Synthesis of the antitumor alkaloid (+)-pancratistatin using the β- azidonation reaction via a prochiral 4-arylcyclohexanone derivative [14]

Full text of DOI:10.1021/ja980407s

J. Am. Chem. Soc. 1998, 120, 5341-5342
5341
Scheme 1a
Synthesis of the Antitumor Alkaloid
(+)-Pancratistatin Using the â-Azidonation Reaction
via a Prochiral 4-Arylcyclohexanone Derivative
Philip Magnus* and Iyassu K. Sebhat
Department of Chemistry and Biochemistry
UniVersity of Texas at Austin, Austin, Texas 78712
ReceiVed February 4, 1998
In 1984 Pettit and co-workers reported the structure of
pancratistatin 1, which was isolated from the roots of the Hawaiian
Pancratium littorale Jacq.1 Subsequently, pancratistatin 1 has
become an important target for total synthesis because of its
increasing potential as a clinically useful antitumor agent.2 The
supply of 1 is limited, and attempts to synthesize 1 from more
abundant alkaloids such as narciclasine have not been successful.3
There are four reported total syntheses of 1. The synthesis of
the racemate was first reported by Danishefsky,4 and three
enantioselective syntheses described by Hudlicky,5 Trost,6 and
Haseltine7 rely on enzymatic and catalytic chiral palladium
methodology to introduce the correct absolute stereochemistry.
While there is substantial literature describing the synthesis of
Amaryllidaceae alkaloids in general,8 the more highly function-
alized compounds have proven tenaciously difficult to synthesize
in an efficient and practical manner.9
The â-azido triisopropylsilyl (TIPS) enol ether functionalization
reaction provides a unique strategy for the synthesis of pancrat-
istatin 1 and Amaryllidaceae alkaloids in general.10 The 4-prochiral
arylcyclohexanone 211 was treated with lithium (+)-bis(R-
methylbenzyl)amide in THF containing lithium chloride, followed
a Ar ) 3-ClC6H4.
(1) Pettit, G. R.; Gaddamidi, V.; Cragg, G. M.; Herald, D. L.; Sagawa, Y.
J. Chem. Soc., Chem. Commun. 1984, 1693.
(2) Pettit, G. R.; Gaddamidi, V.; Herald, D. L.; Singh, S. B.; Cragg, G.
M.; Schmidt, J. M. J. Nat. Prod. 1986, 46, 995. Gabrielson, B.; Monath, T.
P.; Huggins, J. W.; Kirsi, J. J.; Hollingshead, M.; Shannon, W. M.; Pettit, G.
R. Natural Products as AntiViral Angents; Chu, C. K., Cutler, H. G., Ed.;
Plenum: New York, 1992; p 121.
(3) Pettit, G. R.; Melody, N.; O’Sullivan, M.; Thompson, M. A.; Herald,
D. L.; Coates, B. J. Chem. Soc., Chem. Commun. 1994, 2725.
(4) Danishefsky, S.; Lee, J. Y. J. Am. Chem. Soc. 1989, 111, 4829.
(5) Hudlicky, T.; Tian, X.; Ko¨nigsberger, K.; Maurya, R.; Rouden, J.; Fan,
B. J. Am. Chem. Soc. 1996, 118, 10752. Polt, R. Organic Synthesis: Theory
and Applications; Hudlicky, T., Ed.; JAI Press: Greenwich, CT, 1996; Vol
3, p 109.
(6) Trost, B. M.; Pulley, S. R. J. Am. Chem. Soc. 1995, 117, 10143.
(7) Doyle, T. J.; Hendrix, M.; VanDerveer, D.; Javanmard, S.; Haseltine,
J. Tetrahedron 1997, 53, 11153-11170.
(8) Martin, S. F. The Alkaloids; Brossi, A., Ed.; Academic Press: New
York, 1987; Vol 30, p 251.
(9) The synthesis of 7-deoxypancratistatin has been achieved by several
groups. Tian, X.; Maurya, R.; Ko¨nigsberger, K.; Hudlicky, T. Synlett 1995,
1125. Keck, G. E.; McHardy, S. F.; Murry, J. A. J. Am. Chem. Soc. 1995,
117, 7289. Chida, N.; Iitsuoka, M.; Yamamoto, Y.; Ohtsuka, M.; Ogawa, S.
Heterocycles 1996, 43, 1385. Paulsen, H.; Stubbe, M. Liebigs Ann. Chem.
1983, 535. Ohta, S.; Kimoto, S. Chem. Pharm. Bull. 1976, 24, 2977.
(10) Magnus, P.; Lacour, J.; Evans, P. A.; Roe, M. B.; Hulme, C. J. Am.
Chem. Soc. 1996, 118, 3406.
(11) Bromine-lithium exchange of the known bromide I (Shirasaka, T.;
Takuma, Y.; Imaki, N. Synth. Comm. 1990, 20, 1223), followed by addition
of II, and dehydration of the initial adduct gave III. Hydrogenation of III
and acid hydrolysis provided 2.
by TIPSOTf to give 3 in 95% yield with an ee of g85%.12
Treatment of 3 with (PhIO)n/TMSN3 in CH2Cl2 at -15 °C rapidly
produced 4 (95%) as a mixture of trans- and cis-diastereomers
in a 3.5:1 ratio, Scheme 1. Exposure of the mixture to LiBPh4
did not improve the ratio by equilibration via a putative enonium
ion but led to decomposition and elimination to dienes.10
Consequently, while the trans-/cis-ratio of 4 could not be
improved, the yield of the required trans-4 is approximately 75%.
At this stage the stereoisomers could not be separated. Reduction
of 4 using LiAlH4/Et2O, followed by treatment with ClCO2Me/
pyridine gave 5, as a mixture of trans-/cis-diastereomers. On a
large scale (>6 g) two crystallizations were sufficient to provide
pure 5 (56% from 4).
It was anticipated that epoxidation of 5 would proceed by axial
addition, and eventually, after a series of intermediate steps, form
6.13 Indeed, treatment of 5 with m-chloroperoxybenzoic acid/
CH2Cl2/imidazole gave 6 in excellent yield. Mild acid hydrolysis
of 6 gave 7, which on treatment with KOBut/HMPA at 90 °C
resulted in complete conversion into 8 (91%).
At this stage it was necessary to convert 8 into the derived
R,â-unsaturated ketone 11. This proved to be extremely difficult
(12) At the present time there is no predictive model that suggests a
particular base and ketone will result in a specific chiral enolate. Therefore,
predictions are based upon comparisons with experimental data for 4-substi-
tuted cyclohexanones and must be taken with caution. Simpkins, N. S. Pure
Appl. Chem. 1996, 68, 691-694. Cox, P. J.; Simpkins, N. S. Tetrahedron
Asymm. 1991, 2, 1-26. Bunn, B. J.; Simpkins, N. S. J. Org. Chem. 1993, 58,
533-534. Yamashita, T.; Sato, D.; Kiyoto, T.; Kumar, A.; Koga, K.
Tetrahedron Lett. 1996, 37, 9195-9198.
(13) Magnus, P.; Mugrage, B. J. Am. Chem. Soc. 1990, 112, 462. Magnus,
P.; Lacour, J.; Coldham, I.; Mugrage, B.; Bauta, B. Tetrahedron 1995, 51,
11087.
S0002-7863(98)00407-7 CCC: $15.00 © 1998 American Chemical Society
Published on Web 05/13/1998

