A biogenetically based strategy towards the polycyclic core skeleton of sarain A

Article abstract of DOI:10.1002/ejoc.200600293

An approach to the polycyclic core of sarain A, based upon a proposed biogenetic route, is presented. Condensation of a protected amino acid with a bromoacrylamide and subsequent addition of malonaldehyde sodium salt and methyl iodide afforded, after ster

Full text of DOI:10.1002/ejoc.200600293

FULL PAPER
DOI: 10.1002/ejoc.200600293
A Biogenetically Based Strategy Towards the Polycyclic Core Skeleton of
Sarain A
[
a]
[a]
[a]
[a]
Cheng Sheng Ge, Stéphane Hourcade, Antoine Ferdenzi, Angèle Chiaroni,
[a]
[a]
[a]
Stéphane Mons,* Bernard Delpech, and Christian Marazano*
Keywords: Biomimetic synthesis / Alkaloids / Multicomponent synthesis / Reduction / Pyrrolidines / Sarain A
An approach to the polycyclic core of sarain A, based upon
a proposed biogenetic route, is presented. Condensation of a
protected amino acid with a bromoacrylamide and subse-
quent addition of malonaldehyde sodium salt and methyl io-
dide afforded, after stereoselective hydrogenation, a highly
functionalized diazabicyclo[4.3.0]nonane system 16. This in-
termediate possesses the stereochemistry required for the
synthesis of the core skeleton of sarain A, a model of which
(31) was obtained by an allylsilane strategy previously de-
scribed by Weinreb and co-workers.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2006)
Introduction
possibly be the result of the condensation, in a multistep
reductive process, of four intermediates: a sphingolipid 2
and an amino aldehyde 3 with two malonaldehyde units.
One advantage of such a hypothesis is that it closely links
Since its discovery,[1] sarain A (1), a complex polycyclic
alkaloid extracted from the sponge Reniera sarai, has be-
come one of the most challenging targets in natural prod-
[8]
the origin of sarain A to that, also proposed by us, of
ucts synthesis (Scheme 1),[
2–5]
as a result of which the first
manzamine A, an alkaloid extracted from sponges of the
same order (Haplosclerida). Such considerations are also
consistent with recent observations suggesting that these al-
kaloids are in fact produced by sponge-associated bacte-
[6]
total synthesis has very recently been achieved. We have
[
7]
suggested that the biogenetic origin of sarain A (1) could
[
9]
ria.
While these proposals are still speculative, as no biosyn-
thetic studies are yet available, they offer valuable bases for
the design of new synthetic strategies for sarain A. From
the point of view of the proposed biogenetic pathway de-
picted in Scheme 1, there are many different ways to as-
semble the hypothetical starting units. In addition, alde-
hydes and malonaldehyde are unstable compounds, so the
question of whether to use more stable synthetic equivalents
arises. We chose to study the route summarized in
Scheme 2, as a result of which a first study of this model
Scheme 1. Proposed biogenetic pathway.
[
a] Institut de Chimie des Substances Naturelles, CNRS,
Avenue de la Terrasse, 91198 Gif-sur-Yvette, France
Fax: +33-169077247
E-mail: marazano@icsn.cnrs-gif.fr
mons@icsn.cnrs-gif.fr
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Scheme 2. General strategy.
4106
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 4106–4114

Biogenetic Strategy Towards the Polycyclic Core Skeleton of Sarain A
FULL PAPER
based on Knoevenagel chemistry was recently published.[7] the highly functionalized diazabicyclo[4.3.0]nonane 14 with
We now report a new and efficient approach to highly func- complete cis ring junction stereoselectively in an overall
tionalized diazabicyclo[4.3.0]nonane analogues of 6, to- yield of 44%. This sequence, which does not required puri-
gether with some observations concerning their regio- and fication of any intermediate, is of particular interest since it
stereoselective reduction, which is strongly influenced by consists of the condensation of four different molecules
proximity effects, and lastly the synthesis of an analogue of with formation of two C–C and three C–N bonds in a very
the core skeleton of sarain A.
short process. The stereospecificity of the ring closure reac-
tion can be explained by the nearly planar shape of the five
sp atoms in 13. The enaminal anion can only add to the
2
Results and Discussion
double bond from the same side as the nitrogen atom.
The direct functionalization of vinylogous amide 14 by
organometallic species proved to be difficult, but possible.
The best alternative was the reduction of enaminal 14 in
order to introduce the third stereocenter (Scheme 4) by
using sodium borohydride in acetic acid which fully reduced
the enaminal function but stereospecifically delivered the
undesired alcohol isomer 15. We had shown in our previous
As shown in Scheme 2, four molecules (4, 5, and two
malonaldehyde units) need to be condensed to afford inter-
mediates such as 6. Selective reduction and side-chain
elongation with aldehyde 7 should then afford iminium salt
8, Mannich cyclization of which would be likely to produce
sarain A core skeleton 9.
We first tried to mimic the first sequence of the projected
synthesis leading to intermediates 6, and accordingly to
highly functionalized diazabicyclo[4.3.0]nonane systems
[7]
paper that such a configuration is the thermodynamic
one, as was further confirmed when alcohol 16 was oxidized
to the corresponding aldehyde. Under enolization condi-
tions, this aldehyde was completely epimerized at C-3 (re-
sults not shown). Hydrogenation of 14 in the presence of
platinum oxide resulted in the formation of the desired endo
isomer 16, though with poor selectivity (1.6:1 mixture with
(Scheme 3). -Alanine imino derivative 10 was deproton-
ated and treated with bromoacrylamide derivative 11 to
provide salt 12 in 70% yield after acidic imine hydrolysis.
This crude hydrochloride was then treated with malonal-
dehyde sodium salt, affording enaminal 13. This intermedi-
ate was not isolated, but was directly treated with sodium
hydride, followed by methyl iodide, ultimately to provide
[
10]
15), but isomer 16 could eventually be obtained as a 4.7:1
mixture with 15 through the use of a catalytic amount of tin
[11]
dichloride as catalyst activator. The two diastereoisomers
proved to be difficult to separate and were used as a mixture
in the next steps.
Tosylate 17 could be crystallized and subjected to X-ray
[
12]
analysis (Figure 1), which established the stereochemistry
at C-3 in tosylate 17 and, as a consequence, in its precursor
alcohol 15.
A brief investigation of borohydride reduction of the
imide moiety revealed some interesting features. Imide 15
was reduced to give 18 as a single diastereoisomer, but
NOESY experiments and molecular modeling did not allow
conclusive determination of the relative configuration at the
hemiaminal center. Selective reduction at the more hindered
sites of imides by NaBH is already known, and ring open-
4
Scheme 3. Short route to a diazabicyclo[4.3.0]nonane system.
ing can be prevented by continuously lowering the pH of
Scheme 4. Reduction studies on intermediate 14.
Eur. J. Org. Chem. 2006, 4106–4114
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4107

S. Mons, C. Marazano et al.
FULL PAPER
Figure 1. ORTEP plot of tosylate 17.[13] Ellipsoids are drawn at 30% probability level.
the reaction mixture.[14] In contrast, imide 16 gave the diol
19 as the result of a double reduction process. It is very
likely that the hydroxy moiety in 16 allowed intramolecular
delivery of the reagent and enforced the double reduction.
X-ray analysis of diol 19 unambiguously confirmed the
structure found by NMR spectroscopy (Figure 2).
Reduction of the crude mixture of imides 15 and 16
(
(
1:4.7 ratio, Scheme 5) with DIBAH gave two products 20
from 16, 72% yield) and 21 (from 15, 12% yield). Since no
other isomer could be detected, this reduction can be re-
garded as highly, if not totally, regioselective.
The sequence was followed by transformation into cyclic
N-acylenamine 22, allowing reduction of the double bond
by hydrogenation in the presence of Raney nickel to give
alcohol 23 (Scheme 5). Removal of the benzyl group with
sodium in ammonia delivered alcohol 24 in 71% overall
yield from aminal 20. In order to elongate the side chain as
in model 8 (see Scheme 2), alcohol 24 was activated as a
tosylate. However, attempts to introduce aldehyde enolates
(see 7) or their equivalents (imine or malonate anions) in-
variably failed at this point. This can be interpreted in terms
of S 2 substitution of the tosylate group being disfavored
N
because of the bulkiness and concave shape of the bicyclic
core.
