Synthesis and evaluation of an Iejimalide-archazolid chimera

Article abstract of DOI:10.1016/j.tet.2010.05.043

Even though the macrolides of the iejimalide family are of marine origin, whereas those of the archazolid series derive from terrestrial myxobacteria, a comparison of their constitution, stereochemistry, and biological activity suggests that these natural products are close structural and functional relatives. Guided by this perception, compound 5 was prepared, which hybridizes the macrolactone core of iejimalide B (2) with the tail of archazolid A (3). The cytotoxicity profile of this chimera, as determined with a panel of 12 human cancer cell lines, corresponds to that of the parent compound 2, although its potency is lower. This outcome may be interpreted on the basis of molecular dynamics calculations, which suggest that the low energy conformations of 2 and 5 are similar but the energetic barriers between the relevant conformers are distinctly higher for the hybrid structure. The synthesis of 5 hinged on a regioselective functionalization of 2,4-dibromothiazole 6, a highly selective CBS-reduction of ketone 8, a Suzuki cross coupling of vinyl boronate 17 with the elaborate alkenyl iodide 16, and a productive closure of the macrocycle by RCM, which requires the selective activation of two out of eight double bonds present in the cyclization precursor 20 by the second-generation Grubbs catalyst 21.

Full text of DOI:10.1016/j.tet.2010.05.043

Tetrahedron 66 (2010) 6421e6428
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Synthesis and evaluation of an Iejimalide-archazolid chimera
Emilie Moulin , Cristina Nevado a , Julien Gagnepain , Gerhard Kelter , Heinz-Herbert Fiebig ,
Alois Fürstner
a
a
,
y
,
z
a
b
b
a,
*
Max-Planck-Institut für Kohlenforschung, D-45470 Mülheim/Ruhr, Germany
Oncotest GmbH, Am Flughafen 12-14, D-79108 Freiburg, Germany
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 1 March 2010
Received in revised form 28 April 2010
Accepted 13 May 2010
Available online 20 May 2010
Even though the macrolides of the iejimalide family are of marine origin, whereas those of the archazolid
series derive from terrestrial myxobacteria, a comparison of their constitution, stereochemistry, and
biological activity suggests that these natural products are close structural and functional relatives.
Guided by this perception, compound 5 was prepared, which hybridizes the macrolactone core of ieji-
malide B (2) with the tail of archazolid A (3). The cytotoxicity profile of this chimera, as determined with
a panel of 12 human cancer cell lines, corresponds to that of the parent compound 2, although its po-
tency is lower. This outcome may be interpreted on the basis of molecular dynamics calculations, which
suggest that the low energy conformations of 2 and 5 are similar but the energetic barriers between the
relevant conformers are distinctly higher for the hybrid structure. The synthesis of 5 hinged on a re-
gioselective functionalization of 2,4-dibromothiazole 6, a highly selective CBS-reduction of ketone 8,
a Suzuki cross coupling of vinyl boronate 17 with the elaborate alkenyl iodide 16, and a productive
closure of the macrocycle by RCM, which requires the selective activation of two out of eight double
bonds present in the cyclization precursor 20 by the second-generation Grubbs catalyst 21.
Ó 2010 Elsevier Ltd. All rights reserved.
Dedicated to Professor Steven Ley
on the occasion of his receipt of the
Tetrahedron Prize
Keywords:
Conformational analysis
Cytotoxicity
Macrolides
Metathesis
Natural product hybrids
1. Introduction
an in-depth preclinical evaluation.14 In parallel work, deliberate
digression from the total synthesis allowed us to prepare a set of
As part of our ongoing program on the synthesis and evaluation
more than 20 fully synthetic iejimalide analogs for biological test-
ing, in which the entire skeleton of the parent natural product was
subject to molecular editing.
During this endeavor, our attention was caught by the structure
and activity of archazolid A (3) and B (4), secondary metabolites
1,2
of structurally novel anticancer agents, we devoted considerable
efforts to the iejimalide family of marine macrolides (Scheme 1).
These polyunsaturated compounds were isolated by Kobayashi and
co-workers from the tunicate Eudistoma cf. rigida and later from
11,14
3
e5
a Cystodytes sp. collected off Ie island in Japan.
Early studies had
derived from terrestrial myxobacteria of the strains Archangium
6
15,16
shown that 1 and 2 and their sulfated sister compounds hold
gephyra and Cystobacter sp.
When drawn as shown in Scheme 1,
considerable promise because of their remarkable potency (average
a striking relationship to the iejimalides becomes apparent: both
families are 24-membered macrolides comprising seven double
bonds in the hydrophobic ring; their enoate moieties carry either
7
GI50 of 1 for the NCI 60 cell line cancer panel, 13 nM) and prom-
ising selectivity.4
,8e10
Preliminary data also indicated in vivo ac-
4
a
1
tivity upon intraperitoneal administration in mice.
a hydrogen atom (R ¼H, iejimalide A and archazolid B) or a methyl
1
Because of the extremely limited supply of the iejimalides from
branch (R ¼Me, iejimalide B and archazolid A) at the C.2 position; in
the natural sources, we developed the first total synthesis of these
both series, the methylated compounds are somewhat more active.
The lactone connects the enoate with a secondary alcohol at C.23
flanked by a methyl group in a 1,2-anti relationship. These chiral
centers at C.22 and C.23 are (S,S) configured in the iejimalides andthe
archazolids alike. The structural similarity further extends to the
C.18eC.21 diene and the adjacent 17S-configured methyl ether sub-
stituent displayed in both series. Moreover, it seems that even the
sensitive targets.1
1e13
Since then, the original route was refined and
up-scaled to provide gram amounts of iejimalide B (2) as needed for
*
y
Corresponding author. E-mail address: furstner@kofo.mpg.de (A. Fürstner).
Present address: Institut Charles Sadron, Université de Strasbourg, 23 Rue du
9SeOMe group with its neighboring E-configured trisubstituted al-
Loess, F-67034 Strasbourg Cedex 2, France.
z
kene of the iejimalides has correspondence in the 7SeOH group of
Present address: Organic Chemistry Institute, University of Zürich, Winter-
thurerstrasse 190, CH-8057 Zürich, Switzerland.
the archazolid located in the exact same environment. Overall, it only
0
040-4020/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2010.05.043

6422
E. Moulin et al. / Tetrahedron 66 (2010) 6421e6428
metal/halogen exchange with n-BuLi in ether at low temperature
19
(
Scheme 2). In contrast to literature recommendations, which
suggested trapping of the resulting heteroaryl lithium intermediate
2
0
with isovaleric acid ethyl ester or the corresponding nitrile, the
use of the corresponding Weinreb amide 7 was found to give ke-
2
1,22
tone 8 in much better yield and significantly higher purity.
A
2
3
subsequent CBS-reduction furnished the required secondary al-
cohol with an ee of 95%, the S-configuration of which was con-
firmed by Mosher-ester analysis. O-Silylation followed by lithium/
halogen exchange of the remaining bromide in 10 (Fig. 1) with t-
BuLi and quenching of the reactive species with DMF gave aldehyde
11 in good overall yield.
Scheme 1. Structures of the iejimalides and the archazolids, together with the tar-
geted chimera 5, which combines the iejimalide core and the archazolid side chain.
takes a formal relocation of the5,6-alkenetothe13,14-position in the
southern sector to convert the iejimalide framework into the arch-
azolid scaffold, which, however, is slightly more adorned.
Even though the side chains of the iejimalides and archazolids
are different, it is worth recognizing the common site of attachment
to the macrolactone cores at C.23 in both series. These tails consist
of a rigid hydrophobic spacer (1,3-diene in the iejimalides, thiazole
in the archazolids) terminated by a polar head group (N-formyl-
serine vs N-methyl-carbamate).
In addition to this constitutional and stereochemical similarity,
a comparison of the currently known activity data is also quite
instructive. Although one has to be careful in drawing conclusions
from a preliminary set of biochemical and biological results, it is
tempting to see more than a mere coincidence in the fact that the
iejimalides as well as the archazolids are potent and selective in-
Scheme 2. (a) n-BuLi, Et O, ꢀ78 ꢁC, then 7, 88%; (b) 9 (50 mol %), BH $SMe , THF, ꢀ30 ꢁC,
2
3
2
ꢁ
ꢁ
95%, ee 95%; (c) TBSOTf, 2,6-lutidine, CH
2
Cl
2
, 0 C, 84%; (d) t-BuLi, Et
2
O, ꢀ78 C, then
ꢁ
ꢁ
DMF, 70%; (e) Pd(OAc)
2
(20 mol %), PPh
3
(20 mol %), Et
2
Zn, THF, ꢀ78 C/ꢀ20 C, 71%,
8
,15
dr¼1:1 (only one isomer shown); (f) pivaloyl chloride, DMAP cat., pyridine, 67%; (g)
hibitors of vacuolar-type ATPases (V-ATPases).
