First highly efficient asymmetric synthesis of the Hyrtios erectus diketotriterpenoid

Article abstract of DOI:10.1055/s-2002-34949

The first highly efficient asymmetric synthesis of the diketotriterpenoid 1, isolated from the Indonesian marine sponge Hyrtios erectus, in good overall yield and employing simple starting materials such as butanone, is described. Both stereogenic centres at the C-6 and C-19 positions of the C₂-symmetrical molecule were generated via α-alkylation employing the SAMP/RAMP hydrazone method with high asymmetric inductions (de, ee ≥ 96%). The absolute configuration of the natural material was determined as R,R.

Full text of DOI:10.1055/s-2002-34949

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
2280
PAPER
First Highly Efficient Asymmetric Synthesis of the Hyrtios Erectus
Diketotriterpenoid
First
H
ighly Effici
i
ent
A
sy
e
m
metric
S
y
nthesis
e
of the
H
yr
r
tios
E
rectus
D
E
iketotriterpenoi
n
d
ders,* Thomas Schüßeler
Institut für Organische Chemie, Rheinisch-Westfälische Technische Hochschule, Professor-Pirlet-Str. 1, 52074 Aachen, Germany
Fax +49(241)8092127; E-mail: enders@rwth-aachen.de
Received 22 June 2002; revised 4 July 2002
Dedicated to Professor Ekkehard Winterfeldt on the occasion of his 70th birthday.
symmetric and should have either the S,S- or R,R-config-
Abstract: The first highly efficient asymmetric synthesis of the
uration.2
diketotriterpenoid 1, isolated from the Indonesian marine sponge
Hyrtios erectus, in good overall yield and employing simple starting
materials such as butanone, is described. Both stereogenic centres
at the C-6 and C-19 positions of the C2-symmetrical molecule were
generated via -alkylation employing the SAMP/RAMP hydrazone
method with high asymmetric inductions (de, ee 96%). The abso-
lute configuration of the natural material was determined as R,R.
As metallated hydrazones are the reagents of choice for
the regio- and stereoselective synthesis of substituted ke-
tones, we decided to attempt the first asymmetric synthe-
sis of both enantiomers of the title diketotriterpenoid 1
employing our SAMP/RAMP hydrazone methodology.5
Scheme 1 summarizes the two retrosynthetic concepts of
the approach. In both cases the stereo- and regioselectivity
of the alkylation of butanone are controlled by the use of
SAMP as chiral auxiliary and silyl groups as blocking
groups.
Key words: asymmetric synthesis, hydrazones, alkylations, natural
products, terpenoids
Marine sponges are still of interest to scientists searching
for new anticancer drugs and biologically active com-
pounds in general.1 In 1998 Anderson et al. isolated the
diketotriterpenoid 1 from the marine sponge Hyrtios erec-
tus, which was collected in Indonesia (Figure 1).2 The
structure was elucidated by spectroscopic analysis. Ap-
parently, the diketone 1 is the first triterpenoid known
from the genus Hyrtios, although triterpenoids are rela-
tively common sponge metabolites.3 Numerous sesterter-
penoids and related compounds could be isolated by
previous chemical studies of sponges in the genus Hyr-
tios.4
O
+
H3C
H3C
OPG
+
X
CH3
X
CH3
CH3
B
C
O
H3C
OH
2 x
CH3
CH3
CH3
A
route I
O
CH3
CH3
CH3
CH3
O
CH3
CH3
CH3
CH3
H3C
H3C
CH3
CH3
CH3
O
CH3
CH3
CH3
O
1
1
Diketotriterpenoid of the marine sponge Hyrtios erectus
Figure 1
route II
CH3
O
CH3
Crude extracts of the Indonesian sponge Hyrtios erectus
(order Dictyoceratida, family Thorectidae) exhibited sig-
nificant antimitotic activity. However, the new triter-
H3C
H3C
X
X
+
X
+ 2 x
2 x
CH3
CH3
B
D
penoid
1 was one of the biologically inactive
Scheme 1 Retrosynthetic analysis
chromatography fractions. Comparison of the analytical
data (NMR, MS) indicated that 1 contained a 2-fold ele-
ment of symmetry. Although the absolute configuration at
C-6/C-19 could not be determined because of the limited
quantity of 1 the optical activity confirmed that 1 is C2-
The retrosynthetic Route I uses the C2-symmetry of 1 in
the reductive coupling of the allylic alcohol A, which
should be available by the alkylation of butanone with
electrophiles B and C.