In contrast, introduction of a small nucleophile such as
cyanide anion could be performed at 100 °C, affording ni-
trile 26 in good yield (Scheme 6), DIBAH reduction and
hydrolysis then affording aldehyde 27. This aldehyde al-
lowed us to validate our strategy towards the sarain A core
skeleton quickly with the aid of an allylsilane methodology
previously developed by Weinreb and co-workers.[ Allylsi-
2]
Figure 2. ORTEP plot of alcohol 19. Ellipsoids are drawn at 30%
lane 28 [mixture of (E) and (Z) isomers] was obtained from probability level.
4108
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 4106–4114

Biogenetic Strategy Towards the Polycyclic Core Skeleton of Sarain A
FULL PAPER
Scheme 5. Selective imide 16 reduction.
aldehyde 27 in 70% yield, tosylation gave a mixture of iso- ates also gave encouraging results. Firstly, a hydrogenation
mers 29, which was reduced with DIBAH to afford the tar- procedure allowed the desired endo isomer to be obtained,
geted iminium ion precursors 30, and cyclization with ferric while the regioselectivity can be also controlled for imide
chloride finally delivered the sarain A core skeleton ana- reduction. These results open the way to further develop-
logue 31.
ments of this strategy, which are underway in our labora-
tory.
Experimental Section
General Remarks: NMR spectra were recorded with the following
Bruker spectrometers: AC 250 (250 MHz), AC 300 (300 MHz), Av-
ance 300 (300 MHz), DPX 400 (400 MHz), Avance 500 (500 MHz).
FT-IR spectra were recorded as films on NaCl or in a diamond
(
SensIR DurasampIR II) cell with a Perkin–Elmer Spectrum BX
FT-IR spectrophotometer. Mass spectra were recorded by electro-
spray ionization with a Micromass LCT instrument (ESI-TOF).
Melting points were measured with a Büchi Melting Point B-540
apparatus.
N-Benzylidene-
6.3 mmol) and benzaldehyde (5.7 mL, 56.07 mmol) were added to
a stirred solution of -alanine ethyl ester hydrochloride (8.6 g,
6 mmol) in MeCN (100 mL). The reaction mixture was stirred
overnight and filtered through Celite, and the solvents were evapo-
rated. The residue was diluted with Et O and filtered through Ce-
lite again, and the organic layer was concentrated to give imine 10
L-alanine Ethyl Ester (10): TEA (7.8 mL,
5
5
2
1
(
11.5 g, 100% yield). H NMR (300 MHz, CDCl
3
): δ = 1.27 (t, J
=
7.1 Hz, 3 H,), 1.52 (d, J = 6.9 Hz, 3 H), 4.09–4.23 (m, 3 H), 7.35
1
3
Scheme 6. Synthesis of sarain A model 31 by an allylsilane strat-
(m, 5 H), 8.30 (s, 1 H) ppm. C NMR (75 MHz, CDCl
3
): δ = 14.0,
[
2]
egy.
19.2, 60.8, 67.8, 128.3, 128.4, 130.8, 135.6, 162.6, 172.3 ppm. IR
(
7
neat): ν˜ = 2981, 1730, 1641, 1449, 1369, 1184, 1124, 1046, 1021,
52 cm . HRMS (ESI , MeOH): m/z = 206.1186 [M+H] (calcd.
–
1
+
+
for C12H16NO
2
206.1181).
Conclusions
(
Z)-N-Benzyl-3-bromoacrylamide (11): Propiolic acid (9.0 g,
1
28.5 mmol) was added to a solution of LiBr (11.16 g, 128.5 mmol)
We have shown that it is possible to obtain the sarain A
core skeleton analogue 31 by a route based on the proposed
biogenetic pathway depicted in Scheme 2. The reported
strategy has some advantages over our previous approach.
Of particular interest is the short, stereoselective route to
highly functionalized diazabicyclo[4.3.0]nonane systems
[15]
in MeCN (125 mL). The mixture was heated at reflux overnight,
filtered through Celite, and washed with Et O (3×50 mL) to give
lithium (Z)-3-bromoacrylate (17.73 g, 88% yield). Isobutyl chloro-
formate (6.8 mL, 51.7 mmol) was added at 0 °C to a stirred suspen-
sion of this acrylate salt (8.08 g, 51.5 mmol) in THF (250 mL). The
reaction mixture was allowed to warm to room temp. and stirred
2
such as 14, achieved by successive condensations of four until it turned clear and the solution was cooled to 0 °C, after
different molecules. Reduction studies of these intermedi- which benzylamine (5.9 mL, 54.0 mmol) was added and the mix-
Eur. J. Org. Chem. 2006, 4106–4114
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4109

S. Mons, C. Marazano et al.
FULL PAPER
ture was stirred overnight. AcOEt (500 mL) was added, the organic
was then cooled to 0 °C. H
2
O (10 mL) was added to the reaction
phase was washed with H
and the solvents were evaporated. The resulting oil was dissolved
in a mixture of AcOEt and Et O (1:1) and crystallized at –20 °C
to give (Z)-N-benzyl-3-bromoacrylamide (6.18 g, 50% yield): m.p. vents were evaporated. Column chromatography of the residue on
2
O (2×500 mL) and dried with MgSO
4
,
mixture, which was neutralized with a 50% aqueous NaOH solu-
tion. The mixture was extracted with CH Cl (3×200 mL), the
combined organic extracts were dried with MgSO , and the sol-
2
2
2
4
1
7
5–76 °C. H NMR (300 MHz, CDCl
H), 6.54 (d, J = 8.1 Hz, 1 H), 6.62 (d, J = 8.1 Hz, 1 H), 7.26 (m, 5
H) ppm. 13C NMR (75 MHz, CDCl
): δ = 43.5, 114.1, 127.5, 127.8,
28.7, 128.3, 137.6, 163.7 ppm. IR (neat): ν˜ = 3276, 1606, 1537, H), 2.32 (dd, J = 9.9, 8.7 Hz, 1 H), 2.42 (s, 3 H), 2.65 (dd, J = 9.9,
3
): δ = 4.44 (d, J = 6.0 Hz, 2
silica gel (CH
(148 mg, 84% yield): H NMR (300 MHz, CDCl
H), 1.90 (m, 1 H), 1.95 (s, 1 H), 2.09 (ddd, J = 7.6, 5.0, 2.9 Hz, 1
2 2
Cl /MeOH, 97:3) gave alcohol 15 as a colorless oil
1
3
): δ = 1.36 (s, 3
3
1
1
=
–1
+
494, 1450, 1351, 1247, 1184, 1028 cm . MS (ESI , MeOH): m/z
262.264 [M+Na] . C10H10BrNO (240.10): calcd. C 50.02, H
.20, N 5.83; found C 50.09, H 4.32, N 5.54.