Whether in-
ꢁ
0
Cp
CH
2
Zr(H)Cl, THF, then I
2
, 83%; (h) TBAF, THF, 0 C, 77%; (i) 1,1 -carbonyldiimidazole,
terference with these vital proton pumps is the only and/or the
decisive reason for the impressive cytotoxicity of these macrolides,
ꢁ
ꢁ
2
Cl
2
, 0 C, then MeNH
2
; (j) LiBHEt
3
, CH
2
Cl
2
, 0 C, 91% over both steps.
17
however, remains yet to be investigated in more detail.
The analogies outlined above encouraged us to prepare a chimera,
which mergesthemacrolactoneof theiejimalideswiththeside chain
of the archazolids. If such a hybrid retains the bioactivity profile
common to both parent compounds, further credence is lent to the
supposed relationship between these natural products of distinctly
different origin.18 The design of compound 5 (Scheme 1) was guided
by the perception that the side chain of the iejimalides is more
C8
C6
C5
O1
C7
C12
N1
Br1
C4
14
amenable to structural changes than the core region. The synthesis
and evaluation of this ‘designer natural product’ are outlined below.
C13 C11
C14
C3
C1
C9
Si1
2
2
. Results and discussion
C2
S1
.1. Synthesis
C10
The preparation of the required building block representing
the side chain of the targeted chimera 5 commenced with
,4-dibromothiazole (6), which was subjected to a regioselective
Figure 1. Structure of compound 10 in the solid state. Anisotropic displacement
parameters are drawn at the 50% probability level and hydrogen atoms are omitted for
clarity.
2

E. Moulin et al. / Tetrahedron 66 (2010) 6421e6428
6423
The reaction of 11 with propargyl mesylate 12 under the condi-
tions developed by Marshall and co-workers gave a 1:1 mixture of
2.2. Evaluation and structural investigations
2
4e26
diastereomeric alcohols.
Although the isomers are separable by
The cytotoxicity of the chimera 5 was evaluated in vitro in an
careful flash chromatography (cf. Experimental section), it is also
possible to process the mixture by temporary protection of the al-
cohol as the corresponding pivalate 14 in preparation for a hydro-
zirconation/iodination sequence and subsequent cleavage of the TBS
ether. At this stage, isomerically pure 15 could be readilyattained and
further elaborated to the targeted iejimalide/archazolid hybrid.
To this end, the methyl carbamate moiety was installed by se-
quential treatment of 15 with carbonyldiimidazole and methyl-
assay comprising 12 selected human cancer cell lines (Table
31,32
1).
however, that the lung adeno cancer cell lines LXFA 629L (IC50
0.32 M) and LXFL 529L (IC50 0.63 M), as well as the colorectal
cancer cell line CXF HT29 (IC50 0.49 M) were found to be sig-
An average IC50 of 2.02 mM was determined. It is of note,
m
m
m
nificantly more sensitive than average. Least sensitive were the
pancreas-(PAXF 1657L), the melanoma-(MEXF 462NL), and the
renal (RXF 486L) tumor cell lines. Overall, 5 retains a selectivity
profile similar to that of iejimalide B (2), although it is clearly less
3
amine. Cleavage of the remaining pivalate ester with LiBHEt did
not touch the carbamate and hence provided the fully functional
14,33
potent.
side chain building block 16 in readiness for fragment coupling.
11
Under previously optimized conditions, this alkenyl iodide was
11
first reacted with the known boronate 17 under palladium ca-
1
1
talysis before the somewhat labile acid 19 was attached. The
resulting compound 20 was subjected to ring-closing olefin me-
tathesis (RCM).27 It is remarkable that exposure of this sub-
stratedcontaining no less than eight double bondsdto catalytic
amounts of 2128 resulted in selective activation of the terminal al-
kenes, even though second-generation Grubbs catalysts are well
known for their ability to react with more highly substituted olefins
Table 1
Antitumor activity of compound 5 against selected human tumor cell lines
Typea
Cell line
IC50
(m
M)
IC70 (mM)
CXF HT29
GXF 251L
Colon adeno ca., pd
Gastric adeno ca., pd
Lung large cell ca., pd
Lung adeno ca., pd
Mammary adeno ca., wd
Amelanotic melanoma
Ovarian adeno ca., wd
Pancreatic adeno ca., md
Prostate ca., pd
Pleuramesothelioma
Hypernephroma, pd, clear cell
Endometrium carcino sarcoma, pd
0.49
1.30
0.63
0.32
2.49
5.99
3.49
5.34
2.46
3.27
6.56
2.40
0.94
6.21
2.15
0.71
6.49
14.07
6.17
15.20
5.05
13.22
14.13
4.99
LXFL 529L
LXFA 629L
MAXF 401NL
MEXF 462NL
OVXF 899L
PAXF 1657L
PRXF 22RV1
PXF 1752L
RXF 486L
2
9,30
under mild conditions.
In any case, the desired 24-membered
polyene 5 representing the targeted ‘hybrid natural product’ was
isolated in respectable yield as the E-isomer at the newly formed
double bond (Scheme 3).
UXF 1138L
a
ca¼carcinoma, pd¼poorly differentiated, wd¼well differentiated, md¼moder-
ately differentiated.
Extensive NMR investigations indicate that the common
macrolactone rings of 2 and 5 adopt similar but not identical
3
4
conformations, at least in CD
2
Cl
2
.
In both compounds, the 17-
OMe group resonates at significantly higher field (<3 ppm) than
the 9-OMe substituent (ca. 3.3 ppm). This distinctive shift differ-
ence has previously been attributed to the orientation of the C.17-
4
,35
OMe group toward the inside of the macrolactone ring of 2.
Likewise, the diastereomeric protons of the C.10 methylene
group in 5 shows a characteristic pattern signature similar to that
observed for 2. The comparison of the recorded NOESY/ROESY
data further advocates a high degree of conformational homology,
even though these spectra also reveal subtle differences: thus, the
reported ROESY correlations between H.10a,b and the protons of
4
the adjacent methyl ether at C.9 in 2 are missing for 5; in contrast,
distinct cross peaks in the NOESY spectra of 5 indicate a proximity
between H.20 and the methyl group branching off C.22, as well as
between H.23 and H.21; no such contacts were described for
4
iejimalide 2.
3
6
Molecular dynamic simulations with the MMFF94 force field
37
in the CHARMM program were carried out to further assess the
level of homology between and (for details, see the
2
5
Experimental section). The lowest energy conformations of both
molecules were found to be very similar, with the lateral chain
being folded back over the macrocycle. For iejimalide, however,
ꢀ1
a second conformational cluster of only slightly (0.16 kcal mol
)
higher energy is available, in which the side chain points away from
the macrolactone, the C.2 methyl branch, and the ester carbonyl are
synclinal to each other, and the C.17 methoxy group is directed
toward the interior of the ring, which adopts a puckered confor-
mation (Fig. 2). This second conformational cluster nicely corre-
sponds to the 3D-structure of 2 in solution as deduced from the
NMR spectra.
Scheme 3. (a) [(dppf)PdCl
2
]
(12 mol %), Ba(OH)
, 75%; (c) complex 21 (20 mol %), CH
2
$8H
2
O, DMF, 70%; (b) EDC$HCl,
Cl , 68%.
4
-pyrrolidinylpyridine, CH Cl
2
2
2
2

6424
E. Moulin et al. / Tetrahedron 66 (2010) 6421e6428
were used as references and the chemical shifts converted to the
TMS scale (CDCl h77.0 ppm; residual CHCl h7.26 ppm;
CD Cl h53.8 ppm; residual CHDCl h5.32 ppm); IR: Magna
IR750 (Nicolet) or Spectrum One (Perkin/Elmer) spectrometer,
3
:
d
C
3 H
: d
2
2
:
d
C
2 H
: d
~
ꢀ1
wavenumbers (V) in cm ; ESIMS: ESQ3000 (Bruker), accurate
mass determinations: Bruker APEX III FT-MS (7 T magnet) or Mat
95 (Finnigan). Elemental analyses: Kolbe, Mülheim/Ruhr. All com-
mercially available compounds (Fluka, Lancaster, Aldrich) were
used as received.
Figure 2. Representative of the computed conformational cluster of iejimalide B (2),
which matches the solution structure most closely.
