Asymmetric synthesis of the core cyclopentane of viridenomycin

Article abstract of DOI:10.1039/a903160j

The fully substituted cyclopentene ring with three stereogenic centers has been efficiently constructed for the first time using chiral bicyclic lactams, the insertion of the quatetnary center being the key step; ring opening of chiral sulfates to the correct 1,2-dihydroxy substituents also played a major role in reaching 1 in suitable form (18) for further study.

Full text of DOI:10.1039/a903160j

Asymmetric synthesis of the core cyclopentane of viridenomycin
Mark P. Arrington and A. I. Meyers*
Department of Chemistry, Colorado State University, Fort Collins, Colorado 80523-1872, USA.
E-mail: aimeyers@lamar.colostate.edu
Received (in Corvallis, OR, USA) 20th April 1999, Accepted 28th May 1999
The fully substituted cyclopentene ring with three stereo-
genic centers has been efficiently constructed for the first
time using chiral bicyclic lactams, the insertion of the
quaternary center being the key step; ring opening of chiral
sulfates to the correct 1,2-dihydroxy substituents also played
a major role in reaching 1 in suitable form (18) for further
study.
exo approach to 3c was hypothesized for the present transforma-
tion. The stereochemistry would be confirmed at a later time.
Reduction of the quaternary lactam ester 3c with NaBH4 in
EtOH furnished the corresponding alcohol 4, in high overall
yield (65%) for the first three steps (2–4). The alcohol 4 was
now easily separable from the accompanying a,a-dimethyl
material 3b. An X-ray crystal structure of alcohol 4 was
obtained and confirmed that the hydroxymethyl group in 4 was
indeed on the b-face of the molecule. This is the first published
example of the lactam enolate of 2 acylating preferentially from
the exo face. The hydroxymethyl side chain of 4 was
transformed to its benzyl ether 5 in 96% yield, and the chiral
auxiliary removed by reduction–hydrolysis to yield the cyclo-
pentenone 6 with high efficiency (76%) for the three step
process (5 ? 6).
Recently, a natural product was isolated from the culture broth
of an actinomycete identified as Streptomyces gannmycicus,
and upon structure determination, it was found that this
compound was identical to viridenomycin, a natural product
that had been known for some time, but which had not been
structurally characterized.1 Viridenomycin has been shown to
Studies were undertaken to access the allylic alcohol 7a, via
various reducing agents in anticipation of subsequent olefin
dihydroxylation, to 9. Since the benzyl protected hydroxy-
methyl side chain may be considered larger than the methyl
group in 6, it was expected that reducing agents should approach
preferentially from the bottom face of the cyclopentenone 6, to
yield the b-OH group. It was found (Scheme 2) that the highest
stereoselectivity could be realized (3.5+1) utilizing DIBAL-H7
to afford 7a. This modest selectivity was considered acceptable