3.3 Hz, 1 H), 2.72 (dd, J = 16.7, 5.0 Hz, 1 H), 2.86 (dd, J = 16.7,
2.9 Hz, 1 H), 3.60 (dd, J = 14.9, 0.7 Hz, 1 H), 3.64 (dd, J = 14.9,
0.7 Hz, 1 H), 4.88 (d, J = 13.8 Hz, 1 H), 4.93 (d, J = 13.8 Hz, 1
+
4
1
3
H), 7.20–7.40 (m, 5 H) ppm. C NMR (75 MHz,CDCl
3
): δ = 19.9,
4.2, 35.3, 42.5, 43.0, 43.5, 55.5, 65.7, 66.2, 127.4, 128.4, 128.8,
37.5, 171.4, 173.9 ppm. IR (neat): ν˜ = 3439, 2968, 2927, 2801,
5
(
-Amino-1-benzyl-5-methyl-3,4-dehydroglutarimide Hydrochloride
12·HCl): Ethyl N-benzylidene--alanate 10 (6.5 g, 31.6 mmol) in
THF (80 mL) was added at –78 °C under argon to a stirred solu-
tion of LDA (63 mmol) in dry THF (160 mL). The reaction mix-
ture was maintained at –78 °C for an additional 20 min, after which
solution of (Z)-N-benzyl-3-bromoacrylamide (11) (7.6 g,
1.6 mmol) in THF (80 mL) was added. The mixture was allowed
to warm to room temp. and stirred for 14 h, after which H
3
1
1
–1
+
722, 1672, 1447, 1378, 1353, 1335, 1162 cm . HRMS (ESI ,
303.1696 [M+H]+ (calcd. for
17 23 2 3
C H N O
MeOH): m/z
=
303.1709).
a
3
Alcohol 16: A pressure reaction vessel was charged with enaminal
14 (200 mg, 0.67 mmol), THF (25 mL), PtO (33 mg), and SnCl
2
O
2
2
(
200 mL) was added. The reaction mixture was then diluted with
(2.3 mg). The reaction mixture was then hydrogenated in a Parr
apparatus under pressure (4 bar) overnight. After filtration through
Celite, the solvent was evaporated. Column chromatography of the
residue on deactivated silica gel (eluent AcOEt/acetone, 4:1; silica
gel was shaken in a mixture of AcOEt/acetone/TEA, 20:5:1.25 be-
fore use) afforded a mixture of alcohols 15 and 16 (168 mg, 83%
yield, 4.7:1 diastereoselectivity). An analytically pure sample of
Et
H
2
O (200 mL) and the organic layer was separated, washed with
O (5×200 mL), and dried with MgSO . The organic layer was
concentrated to dryness to give the imine, which was dissolved in
Et O (220 mL). An aqueous HCl solution (0.2 , 162 mL) was
2
4
2
added and the resulting mixture was vigorously stirred for 3 h. The
organic phase was decanted and the aqueous phase was washed
with Et
9% yield). H NMR (300 MHz, CDCl
d, J = 14.7 Hz, 1 H), 5.07 (d, J = 14.7 Hz, 1 H), 6.47 (d, J =
2
O (5×100 mL) and concentrated to give 12·HCl (5.79 g,
alcohol 16 was obtained, but the mixture was usually used as it
1
1
6
(
3
): δ = 1.72 (s, 3 H), 5.01 was. H NMR (300 MHz, CDCl
3
): δ = 1.20 (s, 3 H), 2.33 (s, 3 H),
2.33–2.82 (m, 6 H), 3.46 (d, J = 6.6 Hz, 2 H), 4.83 (s, 2 H), 7.13–
1
3
13
1
0.2 Hz, 1 H), 6.94 (d, J = 10.2 Hz, 1 H), 7.31 (m, 5 H) ppm.
): δ = 26.1, 44.2, 56.7, 124.7, 128.6, 129.3,
29.5, 137.6, 142.3, 163.9, 172.3 ppm. IR (neat): ν˜ = 3385, 2848,
C
7.26 (m, 5 H) ppm. C NMR (75 MHz, CDCl
3
): δ = 17.0, 31.1,
NMR (75 MHz, CDCl
3
35.1, 40.5, 41.6, 43.1, 55.2, 61.6, 65.4, 127.3, 128.3, 128.5, 137.1,
–
1
1
1
2
171.9, 173.6 ppm. IR (neat): ν˜ = 3368, 1671, 1603, 1383, 1094 cm .
–1
+
+
+
728, 1687, 1643, 1196, 1145 cm . HRMS (ESI , MeOH): m/z =
HRMS (ESI , MeOH): m/z = 325.1514 [M+Na] (calcd. for
NaO 325.1528).
–
+
31.1128 [M–Cl ] (calcd. for C13
H
15
N
2
O
2
231.1134).
C
17
H
22
N
2
3
Enaminal 14: Amine 12·HCl (1.85 g, 6.93 mmol) was added under Tosylate 17: TsCl (156 mg, 0.82 mmol), DMAP (10 mg,
argon to a stirred solution of malonaldehyde sodium salt[ (1 g,
16]
0.08 mmol), and freshly distilled TEA (0.114 mL, 0.82 mmol) were
8
.93 mmol) in dry CH Cl (30 mL). The reaction mixture was
2
2
added at 0 °C to a solution of alcohol 15 (124 mg, 0.41 mmol) in
stirred for 3 h and filtered through Celite, and the solvents were
evaporated. The resulting orange solid (intermediate vinylogous
amide 13, 2.03 g, 90% purity by H NMR) was dissolved in dry
dry CH
overnight, a saturated solution of aqueous NaHCO
the system was extracted three times with CH Cl , the combined
organic extracts were dried with MgSO , and the solvents were
2
Cl
2
(4 mL). The reaction mixture was stirred at room temp.
3
was added,
1
2
2
THF (15 mL), and sodium hydride (60% in mineral oil, 332 mg,
4
8.32 mmol) was added. After 3 h, the mixture was cooled to 0 °C,
evaporated. Column chromatography of the residue on silica gel
after which methyl iodide (1.3 mL, 20 mmol) was added. The reac-
tion mixture was stirred at 0 °C for 2 h and then at room temp. for
(heptanes/AcOEt, 1:1) afforded tosylate 13 as a white solid
(128 mg, 68% yield); m.p. 104–105 °C. H NMR (300 MHz,
1
1
h, the solvent was evaporated, and the residue was diluted with
CH Cl (50 mL), the mixture filtered through Celite, and the fil-
trate concentrated to afford an orange solid (1.49 g). Column
chromatography of the residue on silica gel (AcOEt/CH Cl , 40:60)
3
CDCl ): δ = 1.30 (s, 3 H), 1.89–2.04 (m, 2 H), 2.25 (dd, J = 10.2,
2
2
8.2 Hz, 1 H), 2.37 (s, 3 H), 2.45 (s, 3 H), 2.50 (dd, J = 10.2, 3.2 Hz,
1 H), 2.67 (dd, J = 16.7, 4.8 Hz, 1 H), 2.78 (dd, J = 16.7, 2.7 Hz,
1 H), 3.93 (dd, J = 9.6, 7.2 Hz, 1 H), 3.99 (dd, J = 9.4, 6.6 Hz, 1
H), 4.84 (d, J = 13.8 Hz, 1 H), 4.91 (d, J = 13.8 Hz, 1 H), 7.25–
2
2
1
gave enaminal 14 (1.31 g, 63% yield) as a colorless oil. H NMR
300 MHz, CDCl
H), 2.79 (s, 3 H), 3.34 (dd, J = 16.0, 4.6 Hz, 1 H), 3.47 (t, J =
.3 Hz, 1 H), 4.80 (d, J = 14.0 Hz, 1 H), 5.04 (d, J = 14.0 Hz, 1
(
3
): δ = 1.55 (s, 3 H), 2.72 (dd, J = 16.0, 6.1 Hz, 1 7.40 (m, 5 H), 7.41 (d, J = 8.2 Hz, 2 H), 7.75 (d, J = 8.2 Hz, 2 H)
1
3
ppm. C NMR (75 MHz, CDCl
43.0, 43.8, 54.5, 65.9, 72.2, 127.5, 127.9, 128.4, 128.8, 130.0, 132.7,
H), 6.82 (s, 1 H), 7.30 (m, 5 H), 9.24 (s, 1 H) ppm. C NMR 137.4, 145.1, 170.7, 173.6 ppm. IR (neat): ν˜ = 2928, 2803, 1724,
3
): δ = 19.9, 27.1, 33.8, 35.1, 39.7,
5
13
–
1
+
(
1
(
75 MHz, CDCl
27.6, 128.4, 128.6, 136.7, 157.0, 168.8, 170.3, 181.8 ppm. IR
neat): ν˜ = 2923, 2777, 1729, 1678, 1626, 1578, 1438, 1408, 1356,
3
): δ = 20.6, 32.3, 32.7, 41.5, 43.7, 70.4, 115.0,
1673, 1356, 1176, 963, 814, 701 cm . HRMS (ESI , MeOH): m/z
= 457.1755 [M+H]+ (calcd. for C24
S 457.1752).