3.1.1. Weinreb amide 7. A solution of isopropylmagnesium chloride
(
(
(
2 M inTHF,15 mL) was slowly added to a solution of ethyl isovalerate
ꢀ
1
The third conformational cluster of 2, which lies 1.55 kcal mol
1.51 mL, 10 mmol) and N,O-dimethylhydroxylamine hydrochloride
above the global minimum, also orients the C.17 methoxy group
inside the macrocycle; however, the torsional angle for the enoate
ꢁ
1.51 g,15.5 mmol) in THF (20 mL) at ꢀ20 C. After stirring for 40 min
4
at this temperature, the reaction was quenched with satd aq NH Cl
ꢁ
ꢁ
part C.3eC.2eC.1eO.24 is increased from þ51 in cluster 2 to þ161 ,
which brings the C.2 methyl branch in a roughly antiperiplanar
orientation relative to the carbonyl group. The large ring accom-
modates this change by adopting a rather flat shape.
(
10 mL) and warmed to room temperature. The resulting mixture
was extracted with EtOAc (3ꢂ10 mL), the combined organic phases
were washed with brine, dried over MgSO , and concentrated to give
amide 7, which was pure enough for direct use in the next reaction
4
The calculated conformers of the chimera 5 can be grouped into
four clusters. Although their gross features closely resemble those
computed for iejimalide, the energetic barriers between them are
clearly more pronounced. Specifically, the second most stable clus-
ter is already 2.1 kcal mol higher in energy than the global mini-
mum; the conformers, in which the macrocycle of 5 matches the
NMR structure of iejimalide in solution (C.17eOMe inward, C.2eMe,
and estercarbonylsynclinal toeach other, puckered macrocycle), are
even 2.8 kcal mol higher (Fig. 3). One may therefore conclude that
the parent natural product 2 and the ‘hybrid natural product’ 5
populate a qualitatively similar conformational space, but the en-
ergetic barriers are clearly more important in the latter case. Al-
though the lack of information about the conformation of these
macrolides, when bound to their biological receptor(s), makes any
firm conclusion impossible at this stage, the computational results
outlined above mayexplainwhy the selectivity profiles of 2 and 5 are
similar, whereas their potencies are quite distinct.
1
(1.41 g, 97%). H NMR (300 MHz, CDCl
3
):
d
¼3.67 (s, 3H), 3.17 (s, 3H),
2
.29 (d, J¼7.0 Hz, 2H), 2.20e2.13 (m, 1H), 0.96 (d, J¼6.6 Hz, 6H); IR
ꢀ1
(film): 2958, 2872, 1659, 1465, 1414, 1378, 1168, 1004 cm ; HRMS
þ
(ESI) calcd for C
7 15
H
O
2
N [M ] 145.11028; found: 145.11012.
ꢀ1
3
.1.2. Compound 8. A solution of n-BuLi (6.1 mL, 9.72 mmol) was
slowly added to a solution of 2,5-dibromothiazole (1.97 g, 8.1 mmol)
in Et
Weinreb amide 7 (1.41 g, 9.72 mmol) was introduced and stirring
continued for 1 h before the mixture was allowed to reach ambient
temperature over the course of 1 h. The reaction was quenched with
ꢁ
2
O (40.5 mL) at ꢀ78 C. After stirring for 1 h at this temperature,
ꢀ1
satd aq NH
3ꢂ20 mL), the combined organic phases were washed with brine,
dried over MgSO , and evaporated. Purification of the residue by
flash chromatography (SiO , hexanes/EtOAc 80:1) afforded com-
pound 8 as a colorless oil (1.77 g, 88%). H NMR (300 MHz, CDCl
¼7.56 (s, 1H), 3.02 (d, J¼7.0 Hz, 2H), 2.38e2.29 (m, 1H), 1.00 (d,
4 2
Cl (20 mL), the aqueous layer was extracted with Et O
(
4
2
1
3
):
d
1
3
J¼6.6 Hz, 6H); C NMR (75 MHz, CDCl
3
):
d
¼192.5, 167.4, 126.8, 124.7,
4
1
6.8, 24.9, 22.6; IR (film): 3119, 2959, 2924, 2868, 1689, 1457, 1387,
300, 1262, 1210, 986, 930, 890, 836, 746 cm ; MS (EI) m/z (%): 249
ꢀ1
þ
þ
(43, M ), 247 (42, M ), 207 (46), 205 (46),192 (25),190 (24),179 (32),
1
C
77 (31), 165 (32), 163 (32), 83 (32), 57 (100); HRMS (ESI) calcd for
þ
8
H10ONSBr [MþNa ] 269.95588; found: 269.95573.
3.1.3. Compound 10. Solutions of (R)-2-methyl-oxazaborolidine 9
(
1 M in toluene, 3.5 mL) and BH $SMe (3.3 mL, 34.9 mmol) were
3
2
slowly added via two dropping funnels to a solution of ketone 8
ꢁ
(
1.73 g, 6.97 mmol) in THF (42 mL) at ꢀ30 C. After stirring for 1.5 h
at this temperature, the reaction was carefully quenched with EtOH
12 mL) and warmed to room temperature. A standard extractive
work up followed by purification of the crude material by flash
chromatography (SiO , hexanes/EtOAc 50:1/8:1) afforded the
Figure 3. Representative conformer of the chimera 5 belonging to the cluster, in which
(
the macrocycle closely resembles the solution structure of iejimalide. This cluster,
ꢀ
1
however, lies ca. 2.8 kcal mol above the global minimum.
2
corresponding alcohol as a crystalline compound (1.65 g, 95%),
3
3
. Experimental
ee¼95% (HPLC: Chiracel OD-H 250 mm, B 4.6 mm, n-heptane/2-
ꢀ
1
propanol¼95:5, 0.5 mL min , 298 K, UV detection (220 nm), t
R
.1. General
(minor)¼13.74 min, t
R
(major)¼15.09 min), which analyzed as fol-
2
0
1
lows: [
a
]
D
ꢀ18.3 (c 0.42, CH
2 2 3
Cl ); H NMR (300 MHz, CDCl ):
All reactions were carried out under Ar in flame-dried glass-
d
¼7.34 (s, 1H), 5.15e5.08 (m, 1H), 2.06e1.90 (m, 1H), 1.88e1.78 (m,
13
ware. The solvents used were purified by distillation over the
drying agents indicated and transferred under Ar: THF, Et O, 1,4-
dioxane (Mg/anthracene), CH Cl , pyridine (CaH ), hexane, toluene,
Na/K), DMF (Desmodur 15, dibutyl tin dilaurate), EtOH (Mg). Flash
2H), 1.08 (d, J¼6.5 Hz, 3H), 1.07 (d, J¼2.5 Hz, 3H); C NMR (75 MHz,
CDCl ):
¼176.9, 124.4, 116.7, 70.5, 47.0, 24.6, 23.3, 21.6; IR (film):
3359, 2956, 2926, 2870, 1738, 1480, 1367, 1250, 1183, 1081, 1067,
2
3
d
2
2
2
ꢀ
1
þ
þ
(
888, 838, 733 cm ; MS (EI) m/z (%): 251 (10, M ), 249 (10, M ), 207
chromatography (FC): Merck silica gel 60 (230e400 mesh) or
CombiFlash (Teledyne Isco). NMR: spectra were recorded on Bruker
DPX 300, AMX 300, AV 400, and DMX 600 spectrometers in the
(16), 205 (16), 194 (100), 192 (97), 166 (40), 164 (40), 139 (13), 137
þ
(14), 57 (31); HRMS (EI) calcd for C
found: 248.98260.
8
H12ONSBr [M ] 248.98231;
solvents indicated; chemical shifts (
relative to TMS, coupling constants (J) in hertz. The solvent signals
d
) are given in parts per million
2,6-Lutidine (1.92 mL, 16.5 mmol) and TBSOTf (3.0 mL,
13.2 mmol) were added to a solution of this alcohol (1.65 g,

E. Moulin et al. / Tetrahedron 66 (2010) 6421e6428
6425
ꢁ
ꢁ
6
.6 mmol) in CH
stirred for 1.5 h. The reaction was quenched with satd aq NaHCO
10 mL) and extracted with Et
O (3ꢂ15 mL), the combined organic
phases were washed with brine, dried over MgSO , and evaporated.
