An enantiospecific synthesis of the potent immunosuppressant FR901483

Article abstract of DOI:10.1002/1521-3773(20001215)39:24<4593::AID-ANIE4593>3.0.CO;2-X

An azaspiroannulation and an uncommon Mitsunobu reaction are the key steps in a convergent, enantioselective synthesis of the title compound 1. This potent immunosuppressant is derived from a fungus and is distinguished by an unusual monophosphate ester g

Full text of DOI:10.1002/1521-3773(20001215)39:24<4593::AID-ANIE4593>3.0.CO;2-X

COMMUNICATIONS
Table 2. Calculated CGaussian 94, RHF, LANL2DZ) geometric parameters and charges CMulliken) of C6F5Xe-Z molecules[a]
.
C6F5Xe-Z
Molecule
Sym.
Selected geometric parameters[b]
Selected Mulliken charges
CC1)-Xe
Xe-Z
CC2)-CC1)-CC6)
Xe
C6F5
CC1)
Z
C6F5Xe-F
C6F5Xe-C6F5
C6F5Xe-CN
1
2
3
Cs
C1
Cs
2.20
2.34
2.24
2.13
2.34
2.38
117.7
117.4
118.1
1.148
0.980
0.967
À 0.415
À 0.490
À 0.403
À 1.001
À 0.687
À 0.853
À 0.733
À 0.490
À 0.564[c]
for comparison see ref [12]
C6F5Xe ´´´ F-AsF5
ꢁ Cs
2.12
2.16
±
2.56
±
2.03
121.3
122.7
±
1.083
0.886
1.306
À 0.134
0.114
±
À 1.032
À 0.870
±
À 0.948
±
À 0.653
‡
[C6F5Xe]
Cs
FXeF
D/h
[a] Z ˆ second ligand bound to XeII; [b] in [Š] or [8], repectively; [c] Mulliken charges of C in the CN ligand: À 0.466.
[4] H.-J. Frohn, A. Klose, G. Henkel, Angew. Chem. 1993, 105, 114;
Angew. Chem. Int. Ed. Engl. 1993, 32, 99.
[5] H.-J. Frohn, T. Schroer, Angew. Chem. 1999, 111, 2751; Angew. Chem.
Int. Ed. 1999, 38, 2554.
[6] K. Akiba, Chemistry of Hypervalent Compounds, Wiley-VCH, New
York, 1998.
[7] L. Turbini, R. Aikman, R. Lagow, J. Am. Chem. Soc. 1979, 101, 5833.
[8] H.-J. Frohn, M. Theiûen, Abstr. 2P-20 16th Int. Symp. Fluorine Chem.
CDurham, UK), 2000.
[9] H.-J. Frohn, T. Schroer, M. Theiûen, Abstr. B37 12th Europ. Symp.
Fluorine Chem. CBerlin, Germany), 1998; M. Theiûen, H.-J. Frohn,
Abstr. 24 8. Dt. Fluortagung CSchmitten, Germany), 1998; H.-J. Frohn,
A. Klose, V. V. Bardin, A. J. Kruppa, T. V. Leshina, J. Fluorine Chem.
1995, 70, 147.
[10] V. V. Bardin, I. V. Stennikova, G. G. Furin, T. V. Leshina, G. G.
Yakobson, J. Gen. Chem. USSR 1988, 58, 2297; A. P. Lothian, C. A.
Ramsden, Synlett 1993, 753; H.-J. Frohn, V. V. Bardin, J. Organomet.
Chem. 1995, 501, 155.
The results of ab initio calculations for 1 ± 3 show the
following sequence of C-Xe distances: 2 > 3 > 1, thus opposite
to the sequence of Mulliken charges of the ligand Z in
C6F5XeZ CTable 2). The comparison of data for 1 and
C6F5Xe ´´´ FAsF5 [12] elucidates clearly the change when going
from the asymmetric hypervalent C-Xe-F bond to a signifi-
cant C-Xe ´´´ F contact: the negative charge on Z gets closer to
À1 whereas the negative charge of the C6F5 group decreases
significantly. The high anionic character of the C6F5 group in
1 ± 3 agrees withthe observed reactivities towards electro-
philes and explains the lower-frequency chemical shifts of the
‡
p-F atom compared to [C6F5Xe] .