29 2 5
H N O
–1
+
+
4
Derivative 18: NaBH (200 mg, 5.23 mmol) was added at 0 °C to a
1
266, 1162 cm . HRMS (ESI , MeOH): m/z = 299.1429 [M+H]
solution of alcohol 15 (150 mg, 0.49 mmol) in absolute EtOH
15 mL). The resulting mixture was stirred at 0 °C for 5 h with ad-
(calcd. for C17
19 2 3
H N O
299.1395).
(
Alcohol 15: NaBH
to enaminal 14 (175 mg, 0.59 mmol) in glacial acetic acid (12 mL).
The reaction mixture was stirred at room temp. for 15 min, and
4
(480 mg, 12.7 mmol) was slowly added at 0 °C
dition of 3 drops of a solution of HCl gas in EtOH (2 ) every
15 min. The mixture was then acidified until pH = 3, stirred for an
additional 20 h, and made alkaline with 10% KOH in EtOH solu-
4110
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 4106–4114

Biogenetic Strategy Towards the Polycyclic Core Skeleton of Sarain A
tion until pH = 9, the organic solvent was evaporated, and the
FULL PAPER
Derivative 21: H NMR (300 MHz, CDCl ): δ = 1.24 (s, 3 H), 2.02–
3
1
reaction mixture was diluted with saturated aqueous NaHCO
3
2.07 (m, 1 H), 2.21 (dd, J = 8.1, 5.7 Hz, 1 H), 2.61 (dd, J = 5.7,
9.9 Hz, 1 H), 2.65 (s, 3 H), 2.86 (t, J = 9.3 Hz, 2 H), 3.60 (d, J =
5.7 Hz, 2 H), 4.53 (d, J = 15 Hz, 1 H), 4.61 (d, J = 15 Hz, 1 H),
5.09 (dd, J = 7.8, 6 Hz, 1 H), 5.84 (d, J = 8.1 Hz, 1 H), 7.12–7.27
(
10 mL). After three extractions with CH
2
Cl
2
, the combined or-
ganic layers were dried with MgSO
4
and the solvents were evapo-
rated. Column chromatography of the residue on alumina (eluent
CH Cl /TEA, 98:2) gave amide 18 as a colorless oil (107 mg, 71%
yield). H NMR (300 MHz, CDCl
8.9 Hz, 1 H), 2.00 (m, 1 H), 2.15 (s, 3 H), 2.47 (dd, J = 10.5,
1
3
2
2
1
(m, 5 H) ppm. C NMR (75 MHz, CDCl
): δ = 0.70 (s, 3 H), 1.83 (q, J 48.3, 49.1, 55.0, 65.0, 65.6, 108.0, 126.8, 127.4, 127.4, 128.6, 137.2,
173.0 ppm. IR (neat): ν˜ = 3361, 2923, 1650, 1495, 1453, 1384, 1253,
3
): δ = 19.7, 35.8, 46.2,
3
=
–
1
+
9
1
1
.4 Hz, 1 H), 2.60 (d, J = 9.5 Hz, 2 H), 3.16 (dd, J = 9.4, 7.0 Hz,
1218, 1172, 1028, 725, 697 cm . HRMS (ESI , MeOH): m/z =
H), 3.58 (d, J = 5.9 Hz, 2 H), 3.87 (d, J = 14.5 Hz, 1 H), 4.49 (s,
287.1767 [M+H]+ (calcd. for C17
23 2 2
H N O
287.1760).
1
3
H), 5.00 (d, J = 14.5 Hz, 1 H), 7.31 (m, 5 H) ppm. C NMR
): δ = 17.7, 34.4, 37.4, 45.8, 47.8, 50.2, 57.1, 63.5,
5.3, 84.1, 127.4, 128.4, 128.6, 137.5, 171.8 ppm. IR (neat): ν˜ =
Enaminone 22: BF ·Et O (0.28 mL, 2.24 mmol) was added to a
3
2
(75 MHz, CDCl
3
solution of tricyclic derivative 20 (128 mg, 0.44 mmol) in dry
CH Cl (5 mL). The reaction mixture was heated at reflux over-
night and was then diluted with 1  aqueous NaOH solution and
extracted three times with CHCl . The combined organic extracts
were dried with MgSO and the solvents evaporated. Column
chromatography of the residue on silica gel (eluent AcOEt/acetone,
:1) gave unsaturated enaminone 22 as a colorless oil (105 mg, 82%
6
3
–1
2
2
385, 2923, 1651, 1482, 1454, 1385, 1176, 1044, 705 cm . HRMS
+
+
(ESI , MeOH): m/z = 305.1863 [M+H] (calcd. for C17
25 2 3
H N O
3
305.1865).
4
Diol 19: NaBH
of alcohol 16 (74 mg, 0.24 mmol) in absolute EtOH (7 mL). After
h, H O (4.5 mL) was added and the reaction mixture was stirred
4
(93 mg, 2.4 mmol) was added at 0 °C to a solution
2
1
yield). H NMR (300 MHz, CDCl
3
): δ = 1.33 (s, 3 H), 2.41–2.45
1
2
(
m, 1 H), 2.68 (s, 3 H), 2.74–2.77 (m, 1 H), 2.78–2.92 (m, 3 H),
at 55 °C for 1 h. Acetic acid was added dropwise to adjust the pH
to ca. 4, and then the aqueous layer was made alkaline with a
3
4
.55 (dd, J = 5.1, 10.2 Hz, 1 H), 3.64 (dd, J = 5.1, 10.2 Hz, 1 H),
.55 (d, J = 15.3 Hz, 1 H), 4.78 (d, J = 15.3 Hz, 1 H), 5.00 (dd, J
saturated aqueous NaHCO
3
solution. After three extractions with
and
=
5
8.7, 5.1 Hz, 1 H), 6.03 (dd, J = 8.7, 1.5 Hz, 1 H), 7.24–7.36 (m,
H) ppm. C NMR (75 MHz, CDCl ): δ = 19.7, 36.3, 41.8, 48.1,
3
CH Cl , the combined organic layers were dried with MgSO
2
2
4
1
3
the solvents were evaporated. Column chromatography of the resi-
49.3, 56.4, 63.3, 64.3, 103.6, 127.5, 127.6, 128.2, 128.6, 137.0,
due on alumina (AcOEt/acetone, 1:1) gave diol 19 as a white solid
1
172.5 ppm. IR (neat): ν˜ = 3383, 2928, 1658, 1650, 1644, 1453, 1392,
(
38 mg, 50% yield); m.p. 155–156 °C. H NMR (300 MHz,
–
1
+
1
[
028, 728, 697 cm . HRMS (ESI , MeOH): m/z = 287.1743
M+H]+ (calcd. for C17
287.1760). C17 (286.37):
calcd. C 71.30, H 7.74, N 9.78; found C 70.83, H 7.95, N 9.54.
CDCl
3
): δ = 1.20 (s, 3 H), 1.41–1.52 (m, 1 H), 1.70–1.81 (m, 1 H),
H
23
N
2
O
2
22 2 2
H N O
2
2
3
8
3
1
.03–2.25 (m, 2 H), 2.17 (s, 3 H), 2.58 (dd, J = 5.7, 9.6 Hz, 1 H),
.95 (d, J = 9.0 Hz, 1 H), 3.41 (dd, J = 8.4, 11.4 Hz, 1 H), 3.55–
.69 (m, 3 H), 4.33 (qd, J = 12.3, 5.7 Hz, 2 H), 7.18–7.30 (m, 5 H),
Amide 23:
A solution of unsaturated alcohol 22 (100 mg,
13
.13 (m, 1 H) ppm. C NMR (75 MHz, CDCl
3
): δ = 19.0, 29.7,
0
.34 mmol) in MeOH (20 mL) was hydrogenated at atmospheric
pressure in the presence of Raney nickel in H O (4 mL). After stir-
ring overnight, the reaction mixture was filtered through Celite and
the solvent evaporated. The residue was diluted with CHCl , dried
with MgSO , and concentrated to give amide 2 as a colorless oil
95 mg, 95% yield). H NMR (300 MHz, CDCl
H), 1.72–1.79 (m, 1 H), 1.89–1.98 (m, 1 H), 2.25–2.40 (m, 2 H),
5.7, 42.9, 43.4, 52.2, 56.5, 61.9, 62.6, 69.0, 127.5, 127.9, 128.8,
2
–
1
38.5, 175.1 ppm. IR (neat): ν˜ = 3327, 2934, 1651, 1512, 1030 cm .
