Two remarkable epimerizations imperative for the success of an asymmetric total synthesis of (+)-aigialospirol

Article abstract of DOI:10.1007/s11426-010-4176-8

Two remarkable epimerization processes were uncovered during our pursuit of an enantioselective synthesis of (+)-aigialospirol featuring a cyclic acetal tethered ring-closing metathesis. Through modeling, we were able to turn these two unexpected epimeriz

Full text of DOI:10.1007/s11426-010-4176-8

SCIENCE CHINA
Chemistry
January 2011 Vol.54 No.1: 31–42
doi: 10.1007/s11426-010-4176-8
• ARTICLES •
· SPECIAL TOPIC · The Frontiers of Chemical Biology and Synthesis
Two remarkable epimerizations imperative for the success of
an asymmetric total synthesis of (+)-aigialospirol
FIGUEROA Ruth1, FELTENBERGER John B.1, GUEVARRA Christle C.1 &
HSUNG Richard P.1,2*
1Division of Pharmaceutical Sciences and Department of Chemistry, University of Wisconsin, Madison, WI 53705, USA
2School of Pharmaceutical Science and Technology, Tianjin University, Tianjin 300072, China
Received July 29, 2010; accepted August 24, 2010
Two remarkable epimerization processes were uncovered during our pursuit of an enantioselective synthesis of (+)-aigia-
lospirol featuring a cyclic acetal tethered ring-closing metathesis. Through modeling, we were able to turn these two unex-
pected epimerizations to our advantage via modeling to ensure a successful and concise total synthesis, thereby firmly estab-
lishing cyclic acetal tethered RCM as a powerful strategy in natural product synthesis. Most importantly, calculations allowed
us to fully understand the nature and the mechanistic course of these two epimerizations that were imperative to the total syn-
thesis efforts.
cyclic acetal, tethered RCM, (+)-aigialospirol, anomeric effect, and epimerization.
1 Introduction
In the last five years, we have been exploring chemistry of
cyclic acetals [1, 2], which can serve either as an activator
of 2-components for cationic cycloadditions via vinyl
oxocarbenium ions [3], or as a unique tether for in-
tramolecular transformations that could lead to a useful
strategy for constructing spiroketals (Figure 1) [4]. This
program provided an invaluable opportunity for us to estab-
lish a fundamentally different, yet competitive, approach
[5–7] compared to the classical spiroketal construction [4],
while allowing us to explore the impact of the anomeric
effect (~1.5 kcal/mol per anomeric effect) [8] on new reac-
tivities [5, 9] and stereochemical issues [6, 9].
Figure 1 A cyclic acetal tethered strategy for spiroketal synthesis.
tation as protecting groups for carbonyls and alcohols and
their perceived instability under hydrolytic conditions.
However, we recognized that developing cyclic acetal teth-
ered reactions would represent a fertile ground for discov-
ering new synthetic strategies. Endeavors in pursuing
strategies and approaches that are unconventional or even
untapped have historically led to fertile grounds for uncov-
With exception of very few precedents [10–12], there
have been no general studies on cyclic acetal tethered reac-
tions to construct spiroketals. The lack of interest in using
acetals as tethers could be due to their overwhelming repu-
*Corresponding author (email: rphsung@pharmacy.wisc.edu)
© Science China Press and Springer-Verlag Berlin Heidelberg 2011
chem.scichina.com
www.springerlink.com

32
Figueroa R, et al. Sci China Chem January (2011) Vol.54 No.1
ering new chemistry and for the advancement of synthesis.
Of all the methods that have been developed, cyclic
acetal tethered RCM [13] appears to be the most powerful
method (Scheme 1). Combining the prevalence of spiroket-
als in biologically active natural products [1, 2] and the
power of RCM has rendered this a fundamentally different
approach truly significant in natural product synthesis [7,
14]. More specifically, we became involved with the total
synthesis of (+)-aigialospirol 4 [15] to demonstrate a point
that the cyclic acetal tethered RCM strategy via 5 can lead
to a competitive if not superior total synthesis than the clas-
sical spiroketal formation through diol-ketone 6.
stereochemistry at C1′. Transformations in the reverse di-
rection (4  8 and 8  7) are not known. We had recently
completed the first total synthesis of (+)-aigialospirol 4 [23]
featuring the cyclic acetal tethered RCM strategy. We wish
to disclose here in this full paper two very interesting epim-
erizations that were key to the success of this total synthesis
endeavor.
2 Experimental
2.1 General
The isolation of (+)-aigialospirol 4 was reported by Isaka
et al. [15]. It was obtained after an extended fermentation of
the marine fungus Aigialus parvus BCC 5311 that was
found in the mangrove Ascomycete. Although (+)-aigialospirol
4 has not been implicated with any biological activities [16],
it is biosynthetically related to (+)-hypothemycin 7 [17–20],
which was isolated from the same fungus source.
(+)-Hypothemycin 7 is the well-known for its potent anti-
malarial [21] and anticancer [22] activities.
Specifically, 4 is postulated [15] to be derived from
macro-lactone 8 via two steps, trans-lactonization and spi-
roketal formation. The macro-lactone 8 can be derived from
a hydrative opening of the epoxide in 7 with inversion of
All reactions were performed in flame-dried glassware un-
der a nitrogen atmosphere. Solvents were distilled prior to
use. Reagents were used as purchased (Aldrich, Acros),
except where noted. Chromatographic separations were
1
performed using Bodman 60 Å SiO2. H and 13C NMR
spectra were obtained on Varian VI-300, VI-400, and VI-500
spectrometers using CDCl3 (except where noted) with TMS
or residual CHCl3 in the solvent as standard. Melting points
were determined using a Laboratory Devices MEL-TEMP
and are uncorrected/calibrated. Infrared spectra were ob-
tained using NaCl plates on a Bruker Equinox 55/S FT-IR
Spectrophotometer, and relative intensities are expressed
qualitatively as s (strong), m (medium), and w (weak). TLC
analysis was performed using Aldrich 254 nm polyester-
backed plates (60 Å, 250 m) and visualized using UV and
a suitable chemical stain. Low-resolution mass spectra were
obtained using an Agilent-1100-HPLC/MSD and can be
either APCI or ESI, or an IonSpec HiRes-MALDI FT-Mass
Spectrometer. High-resolution mass spectral analyses were
performed at University of Wisconsin Mass Spectrometry
Laboratories. All spectral data obtained for new compounds
are reported. X-ray analyses were performed at the X-ray
facility at University of Minnesota.
2.2 Experimental procedures and characterizations
Scheme 1 A cyclic acetal tethered RCM.
Synthesis of -lactone 9
-Lactone 9 was prepared from (S)-glycidol in 6 steps with
an overall yield of 42% yield (1.28 g) [23].
Rf =0.11 (10% EtOAc in hexanes); On occasions, -lactone
9 solidifies into a white solid upon standing: mp 71–72 C;
1
[]D20 = 27.6 (c 0.71, CHCl3); H NMR (400 MHz, CDCl3)
 1.06 (s, 9H), 1.38 (s, 3H), 1.49 (s, 3H), 2.03 (ddd, J=15.0,
10.5, 3.5 Hz, 1H), 2.11 (dt, J = 15.0, 2.5 Hz, 1H), 3.77 (dd,
J = 11.1, 3.9 Hz, 1H), 3.86 (dd, J = 11.3, 4.5 Hz, 1H), 4.59
(d, J = 6.8 Hz, 1H), 4.63–4.71 (m, 2H), 7.37–7.46 (m, 6H),
7.64–7.68 (m, 4H); 13C NMR (125 MHz, CDCl3)  19.5,
24.3, 26.3, 27.0, 30.6, 65.2, 71.9, 73.1, 75.4, 110.9, 128.0,
130.1, 132.9, 133.2, 135.7, 135.8, 168.3; IR (neat) cm1
2932w, 2857w, 1749m, 1472w, 1427m, 1376m, 1260m,
1209m, 1139m, 1110m, 1068m, 1041m, 1005m, 945m,
871w, 822m, 802m, 737m, 701s; HRMS (MALDI) calcd
Scheme 2 Biosynthetic relation between (+)-aigialospirol and (+)-
hypothemycin.