The retrosynthetic Route II builds up the title ketone 1 by
symmetrical dimerization–alkylation of butanone with the
dihalide D and subsequent double alkylation with electro-
phile C.
Synthesis 2002, No. 15, Print: 29 10 2002.
Art Id.1437-210X,E;2002,0,15,2280,2288,ftx,en;Z09602SS.pdf.
© Georg Thieme Verlag Stuttgart · New York
ISSN 0039-7881

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
PAPER
First Highly Efficient Asymmetric Synthesis of the Hyrtios Erectus Diketotriterpenoid
2281
rectly treated with 33% H2SO4 to provide the homoallylic
iodide 3 in a single operation and good yield.
O
H3C
a
I
CH3
52%
CH3
The allylic halid C was synthesized starting from 1,4-
butenediol (4). Protection of the alcohols gave 5, which
was oxidatively cleaved with ozone to yield the TBS-pro-
tected (t-butyldimethylsilyl) hydroxyacetaldehyde (6).
The aldehyde was converted by a Horner–Wadsworth–
Emmons reaction into the corresponding , -unsaturated
ester 7 using triethyl phosphonopropionate and BuLi in
the presence of lithium bromide.7 A quantitative yield and
high E/Z ratio (94:6) was obtained. Reduction of the ester
7 with DIBALH to the alcohol 8 and further treatment
with mesylchloride and lithium bromide gave the bromide
9. Several attempts to convert the allylic alcohol 8 into the
corresponding iodide were without success. Only decom-
position and allylic rearrangement were observed.
2
3
c
b
HO
OH
TBSO
OTBS
99%
96%
4
5
e
TBSO
CH3
d
TBSO
quant.
O
quant.
COOEt
CH3
H
6
7
TBSO
TBSO
CH3
f
82%
Br
OH
9
8
The hydrazone 11 was prepared from butanone by heating
with (S)-1-amino-2-(methoxymethyl)pyrrolidine (SAMP)
in virtually quantitative yield. Deprotonation was
achieved with lithium tetramethylpiperidide (LiTMP) at 0
°C. The resulting aza-enolate was treated with thexyl-
dimethyl chlorosilane (TDS–Cl) to afford the correspond-
ing -silyl hydrazone 12 in high yield (Scheme 3).8
Scheme 2 Route I Reagents and conditions: (a) (1) MeMgI, Et2O,
reflux, 30 min; (2) H2SO4, H2O, 0 °C, 30 min; (b) TBS–Cl, imidazole,
DMF, 0 °C, 20 h; (c) (1) O3, CH2Cl2, –78 °C; (2), Me2S, r.t., 16 h; (d)
(EtO)2OPCH(Me)CO2Et, BuLi, LiBr, MeCN, r.t., 20 h; (e) DIBALH,
CH2Cl2, –78 °C (up to –20 °C), 1 h; (f) (1) MesCl, Et3N, –40 °C, 1 h;
(2) LiBr, THF, 0 °C, 1 h.
Under the usual deprotonation and alkylation conditions
(LDA, 0 °C, 1 h; RX, –78 °C, 16 h) the silyl hydrazone 12
could not be alkylated with the bromide 9.8 After modifi-
The homoallylic compound B was prepared by a modifi-
cation of the method of Julia (Scheme 2).6 The adduct of
the Grignard reagent and acetylcyclopropane (2) was di-
MeO
MeO
MeO
O
N
N
N
N
N
N
a
d
b
c
H3C
CH3
H3C
H3C
quant.
H3C
OTBS
85%
95%
92%
CH3
SiR3
SiR3 CH3
10
11
12
13 (de ≥ 96%)
MeO
MeO
N
N
N
N
e
f
H3C
OTBS
H3C
OH
56%
10%
CH3
CH3 SiR3 CH3
CH3
CH3 SiR3 CH3
14 (de ≥ 96%)
15 (de ≥ 96%)
MeO
N
N
CH3 SiR3 CH3
CH3
CH3
g
H3C
88%
CH3
CH3 SiR3 CH3
N
N
16 (de ≥ 96%)
OMe
O
CH3
CH3
CH3
CH3
H3C
CH3
CH3
CH3
O
(S,S)-1
R3Si = t-hexylMe2Si (TDS)
Scheme 3 Reagents and conditions: (a) SAMP, 60 °C, 12 h; (b) (1) LiTMP, THF, 0 °C, 1 h; (2) TDS–Cl, –78 °C, 16 h; (c) (1) BuLi, THF, 0
°C, 5 h; (2) TMP, 0 °C, 3 h; (3) 9, 0 °C, 16 h; (d) (1) 2.0 equiv t-BuLi, THF, 0 °C, 5 h; (2) 2.1 equiv 3, –78 °C, 16 h; (e) TBAF, THF, r.t., 16
h; (f) (1) TiCl3, LiAlH4, THF, 0 °C, 15 min; (2) reflux, 16 h; (g) oxalic acid, Et2O, reflux, 14 h.