+
+
MS (ESI , MeOH): m/z = 457 [M+H] . C17
26 2 3
H N O (306.40):
3
calcd. C 66.64, H 8.55, N 9.40; found C 66.48, H 8.54, N 9.10.
4
1
(
3
): δ = 1.28 (s, 3
2 2
Tricyclic Derivative 20: DIBAH (1  in CH Cl , 10 mL, 10 mmol)
was added dropwise at –78 °C to a solution of alcohol 16 (crude
mixture with isomeric 15 in a 4.7:1 ratio, 1.54 g, 5 mmol) in dry
2
9
1
7
3
1
1
.62 (s, 3 H), 2.79 (dd, J = 7.5, 9.9 Hz, 1 H), 2.97 (dd, J = 3.6,
.9 Hz, 1 H), 3.23 (dd, J = 3.3, 8.1 Hz, 2 H), 3.64 (dd, J = 4.2,
2 2
CH Cl (15 mL). The mixture was stirred for 1.5 h and trifluoro-
0.5 Hz, 1 H), 3.74 (dd, J = 4.8, 10.5 Hz, 1 H), 4.59 (s, 2 H), 7.23–
acetic acid (2 mL) was then added. The reaction mixture was al-
lowed to warm to room temp. and stirred for an additional 0.5 h.
An aqueous NaOH solution (5 ) was added to the reaction mix-
ture to ensure an alkaline pH of 8–9, the reaction mixture was
13
3
.35 (m, 5 H) ppm. C NMR (75 MHz, CDCl ): δ = 18.4, 23.0,
6.5, 39.5, 46.1, 47.1, 50.3, 57.1, 62.9, 65.0, 127.3, 127.8, 128.6,
37.2, 172.9 ppm. IR (neat): ν˜ = 3385, 2925, 2847, 1620, 1491,
–1
+
451, 1354, 1169, 1029, 909, 726, 698 cm . HRMS (ESI , MeOH):
diluted with H
bined organic extracts were dried with MgSO
evaporated. Column chromatography of the residue on silica gel
eluent AcOEt/acetone, 4:1) gave the tricyclic derivative 20 as a
white solid (1.02 g, 72% yield) and the unsaturated alcohol 21
from 15) as a colorless liquid (0.18 g, 12% yield).
2
O and extracted three times with CH
2
Cl
2
, the com-
+
m/z = 289.1918 [M+H] (calcd. for C17
25 2 2
H N O 289.1916).
4
and the solvents
Amide 24: A flask containing tert-butyl alcohol (0.58 mL) and
(
THF (7 mL) was cooled to –78 °C and NH gas was condensed
3
into the mixture (approximately 12 mL). Sodium metal (101 mg,
4.4 mmol) was added in small portions, producing a deep blue solu-
tion, and benzylamide 23 (87 mg, 0.3 mmol), dissolved in THF
(2 mL), was added. The mixture was stirred for 6 min, the dry ice
(
Tricyclic Derivative 20: M.p. 82–84 °C. 1H NMR (300 MHz,
CDCl
J = 13.8, 3.0 Hz, 1 H), 2.34 (m, 1 H), 2.49 (m, 1 H), 2.57 (s, 3 H),
3
): δ = 1.39 (s, 3 H), 1.75 (dt, J = 13.8, 2.4 Hz, 1 H), 2.15 (dt, bath was removed, and the reaction was quenched with saturated
aqueous NH Cl solution (5 mL). After the NH had been allowed
to evaporate, the mixture was extracted with CHCl , and the or-
ganic phase was dried with MgSO and concentrated. Column
Cl /MeOH
(2:1) as the eluent gave amide 24 as a white solid (54 mg, 91%
4
3
2
3
1
.95 (dd, J = 9.0, 6.3 Hz, 1 H), 3.63 (dd, J = 7, 12.0, 6.3 Hz, 1 H),
3
.85 (dd, J = 12.0, 6.0 Hz, 1 H), 3.90 (d, J = 15 Hz, 1 H), 4.89 (m,
4
H), 5.15 (d, J = 15 Hz, 1 H), 7.25–7.34 (m, 5 H) ppm. 1 C NMR chromatography of the residue on silica gel with CH
3
2
2
(
75 MHz, CDCl
3
): δ = 20.1, 25.0, 34.5, 35.6, 40.6, 47.9, 57.2, 61.4,
1
6
2
7
4.8, 79.7, 127.2, 128.2, 128.4, 137.5, 172.6 ppm. IR (neat): ν˜ =
790, 2360, 1647, 1493, 1448, 1411, 1362, 1328, 1211, 1175, 1041,
3
yield). H NMR (300 MHz, CDCl ): δ = 1.24 (s, 3 H), 1.81–1.85
(m, 1 H), 1.92 (dq, J = 4.5, 12.0 Hz, 1 H), 2.24–2.29 (m, 1 H),
2.37–2.40 (m, 1 H), 2.53 (s, 3 H), 2.74 (t, J = 9.0 Hz, 1 H), 2.89
–
1
+
+
25 cm . HRMS (ESI , MeOH): m/z = 287.1777 [M+H] (calcd.
287.1760). C17
H 7.74, N, 9.78; found C 70.88, H 7.79, N 9.52.
for C17
H
23
N
2
O
2
H
22
N
2
O
2
(286.37): calcd. C 71.30, (dd, J = 2.5, 9.5 Hz, 1 H), 3.22–3.28 (m, 1 H), 3.33–3.37 (m, 1 H),
3.68 (dd, J = 4.0, 10.0 Hz, 1 H), 3.75 (dd, J = 4.0, 10.0 Hz, 1 H),
Eur. J. Org. Chem. 2006, 4106–4114
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4111

S. Mons, C. Marazano et al.
FULL PAPER
6
3
3
.02 (br.s, 1 H) ppm. 13C NMR (75 MHz, CDCl
6.3, 39.7, 40.9, 47.0, 56.8, 63.1, 64.1, 175.5 ppm. IR (neat): ν˜ =
3
): δ = 17.8, 23.5,
THF (2.5 mL). The mixture was allowed to warm to room temp.,
stirred for 1 h, and cooled to 0 °C. (Iodomethyl)trimethylsilane
(0.085 mL, 0.57 mmol) was added, and the solution was allowed to
warm to room temp. After 2 h, the reaction mixture was treated
–1
299, 1643, 1493, 1461, 1346, 1178, 1098, 1057, 1029 cm . HRMS
+
+
(ESI , MeOH): m/z = 199.1440 [M+H] (calcd. for C10
19 2 2
H N O
199.1447).
2
at –78 °C with methyllithium (0.37 mL, 1.6  in Et O). The mixture
immediately turned red. After the mixture had been stirred at room
temp. for 1 h, a homogeneous red solution had formed, and alde-
hyde 27 (38 mg, 0.18 mmol) in THF (1.8 mL) was added to this
at –78 °C. The mixture was allowed to warm to room temp. and
stirred overnight, and was then diluted with a saturated aqueous
ammonium chloride solution and extracted three times with
Tosylate 25: TsCl (136 mg, 0.70 mmol), DMAP (8.7 mg,
.07 mmol), and TEA (0.075 mL, 0.54 mmol) were added at 0 °C
to a solution of alcohol 19 (70 mg, 0.35 mmol) in dry CH Cl
4 mL). The reaction mixture was allowed to warm to room tem-
perature and stirred for 2 h and was then diluted with a saturated
aqueous NaHCO solution and extracted three times with CH Cl
The combined organic extracts were dried with MgSO and the
0
2
2
(
3
2
2
.