Figueroa R, et al. Sci China Chem January (2011) Vol.54 No.1
33
for C25H32O5NaSi 463.1911, found 463.1906.
(1.01 g) as a colorless oil.
Rf = 0.23 (5% EtOAc in hexanes); []D20 = +3.85 (c 1.82,
Synthesis of lactol 10
1
CH2Cl2); H NMR (500 MHz, C6D6)  1.08 (d, J = 6.3 Hz,
To a solution of -lactone 9 (534.0 mg, 1.21 mmol) in Et2O
(9 mL) was added vinyl magnesium bromide (1.0 M in THF:
1.30 mL, 1.30 mmol, 1.07 equiv) carefully dropwise at 78 C.
The pale yellow solution was stirred at 78 °C for 1 h be-
fore it was quenched with sat aq NH4Cl at that temperature.
H2O was then added until all precipitate was dissolved. The
resulting mixture was allowed to reach rt and was diluted
with Et2O. The aqueous phase was separated and extracted
with two portions of Et2O. The combined organic extracts
were dried (MgSO4) and concentrated in vacuo to give a
cloudy crude oil that was purified via silica gel flash column
chromatography (eluted with 10% EtOAc/90% hexane) to
afford a mixture of lactol 10 and vinyl ketone 11 in 88%
combined yield (500.0 mg) as a colorless oil.
3H), 1.16 (s, 3H), 1.18 (s, 9H), 1.43 (s, 3H), 1.87 (dt, J =
12.4, 1.5 Hz, 1H), 2.13–2.24 (m, 3H), 3.74 (dd, J=10.5, 5.6
Hz, 1H), 3.94 (dd, J = 10.4, 5.6 Hz, 1H), 4.06–4.12 (m, 1H),
4.24 (d, J = 7.6 Hz, 1H), 4.49 (dd, J = 7.4, 2.6 Hz, 1H), 4.56
(dtd, J = 10.7, 5.6, 2.0 Hz, 1H), 4.91–4.99 (m, 2H), 5.28 (dd,
J = 8.8, 3.9 Hz, 1H), 5.68 (ddt, J = 17.2, 10.1, 7.1 Hz, 1H),
5.77–5.86 (m, 2H), 7.22–7.24 (m, 6H), 7.80–7.85 (m, 4H);
13C NMR (125 MHz)  In C6D6: 19.5, 21.5, 24.0, 26.6, 27.1,
31.0, 43.0, 66.9, 68.1, 69.6, 70.9, 76.6, 98.5, 108.3, 117.1,
118.1, 128.1, 130.0, 133.9, 134.1, 135.3, 136.07, 136.15,
138.3 (missing two aryl carbons due to overlap); in CDCl3:
19.6, 21.5, 24.2, 26.5, 27.1, 30.8, 42.7, 66.9, 67.7, 69.3,
70.8, 76.3, 98.0, 108.5, 117.3, 118.5, 127.9, 128.0, 129.9,
133.8, 133.9, 135.0, 135.9, 136.0, 137.7 (missing one aryl
carbon due to overlapping para-carbons of the TBDPS
group at 129.9 ppm); IR (neat) cm1 3073w, 2930m, 2859w,
1375w, 1263w, 1212m, 1107s, 1053s, 998s, 824m, 703s;
HRMS (MALDI) calcd for C32H44O5NaSi 559.2850, found
559.2821.
Cyclic acetal 13: Rf = 0.52 (20% EtOAc in hexanes);
[]D20 = +10.52 (c 1.1, CHCl3); 1H NMR (400 MHz, C6D6) 
7.78–7.85 (m, 4H), 7.20–7.25 (m, 6H), 5.84 (dd, 1H, J =
17.2, 10.4 Hz), 5.75 (dd, 1H, J = 7.2, 5.2 Hz), 5.70 (dd, 1H,
J=7.2, 5.2 Hz), 5.29 (dd, 1H, J=10.4, 2.4 Hz), 5.03 (dt, 1H,
J = 2.0, 1.2 Hz), 4.95 (ddt, 1H, J=9.2, 1.2, 0.8 Hz), 4.49 (m,
2H), 4.23 (d, 1H, J = 7.6 Hz), 3.78 (m, 2H), 3.62 (dt, 1H, J=
9.6, 6.8 Hz), 3.46 (dt, 1H, J = 9.6, 6.8 Hz), 2.23 (ddd, 1H,
J=14.0, 12.4, 3.6 Hz), 2.17 (q, 2H, J=6.8 Hz), 1.84 (dt, 1H,
J = 14.0, 2.0 Hz), 1.44 (s, 3H), 1.17 (s, 9H), 1.15 (s, 3H);
13C NMR (100 MHz, C6D6)  137.8, 136.2, 136.2, 136.1,
134.2, 134.0, 130.0, 130.0, 118.3, 116.3, 108.5, 98.6, 76.8,
71.0, 69.5, 68.1, 60.6, 34.9, 30.3, 27.2, 26.5, 23.9, 19.6; IR
(neat) cm1 3072w, 2931m, 2858m, 2360w, 1782m, 1685w,
1428m, 1262m, 1210m, 1111s, 1061s, 997s; HRMS (ESI)
calcd for C31H42O5Si 545.2699, found 545.2711.
Rf = 0.20 [10% EtOAc in hexanes]; []D20 =9.82 (c 2.50,
1
CHCl3); H NMR (400 MHz, C6D6)  1.07 (s, 3H), 1.14 (s,
9H), 1.16 (s, 12H), 1.169 (s, 9H), 1.171 (s, 3H), 1.41 (s, 3H),
1.45 (s, 3H), 1.49 (s, 3H), 1.65–1.78 (m, 3H), 1.87 (dt, J =
14.2, 3.9 Hz, 1H), 2.07 (dt, J = 14.6, 2.7, Hz, 1H), 2.54 (dt,
J = 14.5, 11.9, 3.0 Hz, 1H), 2.77–2.78 (m, 1H), 2.91–2.92
(m, 1H), 3.28–3.29 (m, 1H), 3.41 (dd, J = 10.6, 0.9 Hz, 1H),
3.59–3.78 (m, 5H), 3.93–4.00 (m, 1H), 4.11 (d, J = 8.0 Hz,
2H), 4.26–4.35 (m, 4H), 4.39 (d, J = 7.6 Hz, 1H), 4.42 (dt,
J = 7.8, 2.7 Hz, 1H), 5.09 (dd, J = 10.5, 1.7 Hz, 1H), 5.17
(dd, J = 10.7, 1.8 Hz, 1H), 5.19 (dd, J = 10.3, 1.9 Hz, 1H),
5.64 (dd, J = 17.3, 1.7 Hz, 1H), 5.66 (dd, J = 17.2, 1.8 Hz,
1H), 6.00 (dd, J = 17.2, 10.5 Hz, 1H), 6.24 (dd, J=17.3, 1.9
Hz, 1H), 6.40 (dd, J = 17.3, 10.6 Hz, 1H), 6.73 (dd, J=17.3,
10.5 Hz, 1H), 7.27–7.18 (m, 18H), 7.74–7.83 (m, 12H); IR
(neat) cm1 3441w, 2931w, 2857w, 1753w, 1697w, 1472w,
1427w, 1373w, 1262m, 1210m, 1110s, 1065m, 1041m,
701s; HRMS (MALDI) calcd for C27H36O5NaSi 491.2224,
found 491.2229.
Preparations of cyclic acetals
Note: Two batches of (each batch 573.6 mg of the mixture
10 and 11) the same amount was done and combined during
the purification process. A round-bottomed flask was charged
with the mixture of 10 and 11 (573.6 mg, 1.22 mmol),
CH2Cl2 (13 mL), activated pulverized 4 Å molecular sieves
(1.10 g) and chiral homo-allylic alcohol 12 [24, 25] (1.05 g,
12.24 mmol). The mixture was cooled to 78 C and a solu-
tion of Tf2NH (344.1 mg, 1.22 mmol) in CH2Cl2 (13 mL)
was added via syringe (Note: This solution was freshly
made prior to be used). The reaction mixture was stirred for
1 h before it was quenched with Et3N (10 mL). The cooling
bath was removed and the mixture was allowed to reach rt.
After filtration through a pad of CeliteTM, the filtrate was
concentrated in vacuo to give a crude yellow oil. At this
point, the two identical batches were combined and purified
via silica gel flash column chromatography (eluted with 5%
EtOAc in hexanes) to afford cyclic acetal 14 in 76% yield
Cyclic acetal tethered RCM
To a purple solution of the Grubbs’ first generation catalyst
(129.0 mg, 0.157 mmol) in toluene (270 mL) was added a
solution of cyclic acetal 14 (676.0 mg, 1.26 mmol) in tolu-
ene (60 mL). The mixture was stirred at rt for 3 h. Subse-
quently, it was concentrated in vacuo to give a crude brown
oil that was purified via silica gel flash column chromatog-
raphy (eluted with 5% EtOAc in hexanes) to yield the pure
spiroketal 16 in 86% yield (554.0 mg) as a colorless oil.
Rf = 0.20 (5% EtOAc in hexanes); []D20 = +20.9 (c 0.77,
1
CHCl3); H NMR (500 MHz, C6D6)  1.03 (d, J = 6.3 Hz,
3H), 1.17 (s, 9H), 1.23 (s, 3H), 1.52 (s, 3H), 1.58–1.68 (m,
3H), 2.06 (ddd, J = 14.1, 3.4, 2.2 Hz, 1H), 2.32 (ddd, J =
14.4, 11.7, 3.2 Hz, 1H), 3.64–3.70 (m, 1H), 3.83 (dd, J =
10.1, 6.2 Hz, 1H), 3.92 (dd, J = 10.0, 4.9 Hz, 1H), 4.42 (dt,