Synthesis 2002, No. 15, 2280–2288 ISSN 0039-7881 © Thieme Stuttgart · New York

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
2282
D. Enders, T. Schüßeler
PAPER
cation of the conditions (BuLi, 0 °C, 5 h, TMP, 0 °C, 3 h;
9, 0 °C, 16 h) the -alkylated hydrazone 13 was formed in
excellent yield (92%) and a diastereomeric excess de
96% (13C NMR).
H
CH3
COOEt
a
b
O
EtOOC
87%
2 steps
O
H
CH3
17
18
19
In order to control the regioselectivity of the subsequent
-alkylation step a variety of conditions was examined.
Finally, deprotonation of 13 employing 2.0 equiv t-BuLi
at 0 °C and alkylation of the resulting aza-enolate with io-
dide 3 at –78 °C gave the -alkylated SAMP-hydrazone
14 in good yield (85%) and high diastereomeric excess de
96% (13C NMR).
CH3
CH3
c
d
Br
OH
73%
80% Br
HO
CH3
CH3
21
20
Scheme 4 Reagents and conditions: (a) (1) O3, CH2Cl2, –78 °C; (2),
Me2S, r.t., 16 h; (b) (EtO)2OPCH(Me)CO2Et, BuLi, MeCN, r.t., 20 h;
(c) DIBALH, CH2Cl2, –78 °C, 16 h; (d) (1) MesCl, Et3N, –40 °C, 1 h;
(2) LiBr, THF, 0 °C, 1 h.
Deprotection of the hydroxy group was carried out using
TBAF to give the allylic alcohol 15 without detectable
epimerization. The yield was only 56% but at the same
time 31% of the starting material could be recovered. Us-
ing the t-butyldiphenylsilyl group for protection of the hy-
droxy function gave a higher yield of 87% in the O–Si
cleavage but caused partial epimerization under the corre-
sponding conditions (reflux with 10% KOH–MeOH, de =
90– 96%).
Taking into account the knowledge about the alkylation
conditons of the hydrazone 12 with bromide 9, we were
able to prepare the bishydrazone 22 under similar condi-
tions (Scheme 5). It was necessary to work in high con-
centration and to reflux the mixture for 30 min to increase
the yield. Under different conditions we got the
monoalkylated hydrazone bromide in yields up to 85%.
Nevertheless, the double hydrazone 22 was formed in
good yield (71%) and de of 96% (13C NMR).
The reductive coupling of an alcohol (ROH) to hydrocar-
bon (RR) was first described by van Tamelen and
Schwartz in 1965.9 Further procedures for the reductive
coupling of allylic alcohols to provide 1,5-dienes through
the action of TiCl3 and MeLi or LiAlH4 have been report-
ed.10 In addition, allylic alcohols dimerize with loss of the
hydroxyl group by employing NbCl4/NaAlH4.11
The following double -alkylation of 22 with iodide 3
was carried out via the corresponding aza-enolate by
deprotonation with t-BuLi at 0 °C. To avoid monoalkyla-
tion it was again necessary to work in high concentration
and to reflux the mixture for 30 minutes. The bishydra-
zone 16 was isolated in excellent yield (94%) and de
96% (13C NMR).
We investigated the coupling of allylic alcohol 15 under
the conditions of the described systems. Only the use of
TiCl3 and LiAlH4 gave the desired dihydrazone 16 in 10%
96%; 13C
Next, we studied the hydrazone cleavage and subsequent
desilylation of the hydrazone 16, which was avaible via
both routes. Usually, the oxidative cleavage of SAMP hy-
drazones with ozone is a clean and quantitative method,
yield and high diastereomeric excess (de
NMR). Subsequently, a modification based on the use of
titanium dichloride was introduced, simplifying the pro-
cedure considerably. But even in this case the yield (6%)
was disappointing. A lot of side reactions such as isomer-
ization, elimination, desilylation and hydrazone cleavage
could be observed.
MeO
MeO
a
N
N
N
CH3 SiR3
Due to the low yield of the reductive coupling of the allyl-
ic alcohol 15 we tried an alternative route to synthesize the
bishydrazone 16. Instead of generating the C2-symmetry
at the end of the synthesis by dimerisation, we envisaged
to build up 16 by symmetrical double alkylations
(Scheme 1, Route II).
