMR2): 67.92 (d, C(1)); 28.51 (t, C(2)); 116.27 (s, C(3)); 65.46 (d, C(4)); 33.04
(t, C(5)); 42.78 (d, C(6)); 132.40 (s, C(7)); 90.80 (s, C(8)); 89.19 (s, C(9)); 104.76 (d, C(10)); 149.74 (s, C(11));
24.92 (q, C(12)); 132.12 (d, C(3')); 121.16 (t, C(7')); 21.22 (q, C(11')); 170.02 (s, MeCOO C(4)); 167.71
.
‡
(s, MeCOO C(3')); 20.70 (q, MeCOO C(4)); 21.31 (q, MeCOO C(3')). EI-MS: 332 (2, M ), 290 (1, [M
.
.
.
‡
‡
‡
CH2CO] ), 272 (1.5, [M CH3COOH] ), 230 (7.5, [M CH2CO CH3COOH] ), 212 (9), 201 (7), 183
(10), 169 (7), 43 (100).
4. Cyclizations of (‡)-(R)-Citronellal (8). 4.1. With ZnBr2. To a soln. of 8 (30.0 mg, 0.19 mmol) in CH2Cl2
(4 ml), ZnBr2 (1 equiv.) was added at r.t. and stirring was continued for 30 min until all 8 had disappeared (TLC
monitoring). Precipitated ZnBr2 was filtered off and the eluate evaporated: 9/10 9 :1, as established by
integration of 1H-NMR signals [10].
4.2. With EtAlCl2. The reagents were mixed as described in 4.1., except for using EtAlCl2 as Lewis acid.
After a few minutes, complete conversion of 8 to 9/10 7 :3 was observed.
5. Alkaline Hydrolysis of 3 and 5. 5.1. To a soln. of 3 (2.3 mg, 0.007 mmol) in MeOH (2 ml), solid K2CO3
was added. Complete conversion of 3 to volvatellin (4a) was observed (TLC monitoring) within a few minutes.
The mixture was passed through a strong anion-exchanger (Merck LiChrolut SAX (500 mg), H2O (elimination
of salts), then MeOH): 4a (2.1 mg, 90%). HPLC (hexane/AcOEt 96 :4, flow 1 ml min 1; l 230 nm) gave
(4R,5S)-5-hydroxy-4-(5-methyl-1-methylenehex-4-en-2-ynyl)cyclohex-1-ene-1-carbaldehyde (4a; tR 9.2) of high
20
purity for polarimetric measurements. [a]2D0 ˆ ‡7, [a] ˆ ‡43 (c ˆ 0.1, CHCl3) ([3]: [a]D
ˆ
88.3 (c ˆ 0.04,
577
‡
1
Et2O)). H-NMR and low-resolution MS: in accordance with [3]. HR-MS: 230.1302 Æ 0.002 (C15H18O2 ; calc.
230.1307).
5.2. On addition of solid K2CO3 to a soln. of 3 or 5 (2 mg) in CD3OD (0.6 ml) in a NMR tube, clean and
complete formation of 4a or the diastereoisomeric aldehyde 7, respectively, was observed. 1H-NMR (CD3OD;
only the most relevant spectral differences are reported2): 3.90 (dt, J ˆ 5.5, 9.1, H C(1) (4a)); 4.40 (dt, J ˆ 3.6,
2.6, H C(1) (7)); 6.88 (m, H C(4) (4a)); 6.97 (m, H C(4) (7)); 2.47 (dt, J ˆ 5.8, 9.5, H C(6) (4a)); 2.39
(dt, J ˆ 2.0, 3.8, H C(6) (7)); 5.35, 5.38 (2 br. s, 2 H C(7') (4a)); 5.37, 5.45 (2 br. s, 2 H C(7') (7)).
6. Acetylation of 4a. Aldehyde 4a (1.8 mg, 0.008 mmol) was treated with excess Ac2O in dry pyridine
(0.5 ml) under stirring at r.t. overnight. Then a sat. aq. CuSO4 soln. (1 ml) and AcOEt (1 ml) were added, and
the mixture was passed through a Whatman phase-separation filter. The org. phase was evaporated and
subjected to prep. TLC (hexane/AcOEt 7:3): pure 4b (2.0, 94%). [a]2D0 ˆ ‡15, [a]577 ˆ ‡54 (c ˆ 0.1, CHCl3)
([3]: [a]D
ˆ
57.2 (c ˆ 0.08, CHCl3). 1H-NMR: in accordance with [3].