34
Figueroa R, et al. Sci China Chem January (2011) Vol.54 No.1
J = 7.6, 2.7 Hz, 1H), 4.48–4.53 (m, 1H), 5.68 (ddd, J=10.3,
4.6, 3.7 Hz, 1H), 6.18 (dt, J = 10.3, 2.0 Hz, 1H), 7.15–7.21
(m, 6H), 7.79–7.83 (m, 4H);13C NMR (125 MHz)  In C6D6:
19.6, 21.9, 24.3, 26.6, 27.2, 28.8, 31.1, 68.0, 68.6, 69.5,
71.1, 75.3, 95.3, 108.9, 126.3, 129.2, 129.82, 129.84, 134.3,
134.4, 136.1, 136.2 (missing two aryl carbons due to over-
lap); in CDCl3: 19.6, 22.0, 24.5, 26.5, 27.2, 28.6, 31.2, 68.0,
68.2, 69.2, 70.9, 75.1, 95.1, 109.1, 127.7, 127.8, 127.9,
128.0, 129.7, 129.8, 134.1, 135.9, 136.0 (missing one aryl
carbon due to overlapping Si-substituted ipso-carbons of the
TBDPS group at 129.9 ppm); IR (neat) cm1 3071w, 3050w,
2931m, 1428w, 1371m, 1262w, 1210m, 1166m, 1139m,
1111s, 1032s, 823m, 700s; HRMS (MALDI) calcd for
C30H40O5NaSi 531.2537, found 531.2550.
Hz, 1H), 7.22–7.24 (m, 6H), 7.80–7.84 (m, 4H); 13C NMR
(125 MHz)  In C6D6: 19.6, 20.8, 27.1, 31.7, 35.2, 64.5,
65.5, 67.2, 69.3, 71.4, 99.3, 128.4, 129.1, 129.99, 130.03,
134.10, 134.14, 136.16, 136.18; in CDCl3: 19.6, 21.4, 27.1,
31.9, 35.3, 64.8, 65.3, 66.9, 71.3, 99.1, 127.5, 127.8, 127.9,
129.9, 130.0, 130.3, 133.8, 133.9, 135.9, 136.0; IR (CH2Cl2
film) cm1 3497br, 3047w, 2930m, 2857m, 1589w, 1427m,
1111s, 1066s, 1012s, 822m, 740m, 702s; HRMS (MALDI)
calcd for C27H36O5NaSi 491.2224, found 491.2269.
Synthesis of aldehyde 19
To a solution of spiroketal 16 (138.5 mg, 0.27 mmol) in
THF (10 mL) was added tetra-n-butylammonium fluoride
(1.0 M in THF: 0.41 mL, 0.41 mmol). The resulting yellow
solution was stirred at rt for 22 h. The mixture was concen-
trated in vacuo to give a crude residue that was purified via
silica gel flash column chromatography (eluted with 35%
EtOAc in hexanes) to afford the desired alcohol in 98%
yield (72.3 mg) as a white solid.
Spiroketal 15: Rf = 0.53 (20% EtOAc in hexanes); []D20 =
1
17.69 (c 0.9, CHCl3); H NMR (400 MHz, C6D6)  7.83
(m, 4H), 7.22 (m, 6H), 6.28 (ddd, 1H, J = 10.4, 2.8, 1.6 Hz),
5.71 (ddd, 1H, J = 10.4, 5.6, 1.6 Hz), 4.50 (ddd, 1H, J=7.6,
2.8, 2.0 Hz), 4.45 (ddd, 1H, J = 10.8, 3.2, 1.6 Hz), 4.13 (d,
1H, J = 7.6 Hz), 3.95 (ddd, 1H, J = 12.0, 11.6, 3.6 Hz), 3.79
(ddd, 2H, J = 19.2, 10.4, 5.2 Hz), 3.44 (dd, 1H, J=10.8, 6.4
Hz), 2.44 (ddd, 1H, J = 14.4, 8.8, 2.8 Hz), 1.98 (m, 1H),
1.91 (ddd, 1H, J = 14.0, 2.8, 2.0 Hz), 1.51 (s, 3H), 1.18 (s,
12H); 13C NMR (100 MHz, C6D6)  136.2, 136.1, 134.3,
134.1, 130.0, 129.9, 128.9, 128.3, 128.1, 128.0, 127.8,
127.5, 108.8, 93.9, 76.7, 71.1, 69.4, 68.5, 57.7, 29.1, 27.2,
26.5, 24.9, 24.1, 19.6; IR (neat) cm1 3071w, 2932m, 2859w,
1473m, 1381m, 1371m, 1266m, 1138s, 1028s; HRMS (ESI)
calcd for C29H38O5Si 494.6950, found 494.6950.
Rf = 0.18 (35% EtOAc in hexanes); mp 131–134 C;
1
[]D20 = +108.9 (c 0.52, CHCl3); H NMR (400 MHz, C6D6)
 0.97 (d, J = 6.1 Hz, 3H), 1.23 (d, J = 0.4 Hz, 3H), 1.44
(dddd, J = 17.1, 5.5, 3.8, 1.4 Hz, 1H), 1.52 (d, J = 0.4 Hz,
3H), 1.60 (ddt, J=17.4, 9.1, 2.6 Hz, 1H), 1.64 (ddd, J=14.4,
4.2, 2.6 Hz, 1H), 2.47 (ddd, J = 14.6, 12.0, 2.8 Hz, 1H), 2.83
(dd, J = 8.9, 2.6 Hz, 1H), 3.37 (ddd, J = 11.5, 8.8, 2.9 Hz,
1H), 3.55 (dqd, J = 9.3, 6.2, 3.6 Hz, 1H), 3.81 (dt, J = 11.4,
2.6 Hz, 1H), 3.97 (d, J = 7.7 Hz, 1H), 4.27–4.33 (m, 2H),
5.65 (ddd, J=10.3, 5.7, 2.6 Hz, 1H), 6.13 (ddd, J=10.3, 2.6,
1.4 Hz, 1H); 13C NMR (125 MHz, C6D6)  21.5, 24.4, 25.4,
26.5, 31.1, 65.5, 68.4, 70.2, 70.8, 74.3, 95.5, 108.9, 126.8,
128.5; IR (neat) cm1 3447brw, 3050w, 2979w, 2921w,
1451w, 1431w, 1382m, 1370m, 1209m, 1168m, 1035s, 1016s;
mass spectrum (APCI): m/e (% relative intensity) 271 (10)
(M + H)+, 253 (7) ([M+H]H2O)+, 227 (5), 213 (100), 195
(37), 167 (80); HRMS (ESI) calcd for C14H22O5NaSi
293.1365, found 293.1353.
Hydrolytic epimerization of spiroketal 16
To a solution of spiroketal 16 (9.00 mg, 0.018 mmol) in
MeOH (1 mL) was added p-TsOH (0.50 mg, 0.0027 mmol,
practical grade). The mixture was stirred at rt for 1 h until
TLC showed completion consumption of 16. (Note: The
reaction time varied, and thus, it should be carefully moni-
tored via TLC analysis.) The reaction mixture was concen-
trated in vacuo and diluted with CH2Cl2 (10 mL). The re-
sulting solution was washed with sat aq NaHCO3 (1 mL).
The aqueous layer was extracted with CH2Cl2 (3 × 5 mL).
The combined organic layers were dried (MgSO4) and con-
centrated in vacuo to afford a crude residue that was purified
via silica gel flash column chromatography (eluted with 2%
MeOH in CH2Cl2) to afford diol 17 in 81% yield (6.70 mg)
as a white solid.
To a solution of the above alcohol (117.1 mg, 0.43 mmol)
in CH2Cl2 (4.6 mL) were added iodobenzene diacetate
(153.5 mg, 0.48 mmol) and then 2,2,6,6-tetramethylpiperidine-
1-oxyl (TEMPO) (8.50 mg, 0.054 mmol). The resulting
orange solution was stirred at rt for 21 h before it was
quenched with sat aq Na2S2O3 (5.0 mL). The mixture was
extracted with CH2Cl2 (3 × 10 mL), and the combined or-
ganic extracts were dried (Na2SO4) and concentrated in
vacuo to give a yellow/orange oil that was purified via silica
gel flash column chromatography (eluted with 15% EtOAc
in hexanes to remove impurities and 30% EtOAc in hexanes)
to afford aldehyde 19 in 78% yield (90.6 mg) as a white
solid.
Rf = 0.16 (2% MeOH in CH2Cl2); mp 43–45 C; []D20 =
+51.8 (c 0.62, CH2Cl2); 1H NMR (500 MHz, C6D6)  0.90 (d,
J = 6.3 Hz, 3H), 1.17 (s, 9H), 1.39–1.48 (m, 2H), 1.62 (ddt,
J = 17.6, 11.0, 2.4 Hz, 1H), 1.91 (ddd, J = 13.9, 3.4, 2.4 Hz,
1H), 2.61 (d, J = 10.9 Hz, 1H), 3.30 (dd, J = 10.4, 3.4 Hz,
1H), 3.49 (d, J = 10.6 Hz, 1H), 3.65 (dd, J = 10.6, 4.3 Hz,
1H), 3.74 (dd, J = 10.6, 5.7 Hz, 1H), 3.98 (dqd, J=10.9, 6.3,
3.3 Hz, 1H), 4.05–4.09 (m, 1H), 4.14–4.19 (m, 1H), 5.56
(ddd, J = 10.0, 2.7, 1.3 Hz, 1H), 5.72 (ddd, J=10.0, 5.9, 1.9
Rf = 0.66 (20% EtOAc in hexanes); mp 62–63 C; []D20 =
1
+136.7 (c 0.73, CHCl3); H NMR (500 MHz, C6D6)  0.96
(d, J = 6.4 Hz, 3H), 1.18 (s, 3H), 1.46 (s, 3H), 1.56–1.59 (m,
2H), 2.04 (ddd, J = 14.4, 5.7, 2.8 Hz, 1H), 2.21 (ddd, J =