Experimental Section
1: A cold solution of [NMe4]F C25 mg, 0.27 mmol) in CH2Cl2 C1 mL) was
added to a suspension of [C6F5Xe][AsF6] C131 mg, 0.27 mmol) in CH2Cl2
C1.5 mL) at À788C in an 8 mm FEP trap. The suspension was stirred over
2 days at À788C until all the fluoride was consumed. The mother liquor was
separated from solid [NMe4][AsF6] and the quantity of 1 was determined
C19F NMR): 0.19 mmol, 70%. The other reactions were usually performed
directly withthe cold solutions of 1. By evaporating CH2Cl2 at 10À2 hPa/ ꢀ
À 558C and later drying at ꢀ À 408C 1 was obtained as a colorless solid,
which after warming to 208C decomposed totally within 4 h. In CH2Cl2
solution noticeable decomposition proceeded above À308C withteh
formation of C6F5H and traces of C6F5Cl.
[11] K. O. Christe, E. C. Curtis, D. A. Dixon, H. P. Mercier, J. C. P. Sanders,
G. J. Schrobilgen, J. Am. Chem. Soc. 1991, 113, 3351.
[12] H.-J. Frohn, A. Klose, T. Schroer, G. Henkel, V. Buû, D. Opitz, R.
Vahrenhorst, Inorg. Chem. 1998, 37, 4884.
An Enantiospecific Synthesis of the Potent
Immunosuppressant FR901483**
2: A cold solution of CdCC6F5)2 C17 mg, 0.04 mmol) in CH2Cl2 C0.5 mL) was
added to a solution of 1 C0.08 mmol) in CH2Cl2 C1.5 mL) at À788C. After
5 min of stirring CdF2 precipitated. The reaction was complete C19F NMR)
after further 10 min. The mother liquor was collected. In addition to 2
C0.06 mmol, 75%) the solution contained C6F5H C4 mmol) and CC6F5)2
C2 mmol). The isolation of solid 2 was achieved as described for 1. Solid 2
decomposes completely at room temperature within 1 h and in CH2Cl2
solution at À408C within 9 h [C6F5H:CC6F5)2 ˆ 1:0.1].
Goetz Scheffler, Hirofumi Seike, and Erik J. Sorensen*
Bond formations induced by phenol oxidations have a rich
history in organic chemistry. The influential two-step synthesis
of usnic acid by Sir Derek Barton and co-workers[1] followed a
set of simple rules that provided guidelines for rationalizing
the course of oxidative phenolic radical couplings occurring in
the biogeneses of a number of natural products.[2] We were
3: A cold solution of Me3SiCN C14 mL, 0.10 mmol; or the labeled 13CN and
C15N derivatives) in CH2Cl2 C0.5 mL) was added to a solution of 1
C0.10 mmol) and CH2Cl2 C1.5 mL) at À788C and stirred. After 5 min the
reaction was complete C19F NMR: quantitative reaction) giving a 1:1
mixture of 3 and Me3SiF. CH2Cl2 and Me3SiF were distilled at ꢀ À 558C/
10À2 hPa and the white solid product was dried at ꢀ À 408C/10À2 hPa. The
solid spontaneously decomposed during the rapid warming to room
temperature. CH2Cl2 solutions of 3 decomposed completely at À408C
within 2 h with the formation of C6F5CN and C6F5H C4:1).
[*] Prof. Dr. E. J. Sorensen, Dr. G. Scheffler, H. Seike
Department of Chemistry and
The Skaggs Institute for Chemical Biology
The Scripps Research Institute
10550 NorthTorrey Pines Road, La Jolla, CA 92037 CUSA)
Fax : C‡1)858-784-2798
Checking the purity of the cold solid products after dissolution in CH2Cl2 at
À788C showed in the case of 1 and 2 degrees of decomposition of up to
10% and for the thermally more sensitive product 3 up to 30%.
E-mail: sorensen@scripps.edu
[**] Dr. G. Scheffler and H. Seike contributed equally to this work. We
thank Dr. L. B. Pasternack and Dr. D. H. Huang for NMR spectro-
scopic assistance, and Dr. G. Siuzdak for mass spectrometric
assistance. We also thank Dr. S. Takase and Dr. H. Setoi from the
Fujisawa Pharmaceutical Co. for physical data for FR901483. This
work was generously supported by The Skaggs Institute for Chemical
Biology, a grant from the Merck Research Laboratories, a Beckman
Young Investigator Award to E.J.S., and a predoctoral fellowship from
the Heiwa Nakajima Foundation to H.S.