5342 J. Am. Chem. Soc., Vol. 120, No. 21, 1998
Communications to the Editor
Scheme 2a
described by Hudlicky5 gave 14, which on acetylation produced
15 (60% from 13).
Treatment of 15 under modified Bischler-Napieralski reaction
conditions (Tf2O/DMAP)15 followed by acid hydrolysis gave 16
(60%) along with a small amount of the product of electrophilic
substitution ortho- to the methylenedioxy group 16a (7:1).16 While
16 and 16a could not be separated at this stage, it was found that
exposure of the mixture to BBr3/CH2Cl2/-78 to 0 °C gave 17
(73%, 16a did not react), which was readily separated from 16a.
Finally, removal of the acetate protecting groups from 17 with
NaOMe/MeOH proceeded cleanly to give pancratistatin 1 (87%),
which was identical with an authentic sample kindly supplied by
Professor Tomas Hudlicky. The optical rotation (+38) closely
matched the literature values (lit. values of +40.9, +48, and +44
see refs 5, 1, and 6, respectively), thus corroborating the emperical
observations of Simpkins and Koga12 that the choice of lithium
(+)-bis(R-methylbenzyl)amide would result in the correct absolute
configuration of 3 necessary to synthesize (+)-1.
Acknowledgment. The National Institutes of Health (GM), The
National Science Foundation, The Robert A. Welch Foundation, and
Merck Research Laboratories are thanked for financial support. Professor
Tomas Hudlicky is thanked for a sample of 1 and spectral data.
Supporting Information Available: Complete spectral information
for compounds 1-17 (33 pages, PDF/print). See any current masthead
page for ordering information and Web access instructions.
JA980407S
(14) The formation of 13c is readily explained by acetoxonium ion 13b,
and since the acetic acid was wet, an ortho ester intermediate leads to 13c.
a Ar ) 3-ClC6H4.
to do in an efficient manner, but eventually it was found that
treatment of 8 with TMSOTf/Et3N gave the bis-TMS adduct 9.
Exposure of 9 to PhSeOCOCF3 gave the selenide derivative 10,
which was oxidized (H2O2/py) and eliminated to give 11 (85%
overall). Treatment of 11 under the standard R,â-enone epoxi-
dation conditions of H2O2/NaOH/MeOH gave 12 in low yield
(25%). It was found that 12 was competitively destroyed by the
strongly alkaline reaction conditions. Changing the reaction
conditions to H2O2/NaHCO3/MeOH/THF gave 12 (75%). Reduc-
tion of 12 with L-selectride/THF gave 13 (63% from 11) as a
single diastereomer. The diacetate derivative of 13 was character-
ized by X-ray crystallography. Interestingly, solvolysis of this
diacetate in AcOH/Ac2O/AcONa proceeded with inversion at C2,
presumably through acetoxonium ion participation.14 Treatment
of 13 with PhCO2Na/H2O under the solvolytic conditions
(15) Banwell, M. G.; Cowden, C. J.; Gable, R. W. J. Chem. Soc., Perkin
Trans 1 1994, 3515.
(16) The Bischler-Napieralski reaction was not regiospecific and produced
about 10% of 16a. The presence of the C1 acetoxy group appears to improve
this ratio, since in related experiments lacking this functionality the ratio of
the corresponding two lactams was 3:1.