Figueroa R, et al. Sci China Chem January (2011) Vol.54 No.1
35
14.4, 11.1, 3.1 Hz, 1H), 3.52–3.58 (m, 1H), 3.84 (d, J=7.4
Hz, 1H), 4.17 (dt, J = 7.6, 2.9 Hz, 1H), 4.34 (dd, J = 11.1,
5.7 Hz, 1H), 5.67 (ddd, J = 10.3, 4.7, 3.6 Hz, 1H), 6.14 (dt,
J = 10.4, 2.0 Hz, 1H), 9.81 (s, 1H); 13C NMR (125 MHz,
C6D6)  21.5, 24.7, 25.3, 26.7, 30.9, 68.2, 70.3, 73.8, 75.1,
95.6, 109.3, 127.2, 127.8, 203.2; IR (neat) cm1 2984w,
2936w, 1735s, 1381s; mass spectrum (APCI): m/e (% rela-
tive intensity) 269 (15) (M + H)+.
(100 MHz, CDCl3)  13.5, 14.4, 18.3, 25.8, 39.5, 43.1, 55.5,
106.0, 106.5, 123.0, 128.8, 152.5, 160.9, 169.3 [missing two
Me carbon resonances from the TBS group due to rotameric
line broadening]; IR (neat) cm1 2933w, 2898w, 2860w,
1632brm, 1608brm, 1463brm, 1428br, 1288m, 1255m,
1166m, 839s, 783s; mass spectrum (APCI): m/e (% relative
intensity) 338 (100) (M + H)+; HRMS (MALDI) calcd for
C18H32O3NSi 338.2146, found 338.2131.
Preparation of amide 20
Addition to aldehyde 19
To a solution of 4-methoxysalicylic acid (2.52 g, 15.0 mmol)
in DMF (12 mL) was added diisopropylethylamine (9.40
mL, 54.0 mmol) and tert-butyldimethylsilyl chloride (5.65 g,
37.5 mL). The mixture was vigorously stirred at rt for 3 h.
The reaction mixture was poured over H2O (30 mL) and
extracted with Et2O (4 × 50 mL). The combined organic
extracts were washed with sat aq NaCl (2 × 25 mL), dried
over MgSO4, and concentrated in vacuo to yield a yellow
oil that was purified via silica gel flash column chromatog-
raphy (eluted with 2% EtOAc in hexanes) to afford the de-
sired silyl ester in 74% yield (4.40 g) as a colorless oil.26
To a round-bottomed flask charged with anhyd THF (1.2 mL)
were added tetramethylethylenediamine (19.8 mg, 0.030
mL, 0.17 mmol) and s-BuLi (1.4 M in cyclohexane: 0.12 mL,
0.17 mmol) at 78 C via syringe. The deep yellow solution
was stirred for 2 min before a solution of amide 20 (57.6 mg,
0.17 mmol) in THF (0.3 mL) was added dropwise also via
syringe. The pale yellow solution was stirred for 45 min at
78 C before a solution of aldehyde 19 (41.6 mg, 0.155
mmol) in THF (0.3 mL) was added. Additional THF (0.2
mL) was used to ensure complete transferred of the alde-
hyde. The mixture was stirred for 1 h at 78 C and 2.5 h at
rt, and it was quenched with H2O (3 mL) and extracted with
Et2O (3 × 5 mL). The combined organic layers were dried
(Na2SO4) and concentrated in vacuo to give a crude mixture
that was purified via silica gel flash column chromatogra-
phy (eluted with 10%–60% EtOAc in hexanes) to afford
secondary alcohols 21-syn (18.5 mg) and 21-anti (28.8 mg)
in a combined 50% yield.
21-syn: colorless thick oil; Rf = 0.34 (35% EtOAc in
hexanes); 1H NMR (400 MHz, C6D6)  0.15 (s, 3H), 0.21 (s,
3H), 0.79 (t, J = 7.1 Hz, 3H), 0.98 (s, 9H), 0.99 (d, J = 6.2
Hz, 3H), 1.09 (t, J=7.1 Hz, 3H), 1.19 (s, 3H), 1.34–1.40 (m,
1H), 1.55 (s, 3H), 1.62 (ddt, J = 13.6, 9.8, 2.3 Hz, 1H),
1.86–1.90 (m, 1H), 2.52–2.58 (m, 1H), 2.94 (q, J = 7.1 Hz,
2H), 3.10 (sextet, J = 7.2 Hz, 1H), 3.39 (s, 3H), 3.49–3.60
(m, 2H), 3.97 (d, J = 7.6 Hz, 1H), 4.32 (dt, J = 7.8, 2.7 Hz,
1H), 4.58 (s, 1H), 5.24–5.27 (m, 2H), 5.63 (ddd, J = 10.3,
5.8, 2.0 Hz, 1H), 6.16 (dd, J = 10.3, 1.8 Hz, 1H), 6.58 (d, J=
2.5 Hz, 1H), 7.39 (d, J = 2.5 Hz, 1H); mass spectrum
(APCI): m/e (% relative intensity) 606 (100) (M + H)+, 588
(33) ([M + H]+-H2O), 570 (20), 548 (8) ([M+H]+tBu), 530
(15), 512 (10), 475 (15).
1
Rf = 0.21 (2% EtOAc in hexanes); H NMR (400 MHz,
CDCl3)  0.19 (s, 6H), 0.32 (s, 6H), 0.97 (s, 9H), 0.99 (s,
9H), 3.78 (s, 3H), 6.37 (d, J = 2.5 Hz, 1H), 6.49 (dd, J=8.8,
2.5 Hz, 1H), 7.75 (d, J = 8.8 Hz, 1H); 13C NMR (100 MHz,
CDCl3)  4.4, 4.2, 18.0, 18.7, 26.0, 26.1, 55.5, 106.9,
107.5, 116.6, 133.7, 158.4, 163.7, 164.8; IR (neat) cm1
2954w, 2930w, 2858w, 1704m, 1606m, 1493w, 1463m,
1443m, 1337w, 1251s, 1169m, 1149m, 1134m, 1082s, 834s,
782s; HRMS (ESI) calcd for C20H37O4Si2 397.2230, found
397.2221.
To a solution of the above silyl ester (250.0 mg, 0.630
mmol) in CH2Cl2 (0.6 mL) was added a drop of DMF. After
cooling to 0 C in an ice-water bath, oxalyl chloride (88.0
mg, 0.060 mL, 0.693 mmol) was added dropwise at a rate to
maintain the control of the reaction. The resulting yellow
solution was stirred 0 °C for 1.5 h before the ice bath was
removed and the mixture was further stirred overnight. Af-
ter which, the reaction mixture was concentrated in vacuo,
re-diluted with fresh CH2Cl2, and re-concentrated in vacuo
to afford the crude acyl chloride.
To a solution of the crude acyl chloride prepared above
in CH2Cl2 (0.6 mL) was added diethylamine (93.0 mg, 0.13
mL, 1.26 mmol) at rt dropwise at a rate to maintain the con-
trol of the reaction. The resulting mixture was stirred for 1 h
before being concentrated in vacuo to afford a crude residue
that was purified via silica gel flash column chromatogra-
phy (eluted with 20% EtOAc in hexanes) to yield the de-
sired amide 20 in 56% yield (119.0 mg) as a colorless oil.
21-anti: colorless thick oil; Rf = 0.20 (35% EtOAc in
hexane); 1H NMR (400 MHz, C6D6)  0.15 (s, 3H), 0.22 (s,
3H), 0.99 (d, J = 6.2 Hz, 3H), 1.01 (s, 9H), 1.07 (t, J = 7.1
Hz, 3H), 1.21 (s, 3H), 1.28 (t, J=7.0 Hz, 3H), 1.30–1.40 (m,
1H), 1.52 (s, 3H), 1.52-1.62 (m, 1H), 1.93 (ddd, J = 15.0,
4.3, 3.1 Hz, 1H), 2.62 (ddd, J = 15.0, 12.0, 2.7 Hz, 1H),
2.96–3.05 (m, 1H), 3.26 (qd, J = 7.7, 1.1 Hz, 2H), 3.32 (s,
3H), 3.48–3.53 (m, 1H), 3.94 (d, J = 7.6 Hz, 1H), 4.00–4.09
(m, 1H), 4.28 (dt, J = 7.8, 2.5 Hz, 1H), 4.40 (s, 1H), 4.69 (dt,
J = 12.1, 4.1 Hz, 1H), 5.60–5.64 (m, 2H), 6.06 (dd, J=10.4,
1.9 Hz, 1H), 6.55 (d, J = 2.3 Hz, 1H), 7.20 (d, J = 2.3 Hz,
1H); mass spectrum (APCI): m/e (% relative intensity) 606
1
Rf = 0.41 (20% EtOAc in hexanes); H NMR (400 MHz,
CDCl3)  0.22 (s, 6H), 0.96 (s, 9H), 1.01 (t, J=7.1 Hz, 3H),
1.23 (t, J = 7.2 Hz, 3H), 3.10–3.34 (broad m, 2H), 3.34–3.70
(broad m, 2H), 3.78 (s, 3H), 6.36 (d, J = 2.3 Hz, 1H), 6.53
(dd, J = 8.5, 2.3 Hz, 1H), 7.12 (d, J=8.4 Hz, 1H); 13C NMR