Purification of the residue by flash chromatography (SiO , pen-
tanes/Et O 200:1/80:1) gave compound 10 as a colorless solid
2.03 g, 84%). [
CDCl ):
¼7.13 (s, 1H), 5.03 (dd, J¼7.8, 4.5 Hz, 1H), 1.84e1.68 (m,
H), 1.65e1.57 (m, 1H), 0.94 (d, J¼6.6 Hz, 3H), 0.90 (s, 9H), 0.92 (d,
2
Cl
2
(10 mL) at 0 C and the resulting mixture was
catalytic amounts of DMAP in pyridine (1.0 mL) at 0 C. The mixture
was stirred overnight at ambient temperature before the reaction
was quenched with aq 1 M HCl. A standard extractive work up
followed by purification of the residue by flash chromatography
3
(
2
4
2
(SiO
oil (68 mg, 67%). [
CDCl ):
2
, hexanes/EtOAc 100:1/50:1) yielded product 14 as a yellow
2
0
1
2
a]
D
ꢀ80.5 (c 1.0, CHCl
3
); H NMR (300 MHz,
20
1
(
a
]
D
ꢀ43.7 (c 1.15, CH
2
Cl
2
); H NMR (400 MHz,
3
d
¼7.15 (s, 1H), 5.84 (d, J¼6.6 Hz, 1H), 5.02 (dd, J¼7.7, 4.8 Hz,
3
d
1H), 3.36e3.23 (m, 1H), 1.99 (d, J¼2.4 Hz, 1H), 1.82e1.68 (m, 2H),
2
1.64e1.50 (m, 1H), 1.24 (s, 9H), 1.14 (d, J¼7.1 Hz, 3H), 0.95e0.86 (m,
13
13
J¼6.6 Hz, 3H), 0.18 (s, 3H), ꢀ0.02 (s, 3H); C NMR (100 MHz,
CDCl ):
¼178.8, 123.8, 116.4, 71.6, 48.9, 25.7, 24.0, 23.4, 22.2, 18.1,
4.80, ꢀ4.86; IR (film): 2957, 2926, 2855, 1481, 1471, 1252, 1188,
15H), 0.07 (s, 3H), ꢀ0.08 (s, 3H); C NMR (75 MHz, CDCl
3
):
d¼177.5,
3
d
174.0, 152.4, 116.3, 84.6, 73.6, 71.6, 69.8, 49.0, 40.2, 30.8, 27.1, 25.7,
ꢀ
24.1, 23.3, 22.3, 18.1, 17.4, ꢀ4.88, ꢀ4.96; IR (film): 2957, 2926, 2855,
ꢀ
1
þ
ꢀ1
1
090, 901, 832, 778 cm ; MS (EI) m/z (%): 308 (100, M ꢀt-Bu), 306
1737, 1463, 1365, 1217, 1150, 1086, 838, 778 cm ; HRMS (ESI) calcd
þ
þ
(
(
[
94, M ꢀt-Bu), 251 (6), 249 (6), 222 (5), 220 (5), 192 (2), 190 (2), 139
for C24H
41
O
3
NSSi [MþNa ] 474.24686; found: 474.24667.
13), 137 (14), 57 (31); HRMS (ESI) calcd for C14
H26ONSSiBr
þ
MþNa ] 386.05801; found: 386.05796.
3.1.7. Compound 15. A solution of compound 14 (68 mg,
.15 mmol) in THF (1.25 mL) was slowly added to a suspension of
0
3
.1.4. Compound 11. A solution of compound 10 (1.96 g, 5.38 mmol)
Cp
2
Zr(H)Cl (58.2 mg, 0.23 mmol) in THF (950
m
L) and the mixture
ꢁ
ꢁ
in Et
2
O (50 mL) was added dropwise at ꢀ78 C to a solution of
stirred in the dark for 1 h before it was cooled to 0 C. A solution of
t-BuLi (2.1 M in pentane, 7.7 mL) in Et
for 5 min, carefully dried DMF (1.25 mL, 16.14 mmol) was in-
troduced and the mixture stirred for 1.5 h before the reaction
was quenched with satd aq NH Cl (30 mL) and the mixture
4
allowed to reach ambient temperature. A standard extractive
work up followed by purification of the crude product by flash
chromatography (SiO
compound 11 as a colorless oil (1.17 g, 70%). [
CH
2
O (50 mL). After stirring
I
2
(63 mg, 0.25 mmol) in THF (1.25 mL) was added dropwise until
a pale yellow color persisted. After stirring for 5 min, the reaction
was quenched with satd aq Na (1 mL), the aqueous phase was
extracted with EtOAc (3ꢂ1 mL), the combined organic layers were
washed with brine, dried over MgSO , and evaporated. The residue
was purified by flash chromatography (SiO hexanes/EtOAc
100:1/60:1) to give the corresponding alkenyl iodide as a yellow
2 2 3
S O
4
2
,
2 2
, pentanes/Et O 80:1/10:1) furnished
2
0
20
a
]
D
ꢀ44.2 (c 0.43,
oil (73 mg, 83%), which analyzed as follows: [
CHCl ); H NMR (300 MHz, CD Cl ): d
3 2 2
8.8 Hz, 1H), 6.05 (dd, J¼14.4, 0.7 Hz, 1H), 5.72 (d, J¼6.6 Hz, 1H), 5.02
(dd, J¼7.5, 5.1 Hz, 1H), 3.04e2.95 (m, 1H), 1.80e1.71 (m, 2H),
1.64e1.56 (m, 1H), 1.21 (s, 9H), 0.97e0.84 (m, 18H), 0.10 (s, 3H),
a
]
D
ꢀ75.1 (c 1.0,
1
1
2
Cl
2
); H NMR (300 MHz, CDCl
3
):
d
¼9.98 (s, 1H), 8.11 (s, 1H),
¼7.06 (s, 1H), 6.46 (dd, J¼14.4,
5
0
3
.10 (dd, J¼7.6, 4.7 Hz, 1H), 1.88e1.71 (m, 2H), 1.70e1.60 (m, 1H),
.96 (d, J¼6.5 Hz, 3H), 0.93 (s, 9H), 0.93 (d, J¼6.4 Hz, 3H), 0.12 (s,
13
H), ꢀ0.02 (s, 3H); C NMR (75 MHz, CDCl
3
):
d
¼185.1, 179.6,
13
1
54.5, 128.0, 71.5, 48.8, 25.7, 24.0, 23.4, 22.3, 18.1, ꢀ4.9; IR (film):
ꢀ0.06 (s, 3H); C NMR (75 MHz, CD
2
Cl
2
):
d
¼177.8, 177.4, 153.4,
2
8
3
956, 2926, 2855, 1708, 1470, 1362, 1252, 1189, 1089, 1001, 900,
147.4, 116.2, 76.6, 74.4, 72.1, 49.5, 45.1, 39.2, 27.3, 25.9, 24.5, 23.4,
ꢀ1
þ
37, 808, 777 cm ; HRMS (ESI) calcd for C15
36.14240; found: 336.14235.
H
27
O
2
NSSi [MþNa ]
22.5, 18.4, 16.3, ꢀ4.67, ꢀ4.78; IR (film): 2956, 2930, 2855, 1735,
ꢀ
1
1462, 1364, 1279, 1257, 1148, 1086, 1003, 837, 777 cm ; HRMS (ESI)
þ
calcd for C24
H
42
O
3
NSSiI [MþNa ] 602.15916; found: 602.15943.
3
.1.5. Compound 13. Pd(OAc)
2
(42 mg, 0.19 mmol) and PPh
3
TBAF (1 M in THF, 250
m
L) was added to a solution of this alkenyl
ꢁ
ꢁ
(
48.2 mg, 0.19 mmol) were added at ꢀ78 C to a solution of
iodide (73 mg, 0.13 mmol) in THF (2 mL) at 0 C and the resulting
mixture stirred for 20 min before the reaction was quenched with
satd aq NH
bined extracts were washed with brine, dried over MgSO
evaporated, and the residue was purified by flash chromatography
(SiO , hexanes/EtOAc 50:1/8:1) to give compound 15 as a colorless
oil (45 mg, 77%). Alternatively, isomerically pure 15 can be obtained at
this stage by flash chromatography of the crude product mixture
resulting from the elaboration of syn/anti-13 primarily formed in the
2
4
mesylate 12 (664 mg, 4.5 mmol) in THF (27.2 mL). After stirring
for 5 min, a solution of aldehyde 11 (1.17 g, 3.7 mmol) in THF
4
Cl (2 mL) and extracted with EtOAc (3ꢂ2 mL). The com-
(
(
12 mL) was introduced, followed by the dropwise addition of ZnEt
1 M in hexane, 11.2 mL). After stirring for 30 min at this temper-
2
4
, and
ꢁ
ature, the solution was warmed to ꢀ20 C over a period of 30 min
2
and stirred overnight. The mixture was carefully quenched with
satd aq NH
temperature. The aqueous phase was extracted with EtOAc
3ꢂ20 mL), the combined organic extracts were washed with brine,
dried over MgSO , and evaporated. Purification of the crude prod-
uct by flash chromatography (SiO , pentanes/Et O 100:1/7:1)
afforded a 1:1 mixture of diastereomers 13 as a colorless oil
976 mg, 71%). A second flash chromatography with the same elu-
ent enriched this mixture in the desired anti-isomer (drꢃ7.5:1),
4
Cl (20 mL) before it was allowed to reach ambient
2
0
1
(
Marshall reaction. [
CDCl ):
a]
D
ꢀ83.2 (c 1.0, CH
2 2
Cl ); H NMR (400 MHz,
4
3
d
¼7.06 (s,1H), 6.44 (dd, J¼14.4, 8.6 Hz, 1H), 6.04 (d, J¼14.4 Hz,
2
2
1H), 5.78 (d, J¼6.5 Hz, 1H), 5.01 (dd, J¼8.1, 5.1 Hz, 1H), 3.05e2.95 (m,
1H), 2.70 (br s, 1H), 1.92e1.82 (m, 1H), 1.78e1.71 (m, 2H), 1.22 (s, 9H),
13
(
1.02e0.96 (m, 9H); C NMR (75 MHz, CD
147.3,116.1, 76.7, 74.4, 70.7, 47.7, 45.1, 39.2, 27.3, 25.0, 23.4, 22.0,16.2;IR
(film): 3451, 2958, 2931, 2870, 1732, 1479, 1464, 1280, 1149, 1069,
947 cm ; MS (EI) m/z (%): 465 (33, M ), 422 (68), 409 (15), 364 (3),
338 (7), 284 (10), 236 (22), 200 (19), 182 (95), 57 (100); HRMS (ESI)
calcd for C18
2 2
Cl ):
d¼177.5, 176.2, 153.9,
1
which showed the following characteristic data:
H
NMR
¼7.14 (d, J¼0.6 Hz, 1H), 5.00 (dd, J¼7.3, 4.9 Hz,
H), 4.65 (d, J¼5.0 Hz, 1H), 3.12e3.02 (m, 1H), 2.88 (br s, 1H), 2.09
ꢀ1
þ
(
300 MHz, CDCl
3
):
d
1
þ
(
d, J¼2.4 Hz, 1H), 1.78e1.69 (m, 2H), 1.62e1.55 (m, 1H), 1.16 (d,
H
28
O
3
NSI [MþNa ] 488.07269; found: 488.07303.