because the major isomer 7a could be isolated in good (78%)
yield, and the undesired a-OH isomer 7b was also recyclable by
oxidizing to the enone 6 with CrO3 in high (86%) yield.
Repetitive reductions with DIBAL-H produced additional
quantities of 7a. The latter was converted to its p-methoxy-
benzyl (PMB) ether 8, (KH, PMBCl) in 95% yield. Osmylation
proceeded smoothly (90%) to give an inseparable mixture of
diastereomeric diols 9 with relatively high (10+1) selectivity.
The protons at C-2, C-3 (viridenomycin numbering) in 9 were
sufficiently discernible in the NMR to determine the level of
selectivity. Cyclic sulfate formation was accomplished using
the Sharpless technique,8,9 producing a single isomer (10) in
Me
O
10
HO
1
5
17
Ph
O
2
3
MeO
4
24
NH
Me
HO
O
Viridenomycin
O
RO
OMe
OR'
MeO
Me
RO
1
act as an antifungal antibiotic, and additionally has prolonged
the survival of mice infected with B16 melanoma. It is
structurally very similar to the natural product hitachimycin,2,3
although viridenomycin has not yet been the subject of any
reported synthetic efforts. It has, as its core, a highly
functionalized cyclopentene ring 1 with three contiguous
stereogenic centers, one of which is quaternary. Additionally,
the absolute stereochemistry of viridenomycin is still unknown
with the stereochemistry at C-17 still in question.
With our interest in chiral, bicyclic lactams as templates for
a variety of enantiomerically pure compounds containing
quaternary stereocenters, the C-4 position of viridenomycin
suggested the possibility of an asymmetric synthesis.
The requisite chiral bicyclic lactam 2 (Scheme 1) was readily
prepared by condensation of levulinic acid and enantiomerically
pure natural valinol.4 Treatment of 2 with LDA followed by
MeI quench gave a mixture of products, containing primarily
the desired a-methyl bicyclic lactam 3a, in addition to some of
the a,a-dimethylated compound 3b.5 The latter was only
partially separable from the desired compound, so the crude
product was simply carried on and directly treated again with
LDA. The resulting enolate was quenched with ClCO2Me,
giving 3c with high ( > 98%) diastereoselectivity (NMR).
Again, the a,a-dimethylated derivative 3b was difficult to
remove, and the mixture of 3b and 3c was carried forward. It
had been previously observed6 that acylating agents such as acid
chlorides approached the lactam enolate in an exo fashion, so an
O
O
O
i, ii
iii
R
N
N
N
OH
Me
Me
O
O
O
2
3a R = H
3b R = Me
(+)-4
(+)-
3c R = CO2Me
iv
O
O
v–vii
N
OBn
Me
OBn
Me
(+)-6
O
(+)-5
Scheme 1 Reagents and conditions: i, LDA, MeI, THF, 278 °C ? rt; ii,
LDA, ClCO2Me, THF, 278 °C ? rt (50+1 dr); iii, NaBH4 (7 equiv.), EtOH,
0 °C ? rt (65%, 3 steps); iv, NaH, BnCl, DMF, 0 °C ? rt (95%); v, Red-Al
(6.6 equiv.), THF, 225 °C, 2 d; vi, NBu4NH2PO4 (50 equiv.), EtOH–H2O
(1+1); vii, KOH (0.1 equiv.), THF (76%, 3 steps).
Chem. Commun., 1999, 1371–1372
1371