36
Figueroa R, et al. Sci China Chem January (2011) Vol.54 No.1
(100) (M + H)+, 588 (25) ([M + H]+H2O), 570 (15), 548 (7)
([M+H]+-tBu), 530 (10), 512 (8), 475 (15).
3.52 (dd, J = 9.9, 3.3 Hz, 1H), 3.86 (s, 3H), 4.01 (m, 1H),
4.12–4.18 (m, 2H), 5.36 (d, J = 5.9 Hz, 1H), 5.67 (dt, J =
10.0, 2.0 Hz, 1H), 6.17 (dt, J = 10.0, 4.0 Hz, 1H), 6.48 (d,
J = 1.7 Hz, 1H), 6.60 (s, 1H), 7.68 (s, 1H); 13C NMR (100
MHz, CDCl3)  21.1, 31.4, 33.6, 56.0, 65.0, 65.8, 68.3, 70.7,
82.3, 99.1, 101.2, 101.6, 104.3, 126.5, 130.5, 149.1, 157.7,
167.3, 171.3; IR (CHCl3 film) cm-1 3452brm, 3013w,
2970w, 2930w, 1736s, 1612s, 1444m, 1381m, 1316m,
1216m, 1154s, 1063s, 997s, 843w; mass spectrum (APCI):
m/e (% relative intensity) 379 (15) (M+H)+, 361 (100) ([M+
H]H2O)+, 343 (25), 317 (12), 307 (10), 253 (37), 235 (16),
219 (7), 146 (98), 127 (10), 116 (7), 101 (30); HRMS (ESI)
calcd for C19H22O8Na 401.1212, found 401.1208.
Data as mixture 21-syn + 21-anti: []D20 = +24.3 (c 0.070,
C6H6); IR (neat) cm1 3430brm, 3242brm, 2936brm, 1767m,
1675m, 1606s, 1464s, 1257m, 1160s, 1023s, 840s; HRMS
(ESI) calcd for C32H51NO8NaSi 628.3282, found 628.3287.
Synthesis of lactone 22-syn
To a solution of alcohol 21-syn (7.20 mg, 0.0119 mmol) in a
ternary solvent mixture consisting of MeOH (0.6 mL), THF
(0.2 mL), and H2O (0.06 mL) was added KOH (6.70 mg,
0.119 mmol). The resulting mixture was stirred at rt over-
night before being partitioned between H2O (5 mL) and
CH2Cl2 (5 mL). The mixture was extracted with CH2Cl2 (3 ×
5 mL). The combined organic extracts were dried (Na2SO4)
and concentrated in vacuo to give a crude oil that was puri-
fied via preparative TLC (eluted with 5% MeOH in CH2Cl2)
to afford lactone 22-syn in 71% yield (2.60 mg) as a thick
oil along with the recovered but desilylated starting material
(1.70 mg, 29%).
3 Results and discussion
3.1 Epimerization of the C6′ spiroketal carbon
Synthesis of key cyclic acetal RCM precursor commenced
with -lactone 9, which was prepared in grams quantity
from (S)-glycidol in 6 steps [23, 27] (Scheme 3). A high
yielding vinyl magnesium Grignard addition followed by
acetalization using 1.0 equiv of Tf2NH at 78 °C for 1 h [7,
9, 28] led to cyclic acetals 13 and 14 in 76% and 46% yield
as a single diastereomer, respectively, from 3-butene-1-ol (as
a model system), and chiral homo-allylic alcohol 12 [24, 25].
One distinct advantage of cyclic acetal tethered strategies
for spiroketal synthesis is the high level of stereochemical
control at the acetal carbon during the cyclic acetal con-
struction, taking advantage of well-known anomerically
favored axial addition in glycosylation chemistry [8, 29, 30].
However, nOe experiments on both 13 and 14 quickly con-
firm that while the stereoselectivity was very high in these
cyclic acetal formations, the stereochemistry of the acetal
carbon at C6′ is wrong. As shown in Figure 2, the homo-
allyloxy group at C6′ is actually alpha in 13 and 14, thereby
suggesting that the glycosylation process took place in an
anti-anomeric manner via oxocarbenium ion A. The ano-
merically favored addition of ROH to A would have given
the desired C6′ stereochemistry for (+)-aigialospirol (4).
Rf = 0.75 (5% MeOH in CH2Cl2); []D20 = 37.9 (c 0.68,
1
CHCl3); H NMR (400 MHz, CDCl3)  1.30 (d, J = 6.2 Hz,
3H), 1.33 (s, 3H), 1.44 (s, 3H), 1.99 (ddt, J = 17.3, 7.6, 3.0
Hz, 1H), 2.09–2.18 (m, 2H), 2.40 (ddd, J = 14.5, 11.3, 3.1
Hz, 1H), 3.83 (s, 3H), 3.99–4.06 (m, 2H), 4.14 (d, J = 7.6
Hz, 1H), 4.60 (dt, J = 7.6, 2.7 Hz, 1H), 5.34 (d, J = 8.0 Hz,
1H), 5.88 (dt, J = 10.3, 2.0 Hz, 1H), 6.07 (ddd, J=10.2, 4.8,
3.3 Hz, 1H), 6.43 (d, J = 1.7 Hz, 1H), 6.69 (d, J = 0.9 Hz,
1H), 7.65 (br s, 1H); 13C NMR (100 MHz, CDCl3)  21.9,
24.5, 26.5, 27.5, 31.1, 56.1, 68.3, 70.2, 70.4, 74.9, 84.2,
95.5, 101.2, 103.0, 104.5, 109.4, 127.4, 127.8, 150.5, 157.7,
167.2, 171.9; IR (neat) cm1 2928w, 1733s, 1612s, 1480m,
1372m, 1320m, 1253m, 1213m, 1153s, 1032s, 849m; mass
spectrum (APCI): m/e (% relative intensity) 419 (3) (M +
H)+, 401 (7), 361 (100), 343 (25), 317 (20), 167 (5); HRMS
(ESI) calcd for C22H26O8Na 441.1525, found 441.1529.
Synthesis of (+)-aigialospirol 4
To a solution of lactone 22-syn (28.0 mg, 0.067 mmol) in
MeOH (4 mL) was added p-TsOH (2.70 mg, 0.014 mmol,
practical grade). The mixture was stirred at rt for 4 h before
it was concentrated in vacuo and diluted with CH2Cl2 (10
mL). The solution was washed with sat aq NaHCO3 (10
mL). The aqueous layer was extracted with CH2Cl2 (3 × 10
mL), and the combined organic layers were dried (Na2SO4)
and concentrated in vacuo. The resulting crude residue was
purified via preparative TLC (eluted with 5% MeOH in
CH2Cl2) to afford (+)-aigialospirol 4 in 53% yield (13.4 mg)
as a white solid.
Rf = 0.49 (5% MeOH in CH2Cl2); mp 90–94 C; lit.ref
85–89 C; []D20 = +79.2 (c 0.67, CHCl3); lit.ref []D25 =+47 (c
0.50, CHCl3); 1H NMR (500 MHz, CDCl3)  1.27 (d, J=6.1
Hz, 3H), 1.89 (ddd, J = 14.2, 11.9, 2.6 Hz, 1H), 2.03–2.08
(m, 3H), 2.54 (d, J =10.5 Hz, 1H), 3.44 (d, J=10.5 Hz, 1H),
Scheme 3 Cyclic acetal RCM precursors.

Figueroa R, et al. Sci China Chem January (2011) Vol.54 No.1
37
Scheme 4 The key cyclic acetal tethered RCM.
Figure 2 Stereochemistry of the cyclic acetal carbon C6′.
This unexpected outcome is actually quite reasonable
based on conformational considerations of oxocarbenium
ion A, which would fit perfectly Woerpel’s stereoelectronic
model [29]: (a) The C2′ substituent could dominate by
holding an equatorial position based on sterics; (b) the CO
bond at C4′ would be stereoelectronically favored to assume
the equatorial position to minimize unfavorable *C-O over-
lapping with the electron-deficient C6′ oxocarbenium car-
bon; and (c) the CO bond at C5′ is also stereoelectronically
favored to occupy the axial position. With these analyses,
the anomerically favored axial entrance is completely
blocked by the acetonide methyl group. It is also notewor-
thy that cyclic acetals 13 and 14 prefer two different con-
formations in the pyranyl ring, chair in the former and boat
in the latter. This is likely due to the fact that the C10′ Me
group in 14 would experience a greater 1,3-diaxial interac-
tion (than that in 13) with the C2′ methylene group when in
a chair conformation.
Scheme 5 Acid induced C6’ epimerization.
Ring-closing metathesis of cyclic acetal 13 and 14 pro-
ceeded smoothly to give spiroketal 15 and 16 in 86% yield,
respectively, employing 12.5 mol% Grubb’s Gen-I catalyst
(Scheme 4). Once again, nOe experiments on both 15 and
16 further confirmed that the stereochemistry at the C6′
spiroketal center is wrong in the context of (+)-aigialospirol
4 synthesis. In addition, while 15 tended to decompose un-
der acidic conditions, removal the acetonide group in 16
using p-TsOH led to complete epimerization of stereocenter
at the C6′ spiroketal carbon, as evident by both nOe and
X-ray structure of diol 17 (Scheme 5).
B3LYP/6-31G* calculations [carried out using simplified
model where the TBDPS group is replaced with Me group]
revealed that the acetonide protected spiroketal 16 is actu-
ally 0.68 kcal mol1 more stable than its corresponding
C6′-epimer, whereas E is 2.13 kcal mol1 in favor of 17
over its C6′-epimer (Figure 3). Such a huge enhancement in
the stability is likely a result of hydrogen bonding between
C4′-OH and spiroketal oxygen, which was also seen in the
Figure 3 Energetics of 16, 17 and their respective C6′ epimers.
X-ray structure and the minimized molecular model of 17. It
is noteworthy that the minimized conformational models of