Received: August 18, 2000 [Z15656]
[1] D. Naumann, W. Tyrra, J. Chem. Soc. Chem. Commun. 1989, 47; H.-J.
Frohn, S. Jakobs, J. Chem. Soc. Chem. Commun. 1989, 625.
[2] H.-J. Frohn, V. V. Bardin, J. Chem. Soc. Chem. Commun. 1993, 1072.
[3] V. V. Zhdankin, P. J. Stang, N. S. Zefirov, J. Chem. Soc. Chem.
Commun. 1992, 578.
Angew. Chem. Int. Ed. 2000, 39, No. 24
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3924-4593 $ 17.50+.50/0
4593

COMMUNICATIONS
drawn to the possibility
that the azaspiro[4.5]de-
cane nucleus of the power-
by hypervalent iodine-based oxidants. To construct the
remaining ring and C-6 stereocenter of 1, we hoped to utilize
an intramolecular aldol reaction Csee 4 !5).[10] Despite
concerns about diastereoselectivity issues and potential
epimerization problems, this process was deemed well-suited
for the task of completing the architecture of 1.[4] A synthesis
of 5 would then allow us to explore an uncommon strategy for
introducing the conspicuous phosphate ester of 1 Csee below).
A laboratory synthesis of 1 based on these concepts was
achieved as described below.
By straightforward reaction sequences, aldehyde 6 and
amine 7 CScheme 2) were created from commercially avail-
able N-Boc-l-tyrosine[11] and subsequently joined under
reducing conditions to give 2 CPG ˆ pNO2C6H4SO2).[12] After
muchexperimentation, we found that exposure of a solution
of 2 in 1,1,1,3,3,3-hexafluoro-2-propanol to iodobenzene
diacetate at room temperature resulted in the formation of
azaspiro[4.5]decadienone 3 in 51% yield Cbased on 70%
consumed 2). The outcome of this ring closure, which
established a significant portion of the structure of 1, is
noteworthy because analogous amino phenols are converted
into fused ring systems under similar reaction conditions.[9c]
At this stage, it was necessary to replace the 4-nitrobenzene-
sulfonamide protecting group[13] in 3 witha more compatible
tert-butoxycarbonyl group C3 !8, Scheme 2). Treatment of a
solution of 8 in EtOAc/EtOH withRaney nickel under an
atmosphere of hydrogen caused a complete reduction of the
dienone moiety and afforded a mixture of epimeric alcohols
shown as 9. A reduction of the methyl ester with lithium
aluminum hydride then produced a mixture of diols that could
subsequently be converted by Swern oxidation into a single
ketoaldehyde 4 in good overall yield.[14, 15]
OCH3
ful
immunosuppressant
H3C
H
N
HO
FR901483 C1)[3] may form
H
N
H
during the course of an
H
oxidation of
a tyrosine
OH
OH
H
O
P
phenol moiety. This fun-
gal-derived natural prod-
uct comprises a novel tri-
cyclic core framework and
a salient monophosphate
O
1: FR901483
ester. At the outset, we noted that the intriguing structure
of 1 seemed to conceal most of the atoms of two molecules of
tyrosine and surmised that a laboratory synthesis of this
substance could be facilitated by oxidative azaspiroannulation
and intramolecular aldol addition reactions.[4] This general
analysis was guided by the supposition that such reaction
processes may have counterparts in the biosynthesis of this
natural product. Our goal was to evaluate the feasibility of
these reactions in the course of a synthesis of 1. In this
communication, we describe our convergent, enantiospecific
synthesis of this promising immunosuppressant from l-
tyrosine, which was executed along these lines.[5]
À
We anticipated that the N C3 bond of an amino phenol 2
CScheme 1) could be established in the course of a reductive
coupling of two derivatives of l-tyrosine. The cornerstone of
our strategy is the idea that exposure of 2 to an appropriate
oxidant, suchas iodobenzene diacetate, [6] could give rise to 3.