N
71%
H3C
H3C
CH3
SiR3 CH3
N
N
SiR3
22 (de ≥ 96%)
12
OMe
MeO
N
N
CH3 SiR3 CH3
CH3
CH3
In analogy to bromide 9 the dihalide D was synthesized
starting from 1,5-cyclooctadiene (17). Ozonolysis gave
the unstable succinic aldehyde (18) which was directly
converted by a double Horner–Wadsworth–Emmons re-
action into the corresponding , -unsaturated ester 19. A
high E/Z ratio (95:5) and good yield over two steps was
obtained. Reduction of the diester 19 with DIBALH to the
diol 20 and further treatment with mesyl chloride and lith-
ium bromide gave the dibromide 21 (Scheme 4).
H3C
b
CH3
CH3 SiR3 CH3
N
94%
N
16 (de ≥ 96%)
OMe
CH3
O
CH3
CH3
CH3
H3C
c
CH3
CH3
CH3
O
88%
R3Si = t-hexylMe2Si (TDS)
(S,S)-1 (de ≥ 96%)
Scheme 5 Reagents and conditions: (a) (1) BuLi, THF, 0 °C, 5 h;
(2) TMP, 0 °C, 3 h; (3) 0.5 equiv 21, –78 °C °C, 2 h; (4) reflux, 30
min; (b) (1) 4.0 equiv t-BuLi, THF, 0 °C, 5 h; (2) 4.0 equiv 3, –78 °C,
2 h; (3) reflux 30 min; (c) oxalic acid, Et2O, reflux, 14 h.
Synthesis 2002, No. 15, 2280–2288 ISSN 0039-7881 © Thieme Stuttgart · New York

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
PAPER
First Highly Efficient Asymmetric Synthesis of the Hyrtios Erectus Diketotriterpenoid
2283
however, the sensitivity of the double bonds in this case to
further oxidation led us to search for alternative condi-
tions.12 Therefore, we chose oxalic acid and CuCl2 for the
cleavage of the hydrazone 16. These racemisation-free
cleavage procedures are compatible with functionalities
that are sensitive to oxidative cleavage conditions, such as
C–C double bonds. The Cu(II)-mediated hydrolysis as
well as the treatment with oxalic acid solution gave direct-
ly the desilylated title diketotriterpenoid 1 in a single step
in and good yield (88%). The resulting silyl ketone of the
simple cleavage without desilylation could not be isolat-
ed.
All reagents were of commercial quality used from freshly opened
containers. Solvents were dried and purified by conventional meth-
ods prior to use. THF was freshly distilled from sodium/lead alloy
under Ar. All aq solutions were saturated, unless otherwise noted.
All reactions were carried out under an argon atmosphere using an-
hyd solvents unless stated otherwise. BuLi (15% in hexane) and t-
BuLi (15% in pentane) were purchased from Merck, Darmstadt.
Preparative column chromatography: Merck silica gel 60, particle
size 0.040–0.063 mm (230–400 mesh, flash). Analytical TLC: Sili-
ca gel 60 F254 plates, Merck, Darmstadt. Optical rotation values
were measured on a Perkin–Elmer P241 polarimeter, solvents used
were of Merck UVASOL-quality. Microanalyses were obtained
with a Heraeus CHN-O-RAPID element analyzer. Mass spectra:
Finnigan MAT 212 (CI 100 eV; EI 70 eV). IR spectra: Perkin–Elm-
er FT/IR 1750. 1H NMR spectra (300 and 400 MHz), 13C NMR (75
and 100 MHz): Gemini 300 or Varian Inova 400 (CDCl3 as solvent,
TMS as internal standard).
Because it was impossible to obtain separation of the
enantiomers and diastereomers of a racemic sample of 1
neither by NMR nor GC and HPLC on chiral stationary
phases, the de- and ee-values could not be determined by
these methods. Because under the hydrazone cleavage
conditions used no racemization at the stereogenic centers
occurs, (S,S)-1 was assumed to be of high diastereo- and
enantiomeric purity (de, ee 96%). On the other hand, the
comparison of the optical rotation values with those given
in the literature showed a large difference: found (S,S)-1
[ ]D24 + 15.50 (c 0.50, CHCl3); lit.2 [ ]D24 –371 (c 0.055,
CH2Cl2). After getting similar results under different
cleavage conditions we suppose that our polarimetry data
are correct.