38
Figueroa R, et al. Sci China Chem January (2011) Vol.54 No.1
16 and 17 are in close agreement with those predicted from
nOe experiments (Scheme 4).
desilylation followed by oxidation of 16 gave aldehyde 19.
Snieckus’ directed ortho-metallation [31] using amide 20
[26] proved to be a perfect protocol for assembling the
benzolactone unit. Addition of 20-Li afforded a readily
separable mixture of alcohols 21-syn and 21-anti in 50%
overall yield with an isomeric ratio 1:1.4.
This key epimerization provided us with a window of
opportunity to succeed in a total synthesis of (+)-aigialospirol
4 through diol 17. However, the ensuing protecting group
game using 17 was both messy and laborious (Scheme 5).
In particular, the triol intermediate after removing the
TBDPS group in 17 was difficult to handle given its polarity
and water solubility. Fortuitously, B3LYP/6-31G* calcula-
tions once again demonstrate that the natural product
(+)-aigialospirol 4 itself is also more stable than its C6′
epimer with a E is 1.34 kcal mol1 (Figure 4) presumably
also experiencing the favourable hydrogen bonding between
C4′-OH and spiroketal oxygen (blue arrow). Consequently,
we elected to use 16 for the total synthesis and hoped to
achieve a late stage C6′ epimerization under acidic condi-
tions.
Given the uncertainty of which isomer possesses the de-
sired C1′-stereochemistry, both alcohols 21-syn and 21-anti
were used for the lactone formation. However, during these
investigations, it was found that 21-syn and 21-anti were
actually epimerizing to one another. Specifically, pure 21-
anti was epimerizing to 21-syn with prolonged exposure to
silica gel. Therefore, to properly separate 21-syn and 21-anti
via silica gel column chromatography, speed of the purifica-
tion process actually mattered. Subsequently, epimerization
of alcohol 21-anti to 21-syn was observed in C6D6 at rt.
Most remarkably, epimerization was also occurring in fro-
zen C6D6 at 10 °C with the final ratio arresting at 3:1
syn:anti, which matches the energetic differences attained
using B3LYP/6-31G* calculations (carried out using simpli-
fied models where the TBS group is replaced with Me
group) (Scheme 7).
3.2 The unusual C1′-epimerization
As shown in Scheme 6, to continue our total synthesis effort,
Scheme 7 Epimerization of 21-syn and 21-anti in frozen benzene.
Figure 4 Energetics of (+)-aigialospirol 4 and C6′-epi-4.
Scheme 6 Synthesis of epimeric C1′ alcohols via DOM.
Scheme 8 Total synthesis of (+)-aigialospirol 4.

Figueroa R, et al. Sci China Chem January (2011) Vol.54 No.1
39
To complete the total synthesis, lactone formation could
be achieved using alkaline conditions at rt for 18 h, and al-
cohol 21-syn gave lactone 22-syn in 71% yield along with
reacted starting material sans the TBS group (Scheme 8). On
the other hand, while alcohol 21-anti did indeed lactonize
under the same conditions, the yield was very low. Most
intriguingly, 21-anti led to an isomeric lactone mixture 22-syn
and 22-anti with a ratio of ~6:1 in favor of 22-syn, and the
major product was a ~1:1 mixture of 21-syn and 21-anti sans
the TBS group. This outcome implies that 21-anti was actually
readily epimerizing to 21-syn under basic conditions, and that
the rate of lactonization for 21-syn is greater than 21-anti.
At last, as anticipated from earlier calculations (Figure 4),
removal of the acetonide group in lactone 22-syn took place
concomitantly with the strategic C6' epimerization to afford
(+)-aigialospirol 4 that matched the reported spectroscopic
data [15].
3.3 Initial rationale for the C1′-epimerization
We were seriously intrigued by the C1′-epimerization be-
cause this epimerization involves the same C1′ stereocenter
that is critical in the biosynthetic relationship between (+)-
hypothemycin 7 and (+)-aigialospirol 4. While acidic or
basic conditions were not surprising given that C1′ is a
benzylic carbon, the fact that epimerization took place in
frozen benzene implies that the C1′-scrambling is remark-
able and is likely intramolecular in nature.
Figure 5 Initial thoughts for the C1′ epimerization.
The initial mechanistic proposal [23] involved the two
diastereomeric pairs of hemi-ortho-aminal intermediates
24-syn-R and 24-syn-S and, and 24-anti-R and 24-anti-S as
shown in Figure 5. The R and S designate stereochemistry
of the C7 hemi-ortho-aminal carbon. A “turnstile-like” mecha-
nism was postulated in which 24-syn-R and 24-anti-S could
undergo swapping of the original C7 amido carbonyl oxy-
gen atom (blue) with the C1′ hydroxy oxygen atom (red) [23].
A closer look through B3LYP/6-31G* calculations (car-
ried out using simplified model where the TBS group is
replaced with Me group) reveal that while 24-anti-S is more
stable than 24-anti-R, for the syn isomeric pair, 24-syn-S is
actually more stable. However, unlike 24-syn-R, 24-syn-S is
not suitable stereochemically for an intramolecular-like dis-
placement at C1′ by the C7-OH group. Moreover, while
analyzing the energetics of these hemi-ortho-aminal di-
astereomers may not be completely reflective of the actual
mechanism because of the Curtin-Hammett principle, fur-
ther modelling (Figure 6) reveals that stereoelectronically,
the C7-OH group (blue arrow in 24-anti-S-model) is not
capable of such intramolecular displacement at C1′ (red
dotted arrow in 24-anti-S-model). This analysis firmly rules
out our earlier mechanistic considerations.
Figure 6 The turnstile-like mechanism unlikely.
have to be key intermediates; and (2) reaching 24-syn-S
from 21-syn is likely faster than that of 21-anti based on E
difference of 4.02 kcal mol1 (B3LYP/6-31G*: P = Me as
simplified models) between 24-syn-S and 24-anti-S (Figure
7). In addition, we continue to believe that C1′-epimerization
is not a result of direct ionization at C1′ even under acidic
conditions because the adjacent C2′ oxygen functionality
would inductively disfavour this benzylic cation formation,
and that the resorcinol ring cannot be supportive of this
C1′-cation via resonance (inductively at best, albeit the
C7C=O group exerts the opposite effect).
However, another ionization process could take place in a
highly plausible manner. As shown in Figure 7, imidanium
ions 25-syn and 25-anti could be formed from 24-syn-S and
24-anti-S, respectively; or from 24-syn-R and 24-anti-R,
respectively: not shown but cannot be ruled out [32]. This
event renders the C1′ oxygen an excellent leaving group,
allowing (a) intermolecular nucleophilic displacement under
basic and/or acidic conditions; and (b) intramolecular-like
3.4 C1′-epimerization via an imidanium ion-pair
What remain true are: (1) hemi-ortho-aminal diastereomers

