An efficient synthesis of the C-23 deoxy, 17α-hydroxy South 1 hemisphere and its cephalostatin 1 analog

Article abstract of DOI:10.1021/ol034894e

(Matrix presented) Methods for functionalization of the D ring of diene 1 were investigated. This study led to efficient syntheses of 3 and the subsequent 23′-deoxy,17′α-hydroxy cephalostatin 1 analog (4). The bioactivity data of 2 and 4 are in the low na

Full text of DOI:10.1021/ol034894e

ORGANIC
LETTERS
2003
Vol. 5, No. 16
2849-2852
An Efficient Synthesis of the C-23
Deoxy, 17r-Hydroxy South 1
Hemisphere and Its Cephalostatin 1
Analog1
Wei Li and Philip L. Fuchs*
Department of Chemistry, Purdue UniVersity, West Lafayette, Indiana 47907
pfuchs@purdue.edu
Received May 21, 2003
ABSTRACT
Methods for functionalization of the D ring of diene 1 were investigated. This study led to efficient syntheses of 3 and the subsequent
23′-deoxy,17′r-hydroxy cephalostatin 1 analog (4). The bioactivity data of 2 and 4 are in the low nanomolar range for a 10-cell-line minipanel.
The cephalostatins2 and ritterazines3 comprise a family of
45 structurally unique marine natural products that display
extreme cytotoxicity against human tumors (∼1 nM mean
GI50’s in the 2-day NCI-60 screen and 10-14 M GI50’s in
3-day tests in the Purdue minipanel).4 We have reported the
total syntheses of cephalostatin 1 and cephalostatin 7, as well
as many analogs,5,6 but chemical evidence for the site(s) of
reactivity and the mechanism of action of the bissteroidal
pyrazines remain unknown and no scalable synthesis for such
testing has been achieved. As more extensively discussed
in our previous publication,7 we are now pursuing a strategy
that retains all 27 carbon atoms of hecogenin acetate and
employs site-specific oxidation reactions to introduce the
common features found in the cephalostatin targets, including
23-deoxy South 1 and the derived 23′-deoxy cephalostatin
1 analog 2 (Scheme 1). We envisioned that diene 1 could
be a useful intermediate for D-ring-functionalized South 1
Scheme 1
(1) Cephalostatin studies 26. Oxidations 4.
(2) Pettit, G. R.; Tan, R.; Xu, J.-P.; Ichihara, Y.; Williams, M. D.; Boyd,
M. R. J. Nat. Prod. 1998, 61, 953 and references therein.
(3) Fukuzawa, S.; Matsunaga, S.; Fusetani, N. J. Org. Chem. 1997, 62,
4484 and references therein.
(4) LaCour, T. G.; Guo, C.; Ma, S.; Jeong, J. U.; Boyd, M. R.; Matsunaga,
S.; Fusetani, N.; Fuchs, P. L. Bioorg. Med. Chem. Lett. 1999, 9, 2587 and
references therein. Leukemia, renal and CNS lines are particularly sensitive
to cephalostain 1.
(5) (a) Heathcock, C. H.; Smith, S. C. J. Org. Chem. 1994, 59, 6828
and references therein. (b) Jautelat, R.