5-Iodo-2-methyl-2-pentene (3)
A solution of methylmagnesium iodide was prepared from MeI
(14.19 g, 0.1 mol) and magnesium turnings (2.43 g, 0.1 mol) in ab-
solute Et2O (40 mL). To this, a solution of acetylcyclopropane (2)
(8.2 g, 97 mmol) in Et2O (10 mL) was added dropwise. The Grig-
nard adduct thus formed was slowly added to a cold (0 °C) solution
of concd H2SO4 (15 mL) in H2O (30 mL). After the complete addi-
tion, stirring was continued for 30 min. The ethereal layer was sep-
erated, and the aq layer was extracted with Et2O (2 15 mL). The
combined ethereal phases were washed with aq sodium hydrogen
sulfite (5%; 50 mL), aq NaHCO3 (5%; 50 mL) and brine (50 mL),
dried (MgSO4) and concentrated in vacuo. The crude product was
distilled under reduced pressure (bp 65 °C at 14 mbar) to yield io-
dide 3.
O
CH3
Yield: 11.10 g, 52 mmol (52 %); colorless oil.
H3C
H3C
Br
+
I
CH3
Br
2 x
+ 2 x
The analytical data were consistent with the data given in the liter-
ature.6
CH3
CH3
via RAMP-hydrazone
CH3 CH3
5 steps
56%
1,4-Di-[(tert-butyldimethylsilyl)oxy]-2-butene (5)
O
CH3
CH3
To a cooled solution (0 °C) of imidazole (10.2 g, 15 mmol) in DMF
(20 mL) was slowly added butenediol 4 (4.405 g, 50 mmol) and
TBS–Cl (16.6 g of a 50% solution in toluene, 110 mmol). The mix-
ture was stirred at r.t. for 30 h and quenched with H2O (50 mL). The
solution was extracted with Et2O (2 30 mL), the combined organic
extracts were washed with H2O (4 50 mL) and brine (50 mL),
dried (MgSO4), and concentrated in vacuo. Purification by distilla-
tion under reduced pressure (bp 58 °C at 10 mbar) yielded 5.
H3C
CH3
CH3
CH3
O
(R,R)-1 (de, ee ≥ 96%)
Scheme 6 Asymmetric synthesis of (R,R)-1
Nevertheless, analyzing the sign of optical rotation we
propose that the configuration of the natural diketo-tri-
terpenoid is R,R. At this point we decided to synthesize
the other enantiomer (R,R)-1 as well (Scheme 6). Accord-
ingly, the enantiomer (R,R)-1 was synthesized via double
alkylation with the dibromide 21 and the iodide 3 with a
diasteromeric and enantiomeric excess of 96%, starting
from butanone (10) and using RAMP as the chiral auxil-
iary via (R,R)-16; (R,R)-1 [ ]D24 –14.82 (c 0.50, CHCl3).
Yield: 15.12 g, 48 mmol (96%); colorless oil.
The analytical data were consistent with the data given in the liter-
ature.13
[(tert-Butyldimethylsilyl)oxy]acetaldehyde (6)
A solution of 5 (12.54 g, 40 mmol) in CH2Cl2 (20 mL) was ozono-
lized at –78 °C until the solution turned lightly blue. The mixture
was stirred with Me2S (24.6 g, 400 mmol) at r.t. for 12 h. After evap-
oration of the solvent the crude product was distilled under reduced
pressure (bp 62 °C at 10 mbar) to yield 6.
In summary, the first and very efficient asymmetric syn-
thesis of both enantiomers (S,S and R,R) of the diketo-
triterpenoid 1 was achieved via two different routes em-
ploying the SAMP/RAMP-hydrazone method. The yield
Yield: 13.86 g, 79.7 mmol (99%); colorless oil.
The analytical data were consistent with the data given in the liter-
ature.14
of (S,S)- and (R,R)-1 starting from butanone (10) over 7 Ethyl (E)-4-[(tert-Butyldimethylsilyl)oxy]-2-methyl-2-buten-1-
oate (7)
steps was 4% via Route I and 56% over 5 steps via Route
II. The absolute configuration of the material isolated
from the marine sponge Hyrtios erectus could be deter-
mined as R,R by polarimetry.
Anhyd LiBr (3.472 g, 40 mmol) was dissolved in MeCN (60 mL).
After 15 min stirring at r.t. triethyl phosphonopropionate (7.144 g,
30 mmol) was added dropwise and stirred an additional 15 min. At
0 °C BuLi (20 mL, 32 mmol) was added and after stirring for 15
Synthesis 2002, No. 15, 2280–2288 ISSN 0039-7881 © Thieme Stuttgart · New York