This appealing ring-closure would create the azaspiro[4.5]de-
cane substructure of 1 and would not be complicated by the
formation of diastereoisomers.[7±9] Our goal was to effect
intramolecular carbon ± nitrogen bond formations withsub-
strates possessing the natural oxidation state at C-3, and we
maintained our commitment to this plan even in the face of a
potentially problematic oxidation of the secondary amine of 2
À
Our intention was to create the C6 C7 bond and the
remaining ring of 1 in the course of an aldol cyclization,
although we were aware that this objective could be difficult
to achieve owing to the various sites at which compound 4
could undergo deprotonation in the
presence of bases. Nevertheless, we
were mindful of the work of Snider
and co-workers[4] and could convert 4
into the desired compound 5 in 34%
yield under the conditions shown.[16]
When 4 was dissolved in THF and
treated withpyrrolidine C0.3 equiv)
and AcOH C1.2 equiv) at 08C !RT,
the equatorial C-6 epimer of 5 was
formed in 51% yield.[17] Despite the
modest yield for the methoxide-in-
duced aldol cyclization of 4, we could
procure significant quantities of hy-
droxyketone 5.
Owing to its rigid and concave
architecture, hydroxyketone
5 was
converted into diol 10 in a completely
diastereoselective fashion by hydro-
genation. To achieve a synthesis of 1
from 10, inversion of stereochemistry
at C-8 and production of a C-8 mono-
phosphate ester are required. Gratify-
Scheme 1. A strategy for the synthesis of FR901483 C1) featuring oxidative azaspiroannulation and
aldol cyclization reactions. PG ˆ protecting group.
4594
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3924-4594 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2000, 39, No. 24

COMMUNICATIONS
the presence of a free C-6 hydroxyl
group,[20] and directly produces the
natural stereochemistry and the
phosphate ester moiety at C-8. Hy-
drogenolysis of the dibenzylphos-
phate of 11 followed by cleavage of
the Boc group afforded FR901483
C1) as its hydrochloride salt.
OCH3
OCH3
CH3
N
CH3
N
O
H
N
H
3
CO2CH3
H
a
PG
PG
+
H
H2N
CO2CH3
OH
OH
6
7
2
PG = pNO2C6H4SO2
PG = pNO2C6H4SO2
Received: August 21, 2000 [Z15675]
b
OCH3
OCH3
[1] D. H. R. Barton, A. M. Deflorin, O. E.
Edwards, J. Chem. Soc. 1956, 530 ± 534.
[2] For selected reviews, see: a) D. H. R.
Barton, T. Cohen, Festschr. Arthur Stoll
1957, 117 ± 143; b) D. H. R. Barton,
Chem. Br. 1967, 330 ± 337; c) A. I. Scott,
Q. Rev. Chem. Soc. 1965, 19, 1 ± 35;
c) T. Kametani, K. Fukumoto, Synthe-
sis 1972, 657 ± 674; d) D. H. R. Barton,
Half A Century of Free Radical Chem-
istry, Cambridge University Press, New
York, 1993, pp. 7 ± 20.
H
H
3
H3C
H3C
N
CO2CH3
OH
N
CO2CH3
d
N
N
PG
PG
O
9
3: PG = pNO2C6H4SO2
8: PG = Boc
c
PG = Boc
e, f
[3] K. Sakamoto, E. Tsujii, F. Abe, T.
Nakanishi, M. Yamashita, N. Shigemat-
su, S. Izumi, M. Okuhara, J. Antiobiot.
1996, 49, 37 ± 44.
OCH3
OCH3
5
5
[4] After we initiated our studies, Snider
and co-workers demonstrated the fea-
sibility of an intramolecular aldol addi-
tion in the course of a synthesis of
demethylamino-FR901483 Csee B. B.
Snider, H. Lin, B. M. Foxman, J. Org.
Chem. 1998, 63, 6442 ± 6443). In the
following year, Snider and Lin reported
the first synthesis and absolute stereo-
chemistry of FR901483 C1) Csee B. B.
Snider, H. Lin, J. Am. Chem. Soc. 1999,
121, 7778 ± 7786). The authors of this
pioneering report also postulated that
an oxidative cyclization of a methylated
tyrosine ± tyrosine dipeptide may be a
key step in the biosynthesis of 1.