7. Synthesis of the MTPA Esters 6a and 6b. Compound 3 (5 mg; 0.015 mmol) was treated with ( )-(R)-
MTPA-Cl (3 mol-equiv.) and 4-(dimethylamino)pyridine (1.0 mg) in dry pyridine (0.5 ml) under stirring for
24 h at r.t. The mixture was then quenched with sat. aq. CuSO4 soln. (1 ml), followed by Et2O (4 ml), and passed
through a Whatman phase-separation filter. The org. phase was evaporated and subjected to prep. TLC (hexane/
AcOEt 7:3): 6a (7.3 mg, 88%). By a similar procedure, the same amounts of 3 and (‡)-(S)-MTPA-Cl gave 6b
(7.4 mg, 88%). NMR and TLC examination of the crude reaction mixtures showed that both 6a and 6b were
obtained as single diastereoisomers.
(aS)-a-Methoxy-a-(trifluoromethyl)benzeneacetic Acid (1S,2R,4S,5Z)-4-(Acetyloxy)-5-[(acetyloxy)methy-
lene]-2-(5-methyl-1-methylenehex-4-en-2-ynyl)cyclohexyl Ester (6a). 1H-NMR2): 5.10 (td, J ˆ 10.7, 4.8,
H
H
C(1)); 2.25 (ddd, J ˆ 2.3, 10.7, 13.1, Hax C(2)); 2.60 (dd, J ˆ 4.8, 13.1, Heq C(2)); 5.92 (t, J ˆ 2.8,
C(4)); 1.85 (ddd, J ˆ 2.8, 12.8, 14.8, Hax C(5)); 2.05 (ddd, J ˆ 2.8, 3.8, 14.8, Heq C(5)); 2.85 (ddd, J ˆ 3.8,
10.7, 12.8, H C(6)); 5.37 (sept., J ˆ 1.3, H C(10)); 1.86 (s, 3 H C(12)); 7.06 (d, J ˆ 2.0, H C(3')); 5.33, 5.30
(2d, J ˆ 1.8, 2 H C(7')); 1.81 (s, 3 J ˆ 11')); 2.04 (s, AcO C(4)); 2.17 (s, AcO C(3')); 7.50 (m, 2 arom. H); 7.38
(m, 3 arom. H); 3.54 (q, 5J(H,F) ˆ 1.1, MeO). 13C-NMR2): 75.68 (d, C(1)); 30.56 (t, C(2)); 116.91 (s, C(3));
64.52 (d, C(4)); 35.11 (t, C(5)); 44.49 (d, C(6)); 132.64 (s, C(7)); 90.30 (s, C(8)); 88.09 (s, C(9)); 104.74
(d, C(10)); 149.80 (s, C(11)); 24.95 (q, C(12)); 132.67 (d, C(3')); 122.65 (t, C(7')); 21.24 (q, C(11')); 169.86
(s, MeCOO C(4)); 167.67 (s, MeCOO C(3')); 20.66 (q, MeCOO C(4)); 21.24 (q, MeCOO C(3')); 165.79
(s, ROCO C(1)); 84.60 (q, 2J(C,F) ˆ 28, CCO C(1)); 123.12 (q, 1J(C,F) ˆ 288, CF3); 55.72 (s, MeO); 127.30
.
‡
(d, 2 arom. C); 128.28 (d, 2 arom. C); 129.52 (d, arom. C); 129.65 (s, arom. C). EI-MS: 548 (8, M ), 506 (3),
488 (3), 229 (7), 213 (15), 212 (40), 77 (12), 43 (100).