CH
and the solvents evaporated. Column chromatography of the resi-
due on silica gel with CH Cl /MeOH (50:1) as eluent gave a mix-
ture of isomers 28 as a colorless oil (37 mg, 70% yield). H NMR
300 MHz, CDCl ): δ = 0.08 (s, 9 H), 1.21 (s, 3 H), 1.48 (d, J =
.8 Hz, 2 H), 1.62–1.97 (m, 3 H), 2.01–2.09 (m, 3 H), 2.45 (s, 3 H),
.68 (t, J = 9.6 Hz, 1 H), 2.82 (dd, J = 9.3, 7.8 Hz, 1 H), 3.24–3.33
2 2 4
Cl . The combined organic extracts were dried with MgSO
4
solvents evaporated. Column chromatography of the residue on sil-
2
2
ica gel, with acetone as eluent, gave tosylate 25 as a white solid
1
1
(
111 mg, 90% yield); m.p. 149–150 °C. H NMR (300 MHz,
(
3
CDCl
3
): δ = 1.17 (s, 3 H), 1.58–1.61 (m, 1 H), 1.79 (dq, J = 5.0,
7
2
12.5 Hz, 1 H), 2.16–2.21 (m, 1 H), 2.43 (s, 3 H), 2.47 (s, 3 H), 2.61
(
(
7
t, J = 10 Hz, 1 H), 2.73–2.81 (m, 2 H), 3.18–3.29 (m, 2 H), 4.01
dd, J = 7.0, 9.5 Hz, 1 H), 4.10 (t, J = 9 Hz, 1 H), 6.09 (br.s, 1 H),
(
m, 2 H), 5.18–5.26 (m, 1 H), 5.40–5.46 (m, 1 H), 5.89 (br.s, 1 H)
1
3
ppm. C NMR (75 MHz, CDCl
3
): δ = –1.7, 18.7, 21.5, 27.4, 36.1,
.37 (d, J = 8.5 Hz, 2 H), 7.80 (d, J = 8.5 Hz, 2 H) ppm. 13C NMR
3
8.5, 41.4, 46.8, 53.4, 58.3, 65.3, 125.3, 126.7, 174.0 ppm. IR (neat):
(75 MHz, CDCl
3
): δ = 15.4, 21.4, 21.5, 21.6, 35.9, 37.4, 40.9, 41.1,
–
1
ν˜ = 3326, 2950, 1655, 1624, 1492, 1341, 1245, 1153, 1099 cm .
45.4, 54.2, 65.4, 69.6, 127.9, 129.9, 132.7, 145.0, 173.0 ppm. IR
+
+
HRMS (ESI , MeOH): m/z = 295.220 [M+H] (calcd. for
OSi 295.2206).
(
9
neat): ν˜ = 3369, 2929, 1737, 1659, 1353, 1216, 1188, 1174,
61 cm . HRMS (ESI , MeOH): m/z = 353.1520 [M+H] (calcd.
16 31 2
C H N
–1
+
+
24 2 4
for C17H N O
S 353.1535).
Tosylamides 29: Amides 28 (37 mg, 0.12 mmol) were dissolved in
dry THF (1.2 mL) and the mixture was cooled to 0 °C. LiHMDS
(0.25 mL, 1  in hexane, 0.25 mmol) was added to the resulting
solution, and the mixture was stirred for 15 min. p-Toluenesulfonyl
chloride (60 mg, 0.31 mmol) and DMAP (4.6 mg, 0.04 mmol) were
added in one portion, and the solution was allowed to warm slowly
to room temp. over 2 h. The reaction mixture was diluted with satu-
Cyanide 26: NaCN (512 mg, 10.4 mmol) was added to a solution
of tosylate 25 (368 mg, 1.04 mmol) in DMF (10 mL). The reaction
mixture was stirred at 100 °C for 3 h, diluted with a saturated aque-
ous NaHCO
combined organic extracts were dried with MgSO
evaporated. Column chromatography of the residue on silica gel
with CH Cl /MeOH (20:1) as eluent gave cyanide 21 as a white rated aqueous NaHCO
solid (196 mg, 91% yield); m.p. 129–130 °C. H NMR (300 MHz,
CDCl ): δ = 1.23 (s, 3 H), 1.58–1.61 (m, 1 H), 1.70–1.88 (m, 1 H),
.22–2.30 (m, 1 H), 2.39–2.46 (m, 2 H), 2.50 (s, 3 H), 2.78–2.90 (m,
3
solution, and extracted three times with CH
2 2
Cl . The
and the solvents
4
2
2
3
solution and extracted three times with
, the combined organic extracts were dried with MgSO and
1
CH
2
Cl
2
4
3
the solvents evaporated. Column chromatography of the residue
2
3
on silica gel (heptanes/ethyl acetate, 5:1) gave tosylamides 29 as a
1
3
1
H), 3.23–3.39 (m, 2 H), 6.47 (s, 1 H) ppm. C NMR (75 MHz, colorless oil (40 mg, 71% yield). H NMR (300 MHz, CDCl
3
): δ =
CDCl
1
1
3
): δ = 16.1, 18.0, 21.9, 35.2, 35.9, 40.7, 46.6, 57.2, 64.8, 118.9, 0.01 (s, 9 H), 1.12 (s, 3 H), 1.47 (d, J = 8.7 Hz, 2 H), 1.72–1.80 (m,
74.0 ppm. IR (neat): ν˜ = 3197, 2936, 2879, 2827, 2244 (CN), 1659, 1 H), 1.88–1.98 (m, 1 H), 2.03–2.23 (m, 3 H), 2.15 (s, 3 H), 2.36–
494, 1448, 1212, 1363, 1348, 1184, 1094, 979 cm . HRMS (ESI , 2.50 (m, 1 H), 2.42 (s, 3 H), 2.64–2.78 (m, 2 H), 3.74–3.82 (m, 1
–1
+
+
MeOH): m/z = 230.1275 [M+Na] (calcd. for C11
30.1269).
H
17
N
3
ONa
H), 4.10–4.16 (m, 1 H), 5.03–5.24 (m, 1 H), 5.39–5.53 (m, 1 H),
1
3
2
7.31 (d, J = 8.4 Hz, 2 H), 7.90 (d, J = 8.4 Hz, 2 H) ppm. C NMR
75 MHz, CDCl ): δ = –1.8, 16.2, 18.8, 21.6, 23.4, 27.5, 35.0, 38.8,
5.6, 47.1, 58.6, 68.7, 125.0, 127.1, 128.3, 129.2, 136.4, 144.3,
72.8 ppm. IR (neat): ν˜ = 2950, 1693, 1595, 1352, 1246, 1165,
(
3
Aldehyde 27: DIBAH (1  in CH
added dropwise at –78 °C to a solution of nitrile 26 (57 mg,
2 2
Cl , 0.57 mL, 0.57 mmol) was
4
1
1
0.27 mmol) in dry THF (0.5 mL), and the mixture was kept at
–
1
+
+
089 cm . HRMS (ESI , MeOH): m/z = 449.2287 [M+H] (calcd.
SSi 449.2294).
–78 °C for 1 h. DIBAH (1  in CH Cl , 0.57 mL, 0.57 mmol) was
2
2
37 2 3
for C23H N O
added again, and the mixture was then stirred at –78 °C for an
additional 1 h. MeOH and saturated aqueous NH Cl solution were
added to the mixture, which was allowed to warm to room temp.,
diluted with H O, and extracted three times with CH Cl . The
combined organic extracts were dried with MgSO and the solvents
evaporated. Crude aldehyde 27 (42 mg, 74% yield) was obtained as
a colorless oil. 1H NMR (300 MHz, CDCl
): δ = 1.16 (s, 3 H),
.44–1.49 (m, 1 H), 1.64–1.78 (m, 2 H), 2.07–2.15 (m, 1 H), 2.39
s, 3 H), 2.50–2.55 (m, 2 H), 2.66–2.88 (m, 3 H), 3.13–3.26 (m, 2
4
Iminium Precursor 30: DIBAH (1  in CH
2
Cl
.36 mmol) was added at –78 °C to a solution of lactam 29 (40 mg,
.09 mmol) in dry CH Cl (1 mL). The solution was allowed to
O (0.5 mL) was added.