H3C
PG
H
H3C
PG
H
O
N
O
N
H
g
H
N
N
HO
H
6
9
7
O
4
5
PG = Boc
PG = Boc
h
OCH3
OCH3
H3C
H
H
H3C
PG
H
N
HO
N
HO
H
j, k
i
N
N
1•HCl
PG
8
[5] For an informative review of enantio-
specific syntheses of heterocycles from
a-amino acids, see: F. J. Sardina, H.
Rapoport, Chem. Rev. 1996, 96, 1825 ±
1872.
H
OH
OBn
OBn
H
H
O
H
P
O
11
10
PG = Boc
PG = Boc
[6] For reviews on the use of hypervalent
iodine reagents in organic synthesis,
see: a) T. Kitamura, Y. Fujiwara, Org.
Prep. Proced. Int. 1997, 29, 409 ± 458;
b) A. Varvoglis, Tetrahedron 1997, 53,
1179 ± 1255; c) R. M. Moriarty, O. Pra-
kash, Adv. Heterocycl. Chem. 1998, 69,
1 ± 87; d) T. Wirth, U. H. Hirt, Synthesis
1999, 1271 ± 1287.
Scheme 2. Synthesis of FR901483 C1) ´ HCl: a) 6 C1.0 equiv), 7 C1.5 equiv), NaBHCOAc)3 C1.5 equiv), 4 Š
MS, 08C, 1 h, then RT, 24 h, 80%; b) PhICOAc)2, C1.1 equiv), CCF3)2CHOH, RT, 10 min, 51% Cbased on
70% consumed 2); c) NaSPhC1.5 equiv), DMSO, RT, 3 h; then CBoc) 2O C2 equiv), pyr C2 equiv), RT, 12 h,
71% from 3; d) H2, Raney Ni, EtOAc/EtOH C2:1), RT, 15 h, 93%; e) LiAlH4 C3 equiv), THF, À788C, 1 h
then 08C, 2 h, 90%; f) CCOCl)2 C8 equiv), DMSO C16 equiv), CH2Cl2, À788C, 30 min; then iPr2NEt
C20 equiv), À788C, 30 min; then 08C, 1 h, ꢁ100%; g) NaOMe C1.1 equiv), MeOH, 08C, 2 h, 34% of 5 C74%
total yield of aldol adducts); h) H2, Raney Ni, EtOAc/EtOH C1:10), RT, 4 h, 92%; i) dibenzyl hydrogen
phosphate C6 equiv), trisC4-chlorophenyl)phosphine C3 equiv), DIAD C3 equiv), Et3N C10 equiv), toluene,
RT, 26 h, 37 % of 11‡17% of alkene; j) H2, Pd/C, MeOH, RT, 5 h, 95%; k) 4n HCl, dioxane, 08C, 30 min;
then RT, 2 h, 84% of 1 ´ HCl Cprecipitated from MeOH withEt 2O). MS ˆ molecular sieves, RTˆ room
temperature, Boc ˆ tert-butoxycarbonyl, pyrˆ pyridine, DIAD ˆ diisopropyl azodicarboxylate, Bnˆ benzyl.
[7] A cyclization of a compound of type 2
into
3 would produce a chirotopic,
nonstereogenic center Csee K. Mislow,
J. Siegel, J. Am. Chem. Soc. 1984, 106,
3319 ± 3328).
[8] Kita and co-workers pioneered oxidations of phenolic amides with
hypervalent iodine reagents. See: a) Y. Tamura, T. Yakura, J.-i.
Haruta, Y. Kita, J. Org. Chem. 1987, 52, 3927 ± 3930; b) Y. Kita, T.
Yakura, H. Tohma, K. Kikuchi, Y. Tamura, Tetrahedron Lett. 1989, 30,
1119 ± 1120; c) Y. Kita, H. Tohma, K. Kikuchi, M. Inagaki, T. Yakura,
J. Org. Chem. 1991, 56, 435 ± 438.
ingly, it was possible to accomplishbothobjectives in a single
step by a Mitsunobu reaction between 10 and dibenzyl
hydrogen phosphate C10 !11).[18, 19] Despite its modest yield,
this type of reaction is, to our knowledge, unprecedented in
complex natural-product total synthesis, can be achieved in
Angew. Chem. Int. Ed. 2000, 39, No. 24
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3924-4595 $ 17.50+.50/0
4595

COMMUNICATIONS
[9] As we were attempting to achieve oxidative azaspiroannulations,
Ciufolini and co-workers described novel oxidative cyclizations of
oxazolines produced from tyrosine ± tyrosine dipeptides, which thus
demonstrated for the first time that oxidative azaspiroannulations are
viable reaction processes. See: a) N. A. Braun, M. A. Ciufolini, K.