(aR)-a-Methoxy-a-(trifluoromethyl)benzeneacetic Acid (1S,2R,4S,5Z)-4-(Acetyloxy)-5-[(acetyloxy)methy-
lene]-2-(5-methyl-1-methylenehex-4-en-2-ynyl)cyclohexyl Ester (6b). 1H-NMR2): 5.13 (td, J ˆ 10.7, 4.8,
H
H
C(1)); 2.45 (ddd, J ˆ 2.2, 11.2, 13.1, Hax C(2)); 2.63 (dd, J ˆ 4.8, 13.1, Heq C(2)); 5.93 (t, J ˆ 2.8,
C(4)); 1.85 (ddd, J ˆ 2.8, 12.6, 15.0, Hax C(5)); 2.04 (ddd, J ˆ 2.8, 3.9, 15.0, Heq C(5)); 2.82 (ddd, J ˆ 3.9,
10.7, 12.6, H C(6)); 5.38 (sept., J ˆ 1.3, H C(10)); 1.89 (s, 3 H C(12)); 7.07 (d, J ˆ 2.1, H C(3')); 5.08, 5.07
(2d, J ˆ 1.7, 2 H C(7')); 1.82 (s, 3 H C(11')); 2.07 (s, AcO C(4)); 2.17 (s, AcO C(3')); 7.50, 7.38 (2m,

Helvetica Chimica Acta ± Vol. 83 (2000)
701
5 arom. H); 3.51 (q, 5J(H,F) ˆ 1.1, MeO). 13C-NMR2): 75.27 (d, C(1)); 31.02 (t, C(2)); 116.94 (s, C(3)); 64.56
(d, C(4)); 34.96 (t, C(5)); 44.67 (d, C(6)); 132.19 (s, C(7)); 89.94 (s, C(8)); 88.21 (s, C(9)); 104.90 (d, C(10));
149.52 (s, C(11)); 24.94 (q, C(12)); 132.63 (d, C(3')); 122.87 (t, C(7')); 21.25 or 21.19 (q, C(11')); 169.84
(s, MeCOO C(4)); 167.64 (s, MeCOO C(3')); 20.66 (q, MeCOO C(4)); 21.19 or 21.25 (q, MeCOO C(3'));
165.52 (s, ROCO C(1)); 84.59 (q, 2J(C,F) ˆ 28, CCO C(1)); 123.20 (q, 1J(C,F) ˆ 288, CF3); 55.44 (s, MeO);
127.44 (d, 2 arom. C); 128.21 (d, 2 arom. C); 129.44 (d, arom. C); 130.57 (s, arom. C).
REFERENCES
[1] V. J. Paul, ꢁEcological Roles of Marine Natural Productsꢂ, Cornell University Press, Ithaca, N.Y., 1992.
[2] M. J. Garson, Nat. Prod. Rep. 1989, 6, 143; M. T. Davies-Coleman, M. J. Garson, Nat. Prod. Rep. 1998, 15,
477.
[3] A. Fontana, M. L. Ciavatta, E. Mollo, C. G. Naik, S. Wahidulla, L. DꢂSouza, G. Cimino, J. Nat. Prod. 1999,
62, 931.
[4] a) I. Mancini, G. Guella, A. Defant, M. L. Candenas, C. P. Armesto, D. Depentori, F. Pietra, Helv. Chim.
Acta 1998, 81, 1681; b) A. Guerriero, D. Depentori, M. DꢂAmbrosio, F. Pietra, Helv. Chim. Acta 1995, 78,
1759.
[5] I. Ohtani, T. Kusumi, Y. Kashman, H. Kakisawa, J. Am. Chem. Soc. 1991, 113, 4092.
[6] V. Amico, G. Oriente, M. Piattelli, C. Tringali, E. Fattorusso, S. Magno, L. Mayol, Tetrahedron Lett. 1978,
3593.
[7] a) W. Oppolzer, V. Snieckus, Angew. Chem. 1978, 90, 506; b) B. B. Snider, Acc. Chem. Res. 1980, 13, 426;
c) R. J. Loncharich, K. N. Houk, J. Am. Chem. Soc. 1987, 109, 6947; d) B. B. Snider, M. Karras, R. T. Price,
D. J. Rodini, J. Org. Chem. 1982, 47, 4538.
[8] R. W. Hoffmann, Chem. Rev. 1989, 89, 1841.
[9] G. Guella, F. Marchetti, F. Pietra, Helv. Chim. Acta 1997, 80, 684.
[10] Y. Nakatani, K. Kawashima, Synthesis 1978, 147.
[11] Y.-R. Naves, P. Ochsner, Helv. Chim. Acta 1964, 47, 5; Y. Nakatani, K. Kawashima, Synthesis 1995, 78, 1755;
G. D. Yadav, J. J. Nair, Chem. Commun. 1998, 2369.
Received November 15, 1999