The solution was filtered through Celite, which was washed with
CH Cl , the filtrate was concentrated, and the crude product
35 mg, 87% yield) was obtained, H NMR (300 MHz, CDCl
2
, 0.36 mL,
0
0
2
2
2
2
2
4
warm slowly to room temp. over 3 h, and H
2
3
2
2
1
(
1
(
3
;
major isomer): δ = 0.00 (s, 9 H), 1.14 (s, 3 H), 1.42 (d, J = 9 Hz, 2
H), 1.59–1.63 (m, 3 H), 1.83–1.98 (m, 3 H), 2.24 (s, 3 H), 2.43 (s,
13
H), 6.20 (br.s, 1 H), 9.74 (t, J = 1.2 Hz, 1 H) ppm. C NMR
75 MHz, CDCl ): δ = 15.0, 21.3, 31.1, 35.0, 40.1, 43.8, 45.5, 56.9,
4.1, 172.8, 200.1 ppm. IR (neat): ν˜ = 3295, 2930, 2852, 1716, 1650,
(
3
3
3
H), 2.72–2.76 (m, 1 H), 2.87–2.90 (m, 1 H), 3.13–3.16 (m, 1 H),
.29–3.35 (1 H), 5.07–5.25 (m, 3 H), 5.37–5.47 (m, 1 H), 7.31 (d, J
6
1
2
–
1
+
492, 1454, 1345, 1161, 1096 cm . HRMS (ESI , MeOH): m/z =
1
3
=
8.5 Hz, 2 H), 7.85 (d, J = 8.5 Hz, 2 H) ppm. C NMR (75 MHz,
; characteristic signals): δ = –1.8, 80.3, 125.6, 126.9, 127.9,
129.4, 136.0, 143.2 ppm. IR (neat): ν˜ = 3350, 2954, 1736, 1336,
11.1424 [M+H]+ (calcd. for C11
211.1447).
19 2 2
H N O
CDCl
3
Derivatives 28 (Mixture of Isomers): n-Butyllithium (1.6  in hex-
ane, 0.37 mL, 0.59 mmol) was added at 0 °C to a suspension of
methyltriphenylphosphonium bromide (206 mg, 0.57 mmol) in dry
–
1
+
1259, 1162, 1092, 1015, 853, 799 cm . HRMS (ESI , MeOH): m/z
= 451.2461 [M+H]+ (calcd. for C23
SSi 451.2451).
39 2 3
H N O
4112
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 4106–4114

Biogenetic Strategy Towards the Polycyclic Core Skeleton of Sarain A
FULL PAPER
Sarain A Core Skeleton Analogue 31: The above crude aminal was
tallizes in achiral space group C2 as described above. However, in-
dividual crystals can exhibit a very slight enrichment of one enanti-
omer over the other.
dissolved in dry CH
2 2
Cl (15 mL) and the mixture was cooled to
–78 °C. Anhydrous ferric chloride (126 mg, 0.3 mmol) was added
in one portion, and the resulting solution was warmed to room
temp. After 30 min, the solution was diluted with 1  aqueous
NaOH (2 mL) and stirred for 1 h, the mixture was extracted three
X-ray Crystallography of Diol 19: Colorless prismatic crystal
(
3
0.60×0.35×0.30 mm). Empirical formula C17
06.40, T = 293 K. Monoclinic system, space group P2
The unit cell [parameters: a = 8.033(4), b = 12.395(5), c =
H
26
N
2
O
3
, M =
1
/n, racemic.
times with CH
2 2
Cl , the combined organic extracts were dried with
MgSO and the solvents evaporated. Column chromatography of
4
3
1
(
6.817(7) Å, β = 94.98(2)°, V = 1668 Å ] contains four molecules
the residue on silica gel (acetone/ethyl acetate, 1:4) gave tricyclic
–
3
–1
Z = 4), d
C
= 1.220 gcm , F(000) = 664, µ = 0.084 mm , λ(Mo-
1
product 31 as a colorless oil (17 mg, 61% yield). H NMR
K
α
) = 0.71073 Å. Intensity data were measured with a Nonius
(
1
300 MHz, CDCl
.88 (m, 4 H), 2.00–2.14 (m, 2 H), 2.22–2.45 (m, 1 H), 2.41 (s, 3
H), 2.47 (s, 3 H), 3.16 (ddd, J = 5.4, 12.9, 13.2 Hz, 1 H), 3.31–3.36
m, 1 H), 3.62 (dd, J = 8.7, 13.5 Hz, 1 H), 3.93 (d, J = 4.5 Hz, 1
3
): δ = 1.14 (s, 3 H), 1.44–1.52 (m, 1 H), 1.76–
Kappa-CCD area detector diffractometer with use of graphite-mo-
nochromated Mo-K radiation, in φ- and ω-scans, up to θ = 29.1°.
7243 data were collected giving 7835 monoclinic (hkl and –h–k–l)
α
1
(
reflections, of which 4444 were unique. The structure was solved
H), 4.74 (dt, J = 10.5, 1.5 Hz, 1 H), 4.97 (dt, J = 16.8, 1.5 Hz, 1
H), 5.54–5.65 (m, 1 H), 7.26 (d, J = 8.7 Hz, 2 H), 7.66 (d, J =
with the SHELXS86 program and refined by full-matrix least
2
squares based upon unique F with the SHELXL97 program. The
8
2
1
1
.7 Hz, 2 H) ppm. 13C NMR (75 MHz, CDCl
2.5, 32.5, 33.6, 38.1, 38.9, 40.3, 41.6, 58.4, 59.5, 60.7, 114.5, 127.2,
3
): δ = 20.8, 21.4,
hydrogen atoms, located on difference Fourier maps, were fitted at
theoretical positions and treated as riding, except for the H–N15
atom (refined). They were assigned an isotropic displacement pa-
rameter equivalent to 1.15× that of the bonded atom (1.20× for
29.2, 138.5, 139.7, 142.5 ppm. IR (neat): ν˜ = 2924, 1681, 1453,
–1
+
327, 1155, 1093, 997 cm . HRMS (ESI , MeOH): m/z = 361.1968
S 361.1950).
[
M+H]+ (calcd. for C20
H
29
N
2
O
2
those of the methyl and hydroxy groups). Refinement of 206 pa-
2
rameters on F gave R
1
(F) = 0.0497 calculated with the 3022 ob-
X-ray Crystallography of Tosylate 17: Small prismatic colorless
crystal (0.55×0.35×0.35 mm). Empirical formula C24 S,
2
served reflections with I Ն 2σ(I) and wR
2
(F ) = 0.1352 considering
28 2 5
H N O
all the 4444 data. Goodness of fit = 1.006. The residual density
M = 456.54, T = 293 K. Monoclinic system, space group C2, chiral.
–
3
was found between –0.20 and 0.19 eÅ . One of the two enantio-
mers appears in Figure 2, showing all the pyrrolidine ring substitu-
ents fixed in an axial position. The pyrrolidine ring exhibits an
envelope conformation in C2, this atom deviating by 0.549(2) Å
from the mean plane of the other four atoms. An intramolecular
hydrogen bond is observed between the N15–H atoms and the ni-
trogen atom N1 [distances N15···N1 2.612(2) Å, HN15···N1 2.01 Å,
angle N–H···N 117.3°]. The crystal packing study shows that all
the hydroxy groups are engaged in intermolecular hydrogen bonds,
so the molecules are assembled in dimers through centosymmetric
centers, by hydrogen bonds formed between the O11–H hydroxy
groups towards the O8 oxygen atoms according to the following
scheme: distances O11–H(x, y, z)···O8(2–x, –y, 1–z) 2.801(3) Å,
The unit cell [parameters: a = 22.973(9), b = 7.806(3), c =
3
1
2.919(4) Å, β = 91.33(2)°, V = 2316 Å ] contains four molecules
–3
–1
(Z = 4), d
C
= 1.309 gcm , F(000) = 968, µ = 0.177 mm , λ(Mo-
K
α
) = 0.71073 Å. Intensity data were measured with a Nonius
Kappa-CCD area detector diffractometer with use of graphite-mo-
nochromated Mo-K radiation, in φ- and ω-scans, up to θ = 28.71°.
1528 data were collected, giving 5793 unique monoclinic (hkl
and –h–k–l) reflections. The structure was solved by direct methods
α
1
[
17]
with the SHELXS86
program and refined by full-matrix least
squares based upon unique F2 with the SHELXL97 program.