Peters, E.-M. Peters, Tetrahedron Lett. 1998, 39, 4667 ± 4670; b) N. A.
Braun, J. D. Bray, M. A. Ciufolini, Tetrahedron Lett. 1999, 40, 4985 ±
4988; c) N. A. Braun, M. Ousmer, J. D. Bray, D. Bouchu, K. Peters, E.-
M. Peters, M. A. Ciufolini, J. Org. Chem. 2000, 65, 4397 ± 4408. This
provocative reaction type is related to several interesting oxidative
spiroannulations. For examples, see: d) A. V. R. Rao, M. K. Gurjar,
P. A. Sharma, Tetrahedron Lett. 1991, 32, 6613 ± 6616; e) P. Wipf, Y.
Kim, Tetrahedron Lett. 1992, 33, 5477 ± 5480; f) P. Wipf, Y. Kim, P. C.
Fritch, J. Org. Chem. 1993, 58, 7195 ± 7203; g) P. Wipf, Y. Kim, D. M.
Goldstein, J. Am. Chem. Soc. 1995, 117, 11106 ± 11112; h) A.
McKillop, L. McLaren, R. J. K. Taylor, R. J. Watson, N. Lewis, Synlett
1992, 201 ± 203; i) A. McKillop, L. McLaren, R. J. K. Taylor, R. J.
Watson, N. J. Lewis, J. Chem. Soc. Perkin Trans. 1 1996, 1385 ± 1393;
j) K. Marshall Aubart, C. H. Heathcock, J. Org. Chem. 1999, 64, 16 ±
22; k) D. Yang, M.-K. Wong, Z. Yan, J. Org. Chem. 2000, 65, 4179 ±
4184.
[10] For applications of the aldol reaction in syntheses of bicyclo[3.3.1]-
nonanes and azabicyclo[3.3.1]nonanes, see: a) W. A. Kinney, G. D.
Crouse, L. A. Paquette, J. Org. Chem. 1983, 48, 4986 ± 5000; b) R. J. K.
Taylor, S. M. Turner, D. C. Horwell, O. W. Howarth, M. F. Mahon,
K. C. Molloy, J. Chem. Soc. Perkin Trans. 1 1990, 2145 ± 2150; c) H.-J.
Teuber, C. Tsaklakidis, J. W. Bats, LiebigsAnn. Chem . 1990, 781 ± 787;
d) S. Patir, P. Rosenmund, P. H. Götz, Heterocycles 1996, 43, 15 ± 22.
[11] Aldehyde 6 was prepared from the known amino ester 7 by the
following reaction sequence: 1) 4-nitrobenzenesulfonyl chloride, iPr2-
NEt, CH2Cl2, RT, 0.5 h, 88%; 2) MeI, K2CO3, DMF, RT, 1 h, 91%;
3) AlCl3, EtSH, CH2Cl2, RT, 2.5 h, 92%; 4) BH3 ´ THF, THF, 08C, 1 h;
then RT, 2 h, 94%; 5) SO3 ´ pyr, DMSO, iPr2NEt, RT, 10 min,
approximately 100% C6 was used in crude form).
Total Synthesis of $À)-Spirotryprostatin B and
Three Stereoisomers**
Larry E. Overman* and Mark D. Rosen
Small-molecule natural products play an important role in
contemporary studies to understand and control cellular
proliferation.[1] From the fermentation broth of the fungus
Aspergillus fumigatus, Osada and co-workers recently iden-
tified a group of novel diketopiperazine alkaloids that inhibit
G2/M phase progression of the mammalian cell cycle at
micromolar concentrations.[2, 3] Spirotryprostatins A C1) and B
C2) are the most complex of these alkaloids,[2] all of which
appear to arise biosynthetically by prenylation of a diketo-
piperazine derived from tryptophan and proline.