J¼7.1 Hz, 3H), 0.91e0.86 (m, 6H), 0.89 (s, 9H), 0.07 (s, 3H), ꢀ0.07 (s,
13
3
7
(
H); C NMR (75 MHz, CDCl
1.6, 71.1, 48.9, 33.6, 25.7, 24.0, 23.3, 22.4, 18.1, 17.0, ꢀ4.8, ꢀ4.9; IR
film): 3312, 2956, 2931, 2855, 1471, 1360, 1256, 1087, 1001, 902,
3
):
d¼177.4, 155.9, 114.4, 84.9, 73.3,
3.1.8. Compound 16. A solution of compound 15 (19.5 mg,
0.042 mmol) and 1,1 -carbonyldiimidazole (13.6 mg, 0.084 mmol)
in CH
0
2
Cl
2
(500
m
L) was stirred for 40 min before the mixture was
L) was
ꢀ
1
þ
ꢁ
8
38, 778 cm ; MS (EI) m/z (%): 310 (100, M ꢀt-Bu), 256 (24), 224
cooled to 0 C. A solution of methylamine (2 M in THF, 84
m
(
[
6), 249 (6), 182 (22), 140 (6); HRMS (ESI) calcd for C19
MþNa ] 390.18935; found: 390.18919.
H O
33 2
NSSi
added and stirring continued for 2 h, before the mixture was
acidified with aq HCl (1 M, 1 mL) and extracted with Et
O (3ꢂ1 mL).
The combined organic phases were washed with brine, dried over
MgSO , and evaporated, and the residue was purified by flash
chromatography (SiO , hexanes/EtOAc 20:1/5:1) to give the
þ
2
3.1.6. Compound 14. Pivaloyl chloride (84
mL, 0.68 mmol) was
4
added to a solution of alcohol 13 (83 mg, 0.23 mmol, drꢃ7.5:1) and
2

6426
E. Moulin et al. / Tetrahedron 66 (2010) 6421e6428
corresponding carbamate as a colorless oil (20 mg), which analyzed
3H), 4.84 (br d, J¼4.7 Hz, 1H), 4.03 (td, J¼9.1, 6.4 Hz, 1H), 3.48 (dt,
J¼7.4, 6.1 Hz, 2H), 3.30e3.23 (m, 1H), 3.20 (s, 3H), 3.16 (s, 3H), 3.03
(dt, J¼7.6, 7.2 Hz, 1H), 2.77 (d, J¼4.9 Hz, 3H), 2.36e2.31 (m, 1H),
2.23e2.13 (m, 2H), 1.90e1.81 (m, 2H), 1.85 (d, J¼1.4 Hz, 3H), 1.79 (d,
J¼1.1 Hz, 3H), 1.77 (d, J¼1.2 Hz, 3H), 1.72e1.66 (m, 1H), 1.63e1.58
(m, 1H), 1.49e1.43 (m, 1H), 1.13 (d, J¼6.8 Hz, 3H), 0.96e0.94 (m,
20
1
as follows: [
a
]
D
ꢀ112.2 (c 1.0, CH
2 2 2 2
Cl ); H NMR (400 MHz, CD Cl ):
d
¼7.07 (s, 1H), 6.45 (dd, J¼14.4, 8.7 Hz, 1H), 6.05 (d, J¼14.4 Hz, 1H),
5
3
.97 (dd, J¼8.4, 5.6 Hz, 1H), 5.75 (d, J¼6.3 Hz, 1H), 4.84 (br s, 1H),
.01e2.95 (m, 1H), 2.78 (d, J¼4.8 Hz, 3H), 1.88e1.82 (m, 2H),
13
1.73e1.67 (m, 1H), 1.22 (s, 9H), 0.99e0.95 (m, 9H); C NMR
13
(
75 MHz, CD
2
Cl
2
):
d
¼177.5, 171.6, 156.2, 154.4, 147.2, 116.1, 76.7, 74.5,
9H); C NMR (150 MHz, CD
2
Cl
2
):
d
¼171.6, 167.5, 156.2, 154.6, 145.7,
7
2
5
2.2, 45.2, 44.6, 39.2, 27.8, 27.3, 25.0, 23.1, 22.3, 16.2; IR (film): 3441,
136.8, 135.3, 135.1, 134.0, 133.8, 133.0, 132.9, 132.8, 132.0, 131.6,
131.0, 130.9, 126.7, 116.8, 116.8, 113.4, 81.5, 77.0, 75.6, 72.3, 56.1, 56.0,
44.7, 41.4, 40.3, 36.8, 35.9, 27.8, 24.9, 23.5, 23.0, 22.2, 20.4, 19.8, 16.7,
13.1, 12.6; IR (film): 2956, 2926, 2870, 1732, 1717, 1522, 1448, 1367,
ꢀ1
956, 2931, 2870, 1728, 1524, 1251, 1149, 971 cm ; MS (EI) m/z (%):
þ
22 (6, M ), 465 (53), 448 (21), 422 (30), 409 (18), 395 (6), 257 (21),
þ
1
82 (100), 140 (9); HRMS (ESI) calcd for C20
H
31
O
4
N
2
SI [MþNa ]
ꢀ
1
5
45.09415; found: 545.09449.
A solution of LiBHEt (1 M in THF, 134
tion of this product in CH Cl
1248, 1125, 1097, 991, 968 cm ; HRMS (ESI) calcd for C42
62
H O
6
N
2
S
þ
3
m
L) was added to a solu-
[MþNa ] 745.42208; found: 745.42210.