R1
O
OPMB
Me
R2
O
i
ii
CO2Me
OBn
i
MeO
(+)-6
MeO
TBDMSO
OBn
OBn
OBn
TBDMSO
Me
Me
Me
17
(–)-7a R1 = OH, R2 = H
(+)-7b R1 = H, R2 = OH
16
(–)-8
ii
iii
OTBDMS
CO2Me
OPMB
OPMB
OPMB
MeO
vi
iv, v
AcO
HO
O
HO
HO
O
OBn
TBDMSO
Me
(+)-18
S
O
OBn
OBn
OBn
O
Me
11
Me
Me
Scheme 4 Reagents and conditions: i, LDA, HMPA (25 equiv.),
NCCO2Me, 278 ? 220 °C; ii, TBDNCl, imidazole, DMAP (40%, 2
steps).
(–)-10
9
Scheme 2 Reagents and conditions: i, DIBAL-H (1.5 equiv.), THF. 278 °C
(99%, 3.5+1 dr); ii, KH, (2.2 equiv.) PMBCl, DMF, 0 °C (95%); iii,
K2OsO2·H2O, NMO, acetone–H2O (3+1) (90%, 10+1 dr); iv, SOCl2, Et3N,
CH2Cl2, 0 °C; v, RuCl3·H2O, NaIO4, CCl4–H2O–MeCN (1+1.4+1) (85%, 2
steps); vi, CsOAc (5 equiv.), DMF, 40 °C, then H2SO4, THF (99%, 9+1
dr).
the cleavage was performed with DDQ, which quickly removed
the PMB group in good yield (93%) and gave the free C-1
alcohol 15. Mild oxidation under Swern conditions produced
the desired ketone 16 (91%).
The enolate of 16 was generated in the presence of HMPA10
as cosolvent and warmed to 240°C before quenching with
NCO2Me. Enolate formation generated small amounts of the
epimerized 16 and elimination products. However, the desired
product 17 was the major constituent of the mixture, along with
still significant quantities of undesired products. It appears from
all of these experiments that the source of epimerization was not
initial generation of the incorrect enolate, but was due to facile
proton transfer processes after the addition of the electrophile.
Having obtained a quantity of the carboxylated derivative,
17, attention was turned to isolation of the enol ester 18, the
intended target of this study (Scheme 4). Attempts to silylate 17
gave an enol ether, which was poorly characterized. By
combining the two steps from 16 to 18 (no attempt was made to
purify the b-keto ester 17), the crude material was directly
silylated and gave an overall yield of 18 of 40%. With the
acquisition of 18, all of the correct regio- and stereo-chemical
features have been installed and the substituents at the 4- and
5-positions are in a position to further extend the synthesis to
viridenomycin itself.
satisfactory (77%) yield for the two steps. Caesium acetate
opening of the sulfate 10 occurred with high regioselectivity
(10+1) and yield (99%) to furnish the acetoxy alcohol 11 as the
major regioisomer. In this case, the regioisomeric products were
not separable and thus the mixture was carried forward.
Protection of the alcohol 11 (Scheme 3) was accomplished in
excellent yield (96%) with TBDMSCl and imidazole affording
the acetate 12 which was removed by K2CO3 in MeOH to give
a quantitative crude yield of 13. Methyl ether formation was
accomplished on crude 13 in the usual manner to furnish
methoxy derivative 14. At this stage the regioisomeric impurity
in 11 obtained by opening the cyclic sulfate was found to be
separable and the slightly lower yield of this step (86%) reflects
the removal of the undesired isomer.
It was now necessary to remove the PMB ether in 14.
Oxidative cleavage with CAN resulted in several products, thus
OPMB
OPMB
i
HO
11
AcO
Further studies are in progress to introduce into 18 the
24-membered polyene ring of viridenomycin.
TBDMSO
OBn
TBDMSO
OBn
Me
Me
We are grateful to the NIH for financial support of this
work.
12
13
iii
OH
OPMB
Notes and references
1 M. Nakagawa, K. Furihata, Y. Hayakawa and H. Seto, Tetrahedron
Lett., 1991, 32, 659.
iv
MeO
MeO
TBDMSO
OBn
2 T. Hasegawa, T. Kamiya, T. Henmi, H. Iwasaka and S. Yamatodani,
J. Antibiot., 1975, 28, 167.
TBDMSO
OBn
Me
Me
(+)-15
(+)-14
3 S. Omura, A. Nakagawa, K. Shibata and H. Sano, Tetrahedron Lett.,
1982 , 23, 4713; I. Umezawa, H. Takeshima, K. Komiyama, Y. Koh, H.
Yamamoto and M. Kawaguchi, J. Antibiot., 1981, 34, 259.
4 A. B. Smith, III, T. A. Rano, N. Chida, G. A. Sulikowski and J. L. Wood,
J. Am. Chem. Soc., 1992, 114, 8008.
v
O
5 A. I. Meyers and D. Romo, Tetrahedron, 1991, 47, 9503. A. I. Meyers
and G. P. Brengel, Chem. Commun., 1997, 1.
MeO
2
5
6 D. A. Sandham and A. I. Meyers, unpublished results.
7 A. Saito, A. Tanaka and T. Oritani, Tetrahedron: Asymmetry, 1996, 7,
2923.
8 Y. Gao and K. B. Sharpless, J. Am Chem. Soc., 1988, 110, 7538.
9 B. Lohray, Synthesis, 1992, 1035.
TBDMSO
OBn
Me
(+)-16
Scheme 3 Reagents and conditions: i, TBDMSCl, imidazole, DMAP
(91%); ii, K2CO3, MeOH (100%); iii, NaH (2.5 equiv.), MeI, 0 °C (86%);
iv, DDQ (1.3 equiv.), CH2Cl2–H2O (5+1) (93%); v, (COCl)2, DMSO, Et3N,
278 °C (91%).
10 C. Liotta and T. C. Caruso, Tetrahedron Lett., 1985, 26, 1599.
Communication 9/03160J
1372
Chem. Commun., 1999, 1371–1372