O
O
N
N
H
H
H
N
N
O
O
MeO
N
O
N
O
H
H
Spirotryprostatin A (1)
Spirotryprostatin B (2)
The novelty of their structures and the potential utility of
cell-cycle inhibitors make the spirotryprostatins attractive
targets for total synthesis. If such an undertaking were to be
stereocontrolled, a central challenge would be to relate the
stereochemistry of the quaternary spiro carbon to the
adjacent stereocenter bearing the 2-methylpropenyl side
chain. In 1998, Danishefsky and co-workers reported the
total synthesis of spirotryprostatin A C1), which constituted
the first total synthesis in this area.[4] Earlier this year, the
groups of Williams,[5] Danishefsky,[6] and Ganesan[7] disclosed
inaugural total syntheses of CÀ)-spirotryprostatin B C2).
Our approachto spirotryprostatin B C 2) and congeners is
distinctly different from the previous syntheses CScheme 1).
The logic of our strategy is to correlate the relative config-
urations of C3 and C18 in 2[8] to the geometry of the internal
double bond of a triene cyclization substrate 4 by capitalizing
on the stereochemical selectivity of two palladium-catalyzed
[12] A. F. Abdel-Magid, K. G. Carson, B. D. Harris, C. A. Maryanoff, R. D.
Shah, J. Org. Chem. 1996, 61, 3849 ± 3862.
[13] T. Fukuyama, C.-K. Jow, M. Cheung, Tetrahedron Lett. 1995, 36,
6373 ± 6374.
[14] A. J. Mancuso, D. Swern, Synthesis 1981, 165 ± 185.
[15] Aldehyde 4 was used in crude form because partial epimerization at
C-5 occurred during chromatography on silica. For stimulating
discussions of the stability and chemistry of optically active a-amino
aldehydes, see: A. G. Myers, D. W. Kung, B. Zhong, J. Am. Chem. Soc.
2000, 122, 3236 ± 3237.
[16] Compound 5 was the major component of a mixture of three aldol
cyclization products. The equatorial C-6 epimer of 5 was formed in
approximately 19% yield and a compound tentatively assigned as a
C6 ± C9 aldol adduct was obtained in approximately 21% yield.
[17] This aldol procedure affords 5 in 17% yield and a substance that we
tentatively assigned as the C6 ± C9 aldol adduct in 12% yield Csee
ref. [16]). We have not yet been able to utilize the C-6 epimer of 5 in
our synthesis.
[18] a) O. Mitsunobu, K. Kato, J. Kimura, J. Am. Chem. Soc. 1969, 91,
6510 ± 6511; b) D. A. Campbell, J. C. Bermak, J. Org. Chem. 1994, 59,
658 ± 660.
[19] The results of COSY, HMQC, and ROESY NMR spectroscopy
experiments were fully consistent withthe structure of compound 11.
In addition to 11, a C8 ± C9 alkene was formed in 17% yield due to
elimination.
[*] Prof. L. E. Overman, M. D. Rosen
Department of Chemistry
University of California, Irvine
516 Rowland Hall, Irvine, CA 92697-2025 CUSA)
Fax : C‡1)949-824-3866
[20] Mitsunobu reactions are sensitive to steric environments; therefore,
site-selective modification of polyols is feasible. For examples, see:
a) D. A. Evans, J. R. Gage, J. L. Leighton, J. Am. Chem. Soc. 1992,
114, 9434 ± 9453; b) R. S. Coleman, J. R. Fraser, J. Org. Chem. 1993,
58, 385 ± 392.
E-mail: leoverma@uci.edu
[**] We thank NIH NIGMS CGM-30859) for financial support, Prof. A. J.
Shaka and N. D. Taylor for DPFGSE experiments, Prof. H. Osada for
providing spectral data for natural 1, and Prof. R. M. Williams and
P. R. Sebahar for providing spectral data and a sample of synthetic ent-
21. We are also grateful to Prof. R. M. Williams and Prof. S. J.
Danishefsky for open exchange of information prior to publication.
NMR and mass spectra were determined at UC Irvine withinstru-
ments purchased with the assistance of the NSF and NIH shared
instrumentation programs.
Supporting information for this article is available on the WWW under
http://www.wiley-vch.de/home/angewandte/ or from the author.
4596
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3924-4596 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2000, 39, No. 24