ꢁ
2
2
(2.2 mL) at 0 C and the resulting
mixture stirred for 1 h before the reaction was quenched with
EtOAc (1 mL). A standard extractive work up followed by flash
3.1.11. Compound 5. Complex 21 (0.6 mg, 0.0007 mmol) was added
to a solution of compound 20 (5.1 mg, 0.007 mmol) in CH Cl
2
2
chromatography of the crude material (SiO
2
, hexanes/EtOAc,
(20 mL) and the mixture was stirred at ambient temperature for
24 h. At this point, a second portion of 21 (0.6 mg, 0.0007 mmol)
was introduced and stirring continued for another 24 h before the
reaction was quenched with ethyl vinyl ether (100 mL). After stirring
for 1 h, all volatile materials were evaporated and the residue pu-
1
0:1/2:1) furnished product 16 as a colorless oil (16.5 mg, 91%
20
1
over both steps). [
CD
a
]
D
ꢀ83.5 (c 1.65, CH
2 2
Cl ); H NMR (300 MHz,
2
Cl
2
):
d
¼7.09 (d, J¼0.5 Hz, 1H), 6.54 (dd, J¼14.5, 8.1 Hz, 1H), 6.06
(
dd, J¼14.5, 0.9 Hz, 1H), 5.97 (dd, J¼8.6, 5.3 Hz, 1H), 4.85 (br s, 1H),
4
3
0
.59 (dd, J¼5.8, 5.8 Hz, 1H), 2.80e2.71 (m, 1H), 2.77 (d, J¼4.8 Hz,
rified by flash chromatography (hexanes/EtOAc, 10:1/4:1) to give
2
0
H), 2.63 (d, J¼6.0 Hz, 1H), 1.89e1.80 (m, 2H), 1.75e1.68 (m, 1H),
product 5 as a pale yellow oil (3.3 mg, 68%). [
CH Cl ); H NMR (600 MHz, CD Cl ): d
2 2 2 2
a
]
D
ꢀ78.7 (c 0.15,
1
3
1
.99 (d, J¼6.9 Hz, 3H), 0.96 (d, J¼6.5 Hz, 6H); C NMR (75 MHz,
¼7.24 (s,1H), 6.60 (dd, J¼10.3,
CD
2
Cl
2
):
d
¼171.8, 157.9, 148.0, 125.8, 114.8, 76.4, 74.0, 72.4, 47.2,
1.4 Hz, 1H), 6.46 (d, J¼15.5 Hz, 1H), 6.14 (dd, J¼15.0, 10.5 Hz, 1H),
6.03e5.98 (m, 2H), 5.90 (d, J¼15.5 Hz, 1H), 5.75 (d, J¼9.8 Hz, 1H),
5.53 (ddd, J¼15.2, 10.1, 4.9 Hz, 1H), 5.50e5.39 (m, 3H), 5.18 (dd,
J¼10.3, 5.8 Hz, 1H), 5.09 (d, J¼9.6 Hz, 1H), 4.85 (br q, J¼4.7 Hz, 1H),
4
4.8, 30.5, 25.0, 23.1, 22.3, 15.5; IR (film): 3337, 2958, 2870, 1706,
ꢀ1
1525, 1467, 1369, 1255, 1130, 950 cm ; HRMS (ESI) calcd for
þ
C
H
15 23
O
3
N
2
SI [MþNa ] 461.03663; found: 461.03702.
4
.11 (td, J¼9.8, 2.8 Hz,1H), 3.30e3.26 (m,1H), 3.20 (s, 3H), 3.18e3.12
3
(
.1.9. Compound 18. Ba(OH)
dppf)PdCl (4.0 mg, 0.0055 mmol) were added to a degassed so-
lution of compound 16 (16.2 mg, 0.037 mmol) and boronate 17
13.0 mg, 0.044 mmol) in DMF (810 L). The mixture was stirred for
h at ambient temperature before it was poured on ice-cold aq HCl
1 M, 2 mL). The aqueous phase was extracted with EtOAc (3ꢂ3 mL)
and the combined organic phases were washed with brine, dried
over MgSO , and evaporated. The residue was purified by flash
chromatography (SiO , hexanes/EtOAc 10:1/1:1) to give product
8 as a colorless oil (12.3 mg, 70%). [
NMR (400 MHz, CD Cl ):
¼7.09 (s, 1H), 6.78 (dd, J¼17.3, 10.8 Hz,
H), 6.17e6.07 (m, 2H), 5.98 (dd, J¼8.7, 5.2 Hz, 1H), 5.65 (dd, J¼14.3,
2
$8H
2
O (17.5 mg, 0.055 mmol) and
(m, 1H), 3.02e2.97 (m, 1H), 2.97 (s, 3H), 2.77 (d, J¼4.8 Hz, 3H), 2.61
(br d, J¼13.8 Hz, 1H), 2.54e2.47 (m, 1H), 2.31 (ddd, J¼13.8, 10.0,
10.0 Hz, 1H), 1.91e1.85 (m, 2H), 1.82e1.79 (m, 1H), 1.79 (d, J¼1.5 Hz,
3H), 1.77 (s, 3H), 1.74 (d, J¼1.2 Hz, 3H), 1.72e1.68 (m, 1H), 1.62e1.57
(m, 1H), 1.34e1.27 (m, 1H), 1.03 (d, J¼6.7 Hz, 3H), 0.96 (d, J¼6.5 Hz,
2
(
6
(
m
13
3H), 0.95 (d, J¼6.5 Hz, 3H), 0.89 (d, J¼6.8 Hz, 3H); C NMR
(150 MHz, CD
2
Cl
2
):
d
¼172.1, 167.7, 156.2, 154.1, 146.1, 137.1, 135.7,
4
133.8,133.7,133.2,132.4,132.3,131.8,131.5,129.7,128.8,125.7,125.3,
117.9, 80.0, 76.9, 74.8, 72.5, 56.4, 55.8, 44.8, 42.8, 40.9, 38.2, 35.2,
27.8, 24.9, 23.2, 23.1, 22.1, 21.4, 20.7, 16.6, 13.1, 12.2; IR (film): 2956,
2
2
0
1
1
a
]
D
ꢀ32.8 (c¼1.0, CH
2 2
Cl ); H
ꢀ
1
2
2
d
2931,1732,1721,1522,1448,1367,1254,1125,1094, 993 cm ; HRMS
þ
1
(ESI) calcd for C40
H
58
O
6
N
2
S [MþNa ] 717.39078; found: 717.39095.
7
.8 Hz, 1H), 5.42 (dd, J¼14.2, 7.9 Hz, 1H), 5.38 (t, J¼7.8 Hz, 1H), 5.19
(
d, J¼17.2 Hz, 1H), 5.07 (dd, J¼11.0, 1.6 Hz, 1H), 4.84 (br s, 1H), 4.57
3.1.12. X-ray crystal structure analysis of compound 10.
ꢀ
1
(
t, J¼5.7 Hz, 1H), 3.51 (dd, J¼7.3, 6.1 Hz, 1H), 3.20 (s, 3H), 2.78 (d,
C
14
H
26BrNOSSi, M
r
¼364.42 gmol , colorless plate, crystal size
J¼5.0 Hz, 3H), 2.78e2.73 (m, 1H), 2.48 (d, J¼5.7 Hz, 1H), 2.25e2.14
0.35ꢂ0.31ꢂ0.16 mm, monoclinic, space group P2
1
, a¼8.9105(13) Å,
ꢁ
3
(
m, 2H), 1.94e1.77 (m, 2H), 1.80 (d, J¼1.0 Hz, 3H), 1.74e1.69 (m, 1H),
.67e1.44 (m, 2H), 0.99 (d, J¼6.9 Hz, 3H), 0.96 (d, J¼6.5 Hz, 6H); IR
b¼10.4868(16) Å, c¼9.6494(14) Å,
b
3
¼94.209(4) , V¼899.2(2) Å ,
1
T¼100 K, Z¼2,
D
calcd¼1.346 gcm ,
l
¼1.54178 Å,
m
(Cu-K
a
)¼
ꢀ
1
(film): 3358, 2956, 2916, 2870, 1717, 1523, 1459, 1369, 1254, 1127,
4.780 mm
,
Gaussian absorption correction (Tmin¼0.35,
ꢀ
1
þ
ꢁ
1096, 991 cm ; HRMS (ESI) calcd for C26
H
40
O
4
N
2
S [MþNa ]
T
max¼0.55), Proteum X8 diffractometer, 4.59<
q
<68.82 , 20,675
4
99.26010; found: 499.25999.
measured reflections, 3223 independent reflections, 3211 re-
flections with I>2 (I), Structure solved by direct methods and re-
fined by full-matrix least-squares against F to R
wR
¼0.060, 179 parameters, Absolute structure parameter¼0.020
s
2
3.1.10. Compound 20. EDC$HCl (2.9 mg, 0.015 mmol) and 4-pyrro-
1
¼0.024 [I>2
s(I)],
lidino-pyridine (0.3 mg, 0.0023 mmol) were added to a solution of
2
ꢁ
acid 19 (4.5 mg, 0.017 mmol) in CH
2
Cl
2
(120
mL) at 0 C. The mixture
(14), H atoms riding, S¼1.108, residual electron density 0.3/
3
ꢀ
was warmed to ambient temperature for 15 min and re-cooled to
ꢀ0.3 e Å . CCDC 763567 contains the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Center via www.
ccdc.ca.ac.uk/data_request/cif.
ꢁ
0
C before alcohol 18 (7.3 mg, 0.015 mmol) was introduced. After
stirring for 70 h at ambient temperature, EtOAc (2 mL) was added,
the organic phase was washed with brine (1 mL), dried over MgSO
and evaporated, and the residue purified by flash chromatography
SiO , hexanes/EtOAc, 10:1/5:1) to yield product 20 as a colorless
oil (8.15 mg, 75%). [
CD Cl ):
¼7.13 (s, 1H), 6.76 (ddd, J¼17.4, 10.9, 0.8 Hz, 1H), 6.61 (dd,
J¼9.8, 1.3 Hz, 1H), 6.13e6.05 (m, 3H), 5.98 (dd, J¼8.9, 5.2 Hz, 1H),
.77 (ddt, J¼17.2, 10.3, 7.0 Hz, 1H), 5.76 (d, J¼7.1 Hz, 1H), 5.61 (dd,
J¼14.4, 8.4 Hz, 1H), 5.59 (dd, J¼15.5, 7.1 Hz, 1H), 5.40e5.35 (m, 2H),
.26 (d, J¼9.2 Hz, 1H), 5.18 (dd, J¼17.3, 1.0 Hz, 1H), 5.07e4.99 (m,
4
,
3.1.13. Computational methods. The molecules were simulated by
molecular dynamics with the Merck Molecular Force Field
(
2
2
0
1
36
37
a]
D
ꢀ30.7 (c 0.51, CH
2
Cl
2
); H NMR (600 MHz,
(MMFF94) in the CHARMM program, applying a dielectric
constant of 80 to simulate the effect of solvent. The simulations
were carried out at 1000 K to ensure that conformational barriers
are readily crossed. Possible trajectories starting from five con-
formationally different starting points were calculated and 500
frames were extracted from each trajectory (0.002 ps per step and
2
2
d
5
5

E. Moulin et al. / Tetrahedron 66 (2010) 6421e6428
6427
5
ꢂ100 steps¼10 ns). For each conformer, an energy minimization
4. (a) Nozawa, K.; Tsuda, M.; Ishiyama, H.; Sasaki, T.; Tsuruo, T.; Kobayashi, J.
Bioorg. Med. Chem. 2006, 14, 1063e1067; (b) Tsuda, M.; Nozawa, K.; Shimbo, K.;
Ishiyama, H.; Fukushi, E.; Kawabata, J.; Kobayashi, J. Tetrahedron Lett. 2003, 44,
was performed by 750 steps of the steepest descent (SD) algorithm
in CHARM, and their MMFF energy was calculated. The 2500 con-
formers (500 conformers per trajectory) were then clustered to
determine the dominant conformational space. The lowest energy
conformation (LEC) was taken as representative of the first cluster,
1395e1399.
5. Gerwick and co-workers suggested that iejimalide might actually be produced
by cyanobacteria, see: Simmons, T. L.; Coates, R. C.; Clark, B. R.; Engene, N.;
Gonzalez, D.; Esquenazi, E.; Dorrestein, P. C.; Gerwick, W. H. Proc. Natl. Acad. Sci.
U.S.A. 2008, 105, 4587e4594.
and all conformations having
a
root-mean-square deviation
6. The naturally occurring iejimalides C and D are the serine-O-sulfate esters of 1
and 2.
(
RMSD) lower than 3 Å were compiled into that cluster. Then, the
7. The activity data are available from the NCI homepage, see: http://www.dtp.nci.
LEC of the remaining conformations was taken as starting point for
the second cluster, and this process was iterated until all 2500
conformers were clustered.
nih.gov/docs/dtp_search.html.
8. Kazami, S.; Muroi, M.; Kawatani, M.; Kubota, T.; Usui, T.; Kobayashi, J.; Osada, H.
Biosci. Biotechnol. Biochem. 2006, 70, 1364e1370.
9. Kobayashi co-workers reported that a COMPARE analysis showed no significant
correlation to the activity profile of recorded anticancer agents, cf. Ref. 4a.
10. Wang, W.-L. W.; McHenry, P.; Jeffrey, R.; Schweitzer, D.; Helquist, P.; Tennis-
wood, M. J. Cell. Biochem. 2008, 105, 998e1007.
1. Fürstner, A.; Nevado, C.; Waser, M.; Tremblay, M.; Chevrier, C.; Teplý, F.; Aïssa,
3
.1.14. Cytotoxicity assays. Ten cell lines were established from
patient-derived tumor xenografts passaged subcutaneously in
nude mice; the origin of the donor xenografts has already been
described.31 The other cell lines were obtained from American Type
Culture Collection (PRXF 22RV1), Rockville, MD, USA or the Na-
tional Cancer Institute (CXF HT29), Bethesda, MD, USA. Authenticity
of all cell lines was proven by STR (short tandem repeat) analysis.
1
C.; Moulin, E.; Müller, O. J. Am. Chem. Soc. 2007, 129, 9150e9161.
12. (a) Fürstner, A.; Aïssa, C.; Chevrier, C.; Teplý, F.; Nevado, C.; Tremblay, M. Angew.
Chem., Int. Ed. 2006, 45, 5832e5837; (b) Fürstner, A.; Nevado, C.; Tremblay, M.;
Chevrier, C.; Teplý, F.; Aïssa, C.; Waser, M. Angew. Chem., Int. Ed. 2006, 45,
5
837e5842.
3. For a second total synthesis, see: Schweitzer, D.; Kane, J. J.; Strand, D.; McHenry,
P.; Tenniswood, M.; Helquist, P. Org. Lett. 2007, 9, 4619e4622.
4. Unpublished results.
5. (a) Sasse, F.; Steinmetz, H.; H o€ fle, G.; Reichenbach, H. J. Antibiot. 2003, 56,
20e525; (b) Hassfeld, J.; Farès, C.; Steinmetz, H.; Carlomagno, T.; Menche, D.
1
ꢁ
All cells were grown at 37 C in a humidified atmosphere (95% air,
% CO2) in RPMI 1640 medium (PAA, C o€ lbe, Germany) supple-
1
1
5
ꢀ1
mented with 10% fetal calf serum (PAA) and 0.1 mg mL genta-
5
micin (PAA). A modified propidium iodide assay was used to assess
Org. Lett. 2006, 8, 4751e4754; (c) Farès, C.; Hassfeld, J.; Menche, D.; Carlo-
magno, T. Angew. Chem., Int. Ed. 2008, 47, 3722e3726; (d) Huss, M.; Sasse, F.;
Kunze, B.; Jansen, R.; Steinmetz, H.; Ingenhorst, G.; Zeeck, A.; Wieczorek, H.
BMC Biochem. 2005, 6, 13 (10 pp).
31
the effects of the compounds. Tumor derived cell lines were in-
cubated in 96 multi-well plates. After one day, the compounds
under test were added to the plates at five concentrations in the
16. Total syntheses: (a) Menche, D.; Hassfeld, J.; Li, J.; Rudolph, S. J. Am. Chem. Soc.
ꢀ1
2007, 129, 6100e6113; (b) Roethle, P. A.; Chen, I. T.; Trauner, D. J. Am. Chem. Soc.
range from 0.001 mg up to 10 mg mL and left for further four days.
2
007, 129, 8960e8961; (c) Menche, D.; Hassfeld, J.; Li, J.; Mayer, K.; Rudolph, S.
The inhibition of proliferation was determined by measuring the
DNA content using an aqueous propidium iodide solution
J. Org. Chem. 2009, 74, 7220e7229.
17. For a review on V-ATPases as possible cancer targets, see: Pérez-Sayáns, M.;
Somoza-Martín, J. M.; Barros-Angueira, F.; Gándara Rey, J. M.; García-García, A.
Cancer Treat. Rev. 2009, 35, 707e713.
ꢀ1
(7 mg mL ). Fluorescence was measured using the Cytofluor 4000.
In each experiment, all data points were determined in triplicate.
1
8. Reviews on natural product hybrids: (a) Tietze, L. F.; Bell, H. P.; Chandrasekhar,
S. Angew. Chem., Int. Ed. 2003, 42, 3996e4028; (b) Gademann, K. Chimia 2006,
0, 841e845.
9. Dondoni, A.; Fantin, G.; Fogagnolo, M.; Medici, A.; Pedrini, P. J. Org. Chem. 1988,
3, 1748e1761.
20. Spieß, A.; Heckmann, G.; Bach, T. Synlett 2004, 131e133.
1. Nahm, S.; Weinreb, S. M. Tetrahedron Lett. 1981, 22, 3815e3818.
6
Acknowledgements
1
5
Generous financial support by the MPG, the Chemical Genomics
Center (CGC) of the MPG, the Fonds der Chemischen Industrie, the
Alexander-von-Humboldt-Foundation (fellowships to E.M. and C.
N.), and the Association pour la Recherche sur le Cancer (fellowship
to E.M.), are gratefully acknowledged. We thank Mr. J. Rust and Dr.
C.W. Lehmann for solving the X-ray structure, and the NMR- and
chromatography departments of our Institute for expert support of
our programs.
2
2
2. In our hands, the use of isovaleric acid ethyl ester as the electrophile invariably
resulted in partial loss of the second bromide atom; this by-product is difficult
to separate from 8.
3. Review: Corey, E. J.; Helal, C. J. Angew. Chem., Int. Ed. 1998, 37, 1986e2012.
4. Reviews: (a) Marshall, J. A. J. Org. Chem. 2007, 72, 8153e8166; (b) Marshall, J. A.
Chem. Rev. 2000, 100, 3163e3185.
25. This outcome shows that Marshall propargylations are highly dependent on
the nature of the chosen aldehyde. Whereas aliphatic aldehydes generally
result in excellent syn/anti ratios, aromatic aldehydes usually give low
diastereomeric ratios.
2
2
2
6. For applications of Marshall reactions to natural product synthesis from this
laboratory, see: (a) O’Neil, G. W.; Ceccon, J.; Benson, S.; Collin, M.-P.;
Fasching, B.; Fürstner, A. Angew. Chem., Int. Ed. 2009, 48, 9940e9945; (b)
Fürstner, A.; Kattnig, E.; Lepage, O. J. Am. Chem. Soc. 2006, 128, 9194e9204;
References and notes
1. For selected examples, see: (a) Benson, S.; Collin, M.-P.; O’Neil, G. W.; Ceccon, J.;
Fasching, B.; Fenster, M. D. B.; Godbout, C.; Radkowski, K.; Goddard, R.; Fürstner,
A. Angew. Chem., Int. Ed. 2009, 48, 9946e9950; (b) Fürstner, A.; Bouchez, L. C.;
Morency, L.; Funel, J.-A.; Liepins, V.; Porée, F.-H.; Gilmour, R.; Laurich, D.;
Beaufils, F.; Tamiya, M. Chem.dEur. J. 2009, 15, 3983e4010; (c) Buchgraber, P.;
Snaddon, T. N.; Wirtz, C.; Mynott, R.; Goddard, R.; Fürstner, A. Angew. Chem., Int.
Ed. 2008, 47, 8450e8454; (d) Fürstner, A.; Bindl, M.; Jean, L. Angew. Chem., Int.
Ed. 2007, 46, 9275e9278; (e) Fürstner, A.; Kennedy, J. W. J. Chem.dEur. J. 2006,
(c) Be s¸ ev, M.; Brehm, C.; Fürstner, A. Collect. Czech. Chem. Commun. 2005, 70,
1696e1708.
2
7. (a) Trnka, T. M.; Grubbs, R. H. Acc. Chem. Res. 2001, 34, 18e29; (b) Fürstner, A.
Angew. Chem., Int. Ed. 2000, 39, 3013e3043; (c) Nicolaou, K. C.; Bulger, P. G.;
Sarlah, D. Angew. Chem., Int. Ed. 2005, 44, 4490e4527; (d) Hoveyda, A. H.;
Zhugralin, A. R. Nature 2007, 450, 243e251.
2
2
3
8. Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999, 1, 953e956.
9. Vougioukalakis, G. C.; Grubbs, R. H. Chem. Rev. 2010, 110, 1746e1787.
0. For early work from this laboratory, see: (a) Ackermann, L.; Fürstner, A.;
Weskamp, T.; Kohl, F. J.; Herrmann, W. A. Tetrahedron Lett. 1999, 40,
12, 7398e7410; (f) Fürstner, A.; Domostoj, M. M.; Scheiper, B. J. Am. Chem. Soc.
2006, 128, 8087e8094; (g) Fürstner, A.; Ackerstaff, J. Chem. Commun. 2008,
2870e2872; (h) Fürstner, A.; Nagano, T. J. Am. Chem. Soc. 2007, 129, 1906e1907.
2
. For the synthesis-driven mapping of cytotoxic natural products from this lab-
4787e4790; (b) Fürstner, A.; Thiel, O. R.; Ackermann, L.; Schanz, H.-J.; Nolan, S.
oratory, see: (a) Nagano, T.; Pospí ꢀs il, J.; Chollet, G.; Schulthoff, S.; Hickmann, V.;
P. J. Org. Chem. 2000, 65, 2204e2207; (c) Fürstner, A.; Ackermann, L.; Gabor, B.;
Goddard, R.; Lehmann, C. W.; Mynott, R.; Stelzer, F.; Thiel, O. R. Chem.dEur. J.
Moulin, E.; Herrmann, R.; Müller, R.; Fürstner, A. Chem.dEur. J. 2009, 15,
9697e9706; (b) Buchgraber, P.; Domostoj, M. M.; Scheiper, B.; Wirtz, C.; My-
2
2
001, 7, 3236e3253; (d) Ackermann, L.; El Tom, D.; Fürstner, A. Tetrahedron
000, 56, 2195e2202.
nott, R.; Rust, J.; Fürstner, A. Tetrahedron 2009, 65, 6519e6534; (c) Fürstner, A.;
Flügge, S.; Larionov, O.; Takahashi, Y.; Kubota, T.; Kobayashi, J. Chem.dEur. J.
31. For a description of the assays, see: (a) Fiebig, H. H.; Berger, D. P.; Dengler, W. A.;
2
009, 15, 4011e4029; (d) Gao, W.; Chen, A. P.-C.; Leung, C.-H.; Gullen, E. A.;
Fürstner, A.; Shi, Q.; Wei, L.; Lee, K.-H.; Cheng, Y.-C. Bioorg. Med. Chem. Lett.
008, 18, 704e709; (e) Bindl, M.; Jean, L.; Herrmann, J.; Müller, R.; Fürstner, A.
Wallbrecher, E.; Winterhalter, B. R. Contrib. Oncol. In Immunodeficient Mice in
Oncology; Berger, D. P., Fiebig, H. H., Eds.; Karger: Basel, 1992; Vol. 42,
pp 321e351; (b) Roth, T.; Burger, A. M.; Dengler, W.; Willmann, H.; Fiebig, H. H.
Contrib. Oncol. In Relevance of Tumor Models for Anticancer Drug Development;
Fiebig, H. H., Burger, A. M., Eds.; Karger: Basel, 1999; Vol. 54, pp 145e156; (c)
Dengler, W. A.; Schulte, J.; Berger, D. P.; Mertelsmann, R.; Fiebig, H. H. Anti-
Cancer Drugs 1995, 6, 522e532.
2
Chem.dEur. J. 2009, 15, 12310e12319; (f) Fürstner, A.; Nagano, T.; Müller, C.;
Seidel, G.; Müller, O. Chem.dEur. J. 2007, 13, 1452e1462; (g) Fürstner, A.; Kirk,
D.; Fenster, M. D. B.; Aïssa, C.; De Souza, D.; Nevado, C.; Tuttle, T.; Thiel, W.;
Müller, O. Chem.dEur. J. 2007, 13, 135e149.
3
. (a) Kobayashi, J.; Cheng, J.; Ohta, T.; Nakamura, H.; Nozoe, S.; Hirata, Y.;
Ohizumi, Y.; Sasaki, T. J. Org. Chem. 1988, 53, 6147e6150; (b) Kikuchi, Y.; Ishi-
bashi, M.; Sasaki, T.; Kobayashi, J. Tetrahedron Lett. 1991, 32, 797e798.
32. Fürstner, A.; Kattnig, E.; Kelter, G.; Fiebig, H.-H. Chem.eEur. J. 2009, 15,
4030e4043.

6428
E. Moulin et al. / Tetrahedron 66 (2010) 6421e6428
3
3. In this context it is of note that the few other synthetic archazolid analogues
known in the literature were also invariably less active than the parent natural
product, see: Menche, D.; Hassfeld, J.; Sasse, F.; Huss, M.; Wieczorek, H. Bioorg.
Med. Chem. Lett. 2007, 17, 1732e1735.
only one of the eOMe groups was noticed during the RCM experiments in the
iejimalide series, cf. Ref. 11.
36. (a) Halgren, T. A. J. Comput. Chem. 1996, 17, 616e641; (b) Halgren, T. A.; Nachbar,
R. B. J. Comput. Chem. 1996, 17, 587e615; (c) Halgren, T. A. J. Comput. Chem. 1996,
17, 553e586; (d) Halgren, T. A. J. Comput. Chem. 1996, 17, 520e552; (e) Halgren,
T. A. J. Comput. Chem. 1996, 17, 490e519.
3
4. Kobayashi and co-workers showed that the conformations of 2 in CD
2 2
Cl and
4
[D ]-methanol are very similar, cf. Ref. 4a.
3
5. The notion that this characteristic difference is caused by a conformational
effect of the macrocycle is in line with the observation that the eOMe ethers
resonate at 3.16 and 3.20 ppm in the cyclization precursor 20, but at 2.97 and 3.
37. (a) CHARMM_33b1, 2006. (b) Brooks, B. R.; Bruccoleri, R. E.; Olafson, B.
D.; States, D. J.; Swaminathan, S.; Karplus, M. J. Comput. Chem. 1983, 4,
187e217; (c) Brooks, C. L.; Karplus, M. J. Chem. Phys. 1983, 79,
6312e6325.
20 ppm in the ring-closed chimera 5. A similarly characteristic shift change of