[18]
All the hydrogen atoms, located on difference Fourier maps, were
fitted at theoretical positions and treated as riding. They were as-
signed an isotropic displacement parameter equivalent to
HO11···O8 1.99 Å, angle O–H···O 168.5° and reciprocally; same en-
1
.12×that of the bonded atom (1.15 for those of the methyl
2
antiomeric molecules of these dimers are linked together through
hydrogen bonds formed between the O8–H hydroxy groups and the
nearest O14 oxygen atoms; according to a second scheme: distances
O8–H(x, y, z)···O14(1.5–x, 0.5+y, 1.5–z) 2.716 (3) Å, HO8···O14
1.91 Å, angle O–H···O 169.3° and reciprocally, between O14(x, y,
z) and O8–H(1.5–x, –0.5+ y, 1.5–z).
groups). Refinement of 292 parameters on F gave R
1
(F) = 0.0405
calculated with the 4899 observed reflections with I Ն 2σ(I) and
2
wR
2
(F ) = 0.1032 considering all the 5793 data. Goodness of fit =
1
0
.014. The residual density was found between –0.19 and
.16 e·Å . In the enantiomer drawn in Figure 1, the cis ring junc-
–3
tion along C4–C5 appears clearly. Torsion angle values show that
the pyrrolidine ring exhibits an envelope conformation, with atom
C5 deviating by –0.627(2) Å from the mean plane of the other four
atoms, whilst the six-membered ring also adopts an envelope con-
formation with atom C4 situated at –0.586(2) Å from the mean
plane of the other five atoms. The dihedral angle of these two rings
is 100.3°, whilst that of the two phenyl planes is 137.4°. It is note-
worthy that the crystal from a racemic solution belongs to the chi-
ral monoclinic space group C2. Analysis of Bijvoet pair differences
through the sulfur anomalous contribution (also measured with
CCDC-601385 (17) and -601386 (19) contain the supplementary
crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif.
Supporting Information (see footnote on the first page of this arti-
cle): Copies of NMR spectra for compounds 14–20, 22–23, 25–27,
31.
Cu-K
α
radiation on the same crystal) revealed random dispersion
[1] a) Y. Guo, A. Madaio, E. Trivellone, G. Scognamiglio, G. Cim-
ino, Tetrahedron 1996, 52, 8341–8348; b) V. Caprioli, G. Cim-
ino, A. De Giulio, A. Madaio, G. Scognamiglio, E. Trivellone,
Comp. Biochem. Physiol., Part B: Biochem. Mol. Biol. 1992,
between positive and negative differences corroborating the high
value of the Flack parameter (0.56), pointing to a case of racemic
lamellar twinning, so the two enantiomers are present in the crys-
tals. Such behavior prompted us to check whether tosylate 17 was
actually a racemic mixture. The rotatory power proved to be zero
1
03, 293–296; c) G. Cimino, G. Scognamiglio, A. Spinella, E.
Trivellone, J. Nat. Prod. 1990, 53, 1519–1525; d) G. Cimino,
C. A. Mattia, L. Mazarella, R. Puliti, G. Scognamiglio, A. Spi-
nella, E. Trivellone, Tetrahedron 1989, 45, 3863–3872.
D
([α] = 0± 0.06) and chiral HPLC (column OJ, hexane/iPrOH,
65:35, 1 mL/min) on individual crystals showed that the composi-
[
2] a) S. Hong, J. Yang, S. M. Weinreb, J. Org. Chem. 2006, 71,
2078–2089; b) O. Irie, K. Samizu, J. R. Henry, S. M. Weinreb,
J. Org. Chem. 1999, 64, 587–595; c) J. Sisko, J. R. Henry, S. M.
tions of each crystal were between 52:48 to 48:52 in ratio. As a
conclusion, tosylate 17 is definitely a racemic mixture but it crys-
Eur. J. Org. Chem. 2006, 4106–4114
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4113

S. Mons, C. Marazano et al.
FULL PAPER
Weinreb, J. Org. Chem. 1993, 58, 4945–4951; d) J. Sisko, S. M.
Weinreb, J. Org. Chem. 1991, 56, 3210–3211.
[9] A sponge-associated bacteria responsible for the biosynthesis
of manzamine A has been identified as a previously unde-
scribed species of Micromonospora (actinomycetes): R. T. Hill,
M. T. Hamann, O. Peraud, N. Kasanah, Worldwide patent WO
2004/013297 (US 2003/024238), 12 February 2004, pp. 49.
[10] P. Bey, J. P. Vevert, J. Org. Chem. 1980, 45, 3249–3253.
[11] E. B. Maxted, S. Akhtar, J. Chem. Soc. 1959, 3130–3133.
[12] Relative configurations of diazabicyclo[4.3.0]nonane systems
could be guesstimated from NOESY experiments, but the
weakness of the NOE cross-peaks prevented any definitive con-
clusion before X-ray analysis.
[
[
[
3] a) C. J. Douglas, S. Hiebert, L. E. Overman, Org. Lett. 2005,
7
, 933–936; b) S. Jaroch, R. T. Matsuoka, L. E. Overman, Tet-
rahedron Lett. 1999, 40, 1273–1276; c) R. Downham, F. W. Ng,
L. E. Overman, J. Org. Chem. 1998, 63, 8096–8097.
4] a) H. I. Lee, M. J. Sung, H. B. Lee, J. K. Cha, Heterocycles
2
004, 62, 407–422; b) M. J. Sung, H. I. Lee, H. B. Lee, J. K.
Cha, J. Org. Chem. 2003, 68, 2205–2208; c) M. J. Sung, H. I.
Lee, Y. Chong, J. K. Cha, Org. Lett. 1999, 1, 2017–2019.
5] a) D. J. Denhart, D. A. Griffith, C. H. Heathcock, J. Org.
Chem. 1998, 63, 9616–9617; b) D. A. Griffith, C. H. [13] Though tosylate 17 is racemic, as corroborated by its [α]
D
value
Heathcock, Tetrahedron Lett. 1995, 36, 2381–2384; c) C. H.
Heathcock, M. Clasby, D. A. Griffith, B. R. Henke, M. J.
Sharp, Synlett 1995, 467–474; d) B. R. Henke, A. J. Kouklis,
C. H. Heathcock, J. Org. Chem. 1992, 57, 7056–7066.
and by chiral HPLC, its crystals unexpectedly exhibit a chiral
symmetry group. See Experimental Section for more details.
[14] W. N. Speckamp, H. Hiemstra, Tetrahedron 1985, 41, 4367–
4416.
[15] Modification of the procedure described by: S. Ma, X. Lu, Org.
Synth. 1995, 72, 112–115.
[
[
[
6] N. K. Garg, S. Hiebert, L. E. Overman, Angew. Chem. Int. Ed.
2006, 45, 2912–2915.
7] S. Hourcade, A. Ferdenzi, P. Retailleau, S. Mons, C. Marazano,
Eur. J. Org. Chem. 2005, 1302–1310.
[16] A. Gomez-Sanchez, I. Hermosin, J. M. Lassaletta, I. Maya,
Tetrahedron 1993, 49, 1237–1250.
[17] G. M. Sheldrick, Acta Crystallogr., Sect. A 1990, 46, 467–473.
[18] G. M. Sheldrick, SHELXL-97, Program for the Refinement of
Crystal Structures, University of Göttingen, Germany, 1997.
Received: April 4, 2006
8] a) M. d. R. Sanchez-Salvatori, C. Marazano, J. Org. Chem.
2
003, 68, 8883–8889; b) M. Herdemann, A. Al-Mourabit, M.-
T. Martin, C. Marazano, J. Org. Chem. 2002, 67, 1890–1897;
c) K. Jakubowicz, K. Ben Abdeljelil, M. Herdemann, M.-T.
Martin, A. Gateau-Olesker, A. Almourabit, C. Marazano,
B. C. Das, J. Org. Chem. 1999, 64, 7381–7387; d) A. Kaiser, X.
Billot, A. Gateau-Olesker, C. Marazano, B. C. Das, J. Am.
Chem. Soc. 1998, 120, 8026–8034.
Published Online: July 19, 2006
4114
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 4106–4114