40
Figueroa R, et al. Sci China Chem January (2011) Vol.54 No.1
Scheme 9 Final assessment of C1′ stereochemistry in (+)-aigialospirol.
equilibrium, and that the chance in detecting a significant
amount of erosion of the stereochemical integrity at C1′
during either the isolation process or synthetic transforma-
tions would be slim.
In conclusion, we have described here two remarkable
epimerization processes during our pursuit of an enantiose-
lective synthesis of (+)-aigialospirol that featured a cyclic
acetal tethered ring-closing metathesis. Through modeling,
we were able to turn these two unexpected epimerizations to
our advantage to ensure a successful and concise total syn-
thesis, thereby firmly establishing cyclic acetal tethered
RCM as a competitive and powerful strategy in natural
product synthesis. More importantly, careful calculations
allowed us to fully understand the nature and the mechanis-
tic course of these two epimerizations.
Figure 7 A more likely pathway for the C1′ epimerization.
nucleophilic displacement in frozen benzene or neutral con-
ditions, when 25-syn and 25-anti likely exist as close ion-
pairs.
Most importantly, imidanium ions 25-syn and 25-anti are
highly delocalized (see the overall resonance forms), and
that 25-syn is more stable than 25-anti by 2.01 kcal mol1
(B3LYP/6-31G*), thereby also suggesting the relative ease
or faster rate of formation of 25-syn from hemi-ortho-aminal
24-syn-S (or R) over that of 25-anti. Without calculating the
actual activation energy, this assessment is consistent with
the observation that 21-syn underwent the intended lactoni-
zation quite rapidly to afford 22-syn through most likely
hydrolysis of imidanium ion 25-syn (and/or hemi-ortho-
aminal 24-syn-S). On the other hand, the formation of imi-
danium ion 25-anti from 21-anti via 24-anti-S is relatively
sluggish. Consequently, the otherwise slower C1′-epimeri-
zation process, which is formally a SN2 reaction on a highly
hindered secondary carbocenter, is now allowed to compete
favourably with hydrolysis.
Lastly, such C1′-epimerization implies that stereochemical
integrity at C1′ of the natural product (+)-aigialospirol 4
itself could also be in jeopardy through an intermediate
shown as 26-syn/anti whether during the lactonization, dur-
ing silica gel column purification, or being stored in
non-polar/neutral solvents such as benzene (Scheme 9).
However, calculations confirm that if such epimerization is
occurring, C1′-epi-4 is much less stable than 4 (E = 1.58
kcal mol1 via B3LYP/6-31G*), thereby suggesting that the
natural product (+)-aigialospirol 4 would dominate the
Authors thank PRF-AC (42106) for support and Dr. Victor Young and Ben
Kucera (University of Minnesota) for X-ray analysis. CCG thanks UW for
an NIH-CBI Training Grant. We thank Professor Masahiko Isaka [Na-
tional Center for Genetic Engineering and Biotechnology (BIOTEC),
Pathumthani, Thailand] for kindly providing the original spectra data of
(+)-aigialospirol.
1
For a reviews, see; (a) Mukaiyama T, Murakami M. Cross-coupling
reactions based on acetals. Synthesis, 1987: 1043–1054; (b) Alexakis
A, Mangeney P. Chiral acetals in asymmetric synthesis. Tetrahedron
Asy, 1990, 1: 477–511
2
3
Also see: Harmata M, Rashatasakhon P. Cycloaddition reactions of
vinyl oxocarbenium ions. Tetrahedron, 2003, 59: 2371–2395
Ko C, Feltenberger JB, Ghosh SK, Hsung RP. Gassman’s in-
tramolecular [2+2] cationic cycloaddition. Formal total syntheses of
raikovenal and epi-raikovenal. Org Lett, 2008, 10: 1971–1974
For reviews, see: (a) Aho JE, Pihko PM, Rissa TK. Nonanomeric
spiroketals in natural products: Structures, sources, and synthetic
Strategies. Chem Rev, 2005, 105: 4406–4440; (b) Mead KT, Brewer
BN. Strategies in spiroketal synthesis revisited: Recent applications
and advances. Curr Org Chem, 2003, 7: 227–256; (c) Brimble MA,
Fares FA. Synthesis of bis-spiroacetal ring systems. Tetrahedron,
1999, 55: 7661–7706; (d) Fletcher MT, Kitching W. Chemistry of
fruit flies. Chem Rev, 1995, 95: 789–828; (e) Perron F, Albizati KF.
Chemistry of spiroketals. Chem Rev, 1989, 89: 1617–1661
4

Figueroa R, et al. Sci China Chem January (2011) Vol.54 No.1
41
5
For cyclic acetal tethered IMDA, see: Wang J, Hsung RP, Ghosh SK.
Stereoselective ketal-tethered intramolecular Diels-Alder cycloaddi-
tions. An approach to the 2-oxadecalin spiroketal core of antifungal
agent Fusidilactone C. Org Lett, 2004, 6: 1939–1942
Wallace DJ. Relay ring-closing metathesis – A strategy for achieving
reactivity and selectivity in metathesis chemistry. Angew Chem Int
Ed, 2005, 44: 1912–1915
14 For application in spirastrellolide A synthesis, see: (a) Liu J, Hsung
RP. Synthesis of the C11–C23 fragment of spirastrellolide A. A
ketal- tethered rcm approach to the construction of spiroketals. Org
Lett, 2005, 6: 2273–2276; (b) Yang J-H, Liu J, Hsung RP. Synthesis
of the C1-C23 fragment of spirastrellolide A. Org Lett, 2008, 10:
2525–2528
15 Vongvilai P, Isaka M, Kittakoop P, Srikitikulchai, P, Kongsaeree P,
Thebtaranonth Y. Ketene acetal and spiroacetal constituents of the
marine fungus Aigialus parvus BCC 5311. J Nat Prod, 2004, 67:
457–460
6
7
For cyclic acetal tethered Pauson-Khand, see: Ghosh SK, Hsung RP,
Liu J. An anomeric control on remote stereochemistry in the synthe-
sis of spiroketals. J Am Chem Soc, 2005, 127: 8260–8261
For cyclic acetal tethered RCM, see: (a) Ghosh SK, Hsung RP, Wang
J. Ketal-tethered ring-closing metathesis. An unconventional ap-
proach to constructing spiroketals and total synthesis of an insect
pheromone. Tetrahedron Lett, 2004, 45: 5505–5510; (b) Ghosh SK,
Ko C, Liu J, Wang J, Hsung RP. A ketal-tethered RCM strategy to-
ward the synthesis of spiroketal related natural products. Tetrahedron,
2006, 62: 10485–10496
For reviews, see: (a) Postema MHD. Recent developments in the syn-
thesis of C-glycoside. Tetrahedron, 1992, 48: 8545–8599; (b)
Postema MHD. C-Glycoside Synthesis. Ann Arbor: CRC Press, 1995;
(c) Parker KA. Novel methods for the synthesis of C-aryl glycoside
natural products. Pure Appl Chem, 1994, 66: 2135–2138; (d) Levy
DE, Tang C. The Chemistry of C-Glycosides, 1st ed. Pergamon Press.
1995, Vol. 13
16 For biological activities of marine fungi in general, see: Faulkner DJ.
Marine natural products. Nat Prod Rep, 2001, 18: 1–49
8
9
17 (a) Nair MSR, Carey ST. Metabolites of pyrenomycetes. XIII. Struc-
ture of (+)-hypothemycin, an antibiotic macrolide from Hypomyces
trichothecoides. Tetrahedron Lett. 1980, 21: 2011–2012; (b) Nair
MSR, Carey ST, James JC. Metabolites of pyrenomycetes. XIV.
Structure and partial stereochemistry of the antibiotic macrolides hy-
pothemycin and dihydrohypothemycin. Tetrahedron, 1981, 37:
2445–2449
18 Revised structure of hypothemycin: Agatsuma T, Takahashi A,
Kabuto C, Nozoe S. Revised structure and stereochemistry of hypothe-
mycin. Chem Pharm Bull, 1993, 41: 373–375
Ko C, Hsung RP. An unusual stereoselectivity in the anomeric sub-
stitution with carbamates promoted by HNTf2. Org Biomol Chem,
2007, 7: 431–434
10 For RCM, see: (a) Leeuwenburgh MA, Appeldoorn CCM, Van Hooft
PAV, Overkleeft HA, Van der Marel GA, Van Boom JH. Synthesis
of highly functionalized carbohydrate-derived spiroacetals by ring-
closing metathesis and Pauson-Khand reaction of ketoglycosidic
enynes. Eur J Org Chem, 2000: 837–877; (b) Bassindale MJ, Hamley
P, Leitner A, Harrity JPA. Spirocycle assembly through selective
tandem ring closing metathesis reactions. Tetrahedron Lett, 1999, 40:
3247–3250; (c) Donnard M, Tschamber T, Desrat S, Hinsinger K,
Eustache J. Ring-rearrangement metathesis of 3,6-dialkoxy-3,6-dihydro-
2H-pyrans. Tetrahedron Lett, 2008, 49: 1192–1195
19 For a leading review on syntheses of hypothemycin and related re-
sorcylic macrolides, see: Winssinger N, Barluenga S. Chemistry and
biology of resorcylic acid lactones. Chem Commun, 2007: 22–36
20 For a total synthesis of hypothemycin, see: (a) Selles P, Lett R. Con-
vergent stereospecific synthesis of C292 (or LL-Z1640-2), and hy-
pothemycin. Part 1. and Convergent stereospecific synthesis of
LL-Z1640-2 (or C292), hypothemycin and related macrolides. Part 2.
Tetrahedron Lett, 2002, 43: 4621–4625, and 4627-4631; For synthe-
ses of other related resorcylic macrolides, see: (b) Yang ZQ, Geng X,
Solit D, Pratilas CA, Rosen N, Danishefsky SJ. New efficient synthe-
sis of resorcinylic macrolides via ynolides: Establishment of cyclo-
proparadicicol as synthetically feasible preclinical anticancer agent
based on Hsp90 as the Target. J Am Chem Soc, 2004, 126,
7881–7889; (c) Barluenga S, Dakas PY, Ferandin Y, Meijer L,
Winssinger N. Modular asymmetric synthesis of aigialomycin D, a
kinase-inhibitory scaffold. Angew Chem Int Ed, 2006, 45: 3951–3954;
(d) Lu J, Ma J, Xie X, Chen B, She X, Pan X. Enantioselective total
synthesis of aigialomycin D. Tetrahedron Asy, 2006, 17: 1066–1073;
(e) Grosche P, Akyel K, Marzinzik A. Synthesis of macrocyclic
scaffolds from natural products and their utilization for solid-phase
chemistry. Synthesis, 2005: 2015–2021; (f) Garbaccio RM, Stachel SJ,
Baeschlin DK, Danishefsky SJ. Concise asymmetric syntheses of
radicicol and monocillin I. J Am Chem Soc, 2001, 123: 10903–10908;
(g) Lampilas M, Lett R. Convergent stereospecific total synthesis of
monochiral monocillin I related macrolides. And convergent stereo-
specific total synthesis of monocillin I and monorden (or radicicol).
Tetrahedron Lett, 1992, 33: 773–776 and 777–780
11 For [2+2+2] cycloadditions, see: Yamamoto Y, Hashimoto T, Hattori
K, Kikuchi M, Nishiyama H. Synthesis of spirocyclic C-arylribosides
via cyclotrimerization. Org Lett, 2006, 8: 3565–3568
12 For other related RCM templated by acetals, see: (a) Scholl M,
Grubbs RH. Total synthesis of ()- and (±)-frontalin via ring-closing
metathesis. Tetrahedron Lett, 1999, 40: 1425–1428; (b) Keller VA,
Martinelli JR, Strieter ER, Burke SD. Dioxolane-to-bridged acetal-
to-spiroketal via ring-closing metathesis and rearrangement: A novel
route to 1,7-dioxaspiro[5.5]undecanes. Org Lett, 2002, 4: 467–470; (c)
Voight EA, Rein C, Burke SD. Synthesis of sialic acids via desym-
metrization by ring-closing metathesis. J Org Chem, 2002, 67:
8489–8499; (d) Burke SD, Voight EA. Formal synthesis of
(+)-3-deoxy-D-glycero-D-galacto-2-nonulosonic acid (KDN) via de-
symmetrization by ring-closing metathesis. Org Lett, 2001, 3: 237–240;
(e) Burke SD, Muller N, Beaudry CM. Desymmetrization by ring-
closing metathesis leading to 6,8-dioxabicyclo[3.2.1]octanes: A New
route for the synthesis of (+)-exo- and endo-brevicomin. Org Lett,
1999, 1: 1827–1829; (f) Kinderman SS, Doodeman R, Van Beijma
JW, Russcher JC, Tjen KCMF, Kooistra TM, Mohaselzadeh H, Van
Maarseveen JH, Hiemstra H, Schoemaker HE, Rutjes FPJT.
Ring-closing metathesis of allylic O,O- and N,O-acetals. Adv Syn
Catal, 2002, 344: 736–748
21 Isaka M, Suyarnsestakorn C, Tanticharoen M, Kongsaeree P, Theb-
taranonth Y. Aigialomycins A–E, new resorcylic macrolides from the
marine mangrove fungus Aigialus parvus. J Org Chem, 2002, 67,
1561–1566
13 For recent reviews, see: (a) Trnka TM, Grubbs RH. The development
of L2X2Ru:CHR olefin metathesis catalysts: An organometallic suc-
cess story. Acc Chem Res, 2001, 34, 18–29; (b) Hoveyda AH,
Schrock RR. Catalytic asymmetric olefin metathesis. Chem Eur J,
2001, 7: 945–950; (c) Walters MA. Recent advances in the synthesis
of heterocycles via ring-closing metathesis. Prog Heteroc Chem,
2003, 15: 1–36; (d) Nakamura I, Yamamoto Y. Transition-metal-
catalyzed reactions in heterocyclic synthesis. Chem Rev, 2004, 104:
2127–2198; (e) Deiters A, Martin SF. Synthesis of oxygen- and
nitrogen-containing heterocycles by ring-closing metathesis. Chem
Rev, 2004, 104: 2199–2238; (f) McReynolds MD, Dougherty JM,
Hanson PR. Synthesis of phosphorus and sulfur heterocycles via
ring-closing olefin metathesis. Chem Rev, 2004, 104: 2239–2258; (g)
22 (a) Tanaka H, Nishida K, Sugita K, Yoshioka T. Antitumor efficacy of
hypothemycin: A new ras-signaling inhibitor. Jpn J Cancer Res, 1999,
90: 1139–1145; (b) Schirmer A, Kennedy J, Murli S, Reid R, Santi DV.
Targeted covalent inactivation of protein kinases by resorcylic acid lac-
tone polyketides. Proc Nat Aca Sci, 2006, 103: 4234–4239
23 Figueroa R, Hsung RP, Guevarra CC. An enantioselective total syn-
thesis of (+)-aigialospirol. Org Lett, 2007, 9, 4857–4859.
24 Marshall JA, Sabatini JJ. Synthesis of cis- and trans-2,5-disubstituted
tetrahydrofurans by a tandem dihydroxylation-SN2 cyclization se-
quence. Org Lett, 2005, 7: 4819–4822
25 Kumar P, Gupta P, Naidu SV. A simple and efficient approach to
1,3-polyols: application to the synthesis of cryptocarya diacetate.
Chem Eur J, 2006, 12: 1397–1402

42
Figueroa R, et al. Sci China Chem January (2011) Vol.54 No.1
26 Prepared by following Sellès’ conditions: (a) See reference [20a]; (b)
Wissner A, Grudzinskas C. Reaction of tert-butyldimethylsilyl esters
with oxalyl chloride-dimethylformamide: Preparation of carboxylic
acid chlorides under neutral conditions. J Org Chem, 1978, 43:
3972–3974
substituent during glycopyranoside acetolysis. J Org Chem, 1997, 62:
7597–7604
31 (a) Snieckus V. Heterocycles via ortho-lithiated benzamides. Het-
erocycles 1980, 14: 1649–1676; (b) Beak P, Snieckus V. Directed
lithiation of aromatic tertiary amides: An evolving synthetic method-
ology for polysubstituted aromatics. Acc Chem Res, 1982, 15:
306–312; (c) Snieckus V. Directed ortho metalation. Tertiary amide
and O-carbamate directors in synthetic strategies for polysubstituted
aromatics. Chem Rev, 1990, 90: 879–933; (d) Anctil EJ.-G, Snieckus
V. The directed ortho metalation-cross coupling symbiosis. Regiose-
lective methodologies for biaryls and heterobiaryls. Deployment in
aromatic and heteroaromatic natural product synthesis. J Organomet
Chem, 2002, 653: 150–160
32 One of the referees kindly pointed out a third approach that could
account for either the observed epimerization through the direct for-
mation of the proposed imidanium ion pairs. This insightful sugges-
tion is also plausible and cannot be ruled out especially under the
acidic conditions. While it places less significance on the hemi-ortho-
aminal formation, it accentuates the importance of the amido-group
and implies that such epimerization may not take place in the absence
of the amido-group
27 Hansen TV. Synthesis of (R)-()-argentilactone. Tetrahedron Asy,
2002, 13: 547–550
28 For leading references on using HNTf2, see: (a) Sakakura A, Suzuki
K, Nakano K, Ishihara K. Chiral 1,1′-Binaphthyl-2,2′-diammonium
salt catalysts for the enantioselective Diels-Alder reaction with. al-
pha.-acyloxyacroleins. Org Lett, 2006, 8: 2229–2232; (b) Sun J,
Kozmin SA. Bronsted acid-promoted cyclizations of 1-siloxy-1,5-
diynes. J Am Chem Soc, 2005, 127, 13512–13513; For the acidity of
HNTf2, see: (c) Thomazeau C, Olivier-Bourbigou H, Magna L, Luts
S, Gilbert B. Determination of an acidic scale in room temperature
ionic liquids. J Am Chem Soc, 2003, 125: 5264–5265, and references
cited therein.
29 (a) Shenoy SR, Smith DM, Woerpel KA. Nucleophilic additions of
trimethylsilyl cyanide to cyclic oxocarbenium ions: Evidence for the
loss of stereoselectivity at the limits of diffusion control. J Am Chem
Soc, 2006, 128, 8671–8677; (b) Smith DM, Woerpel KA. Using
stereoelectronic effects to explain selective reactions of 4-substituted
five-membered ring oxocarbenium ions. Org Lett, 2004, 6:
2063–2066; (c) Romero JAC, Tabacco SA, Woerpel KA. Stereo-
chemical reversal of nucleophilic substitution reactions depending
upon substituent: Reactions of heteroatom-substituted six-membered-
ring oxocarbenium ions through pseudoaxial conformers. J Am Chem
Soc, 2000, 122, 168–169; (d) Matsutani H, Ichikawa S, Yaruva J,
Kusumoto T, Hiyama T. Synthesis and Lewis acid-catalyzed nucleo-
philic substitution of chiral 1-alkoxyalkyl carboxylates. J Am Chem
Soc, 1997, 119: 4541–4542; (e) Hosokawa S, Kirschbaum B, Isobe M.
1,4-Anti induction in C-glycosylation of pentose glycals. Tetrahe-
dron Lett, 1998, 39: 1917–1920
30 (a) Woods RJ, Andrews CW, Bowen JP. Molecular mechanical in-
vestigations of the properties of oxocarbenium ions. 2. Application to
glycoside hydrolysis. J Am Chem Soc, 1992, 114: 859–864; (b)
Miljkovic M, Yeagley D, Deslongchamps P, Dory YL. Experimental
and theoretical evidence of through-space electrostatic stabilization of
the incipient oxocarbenium ion by an axially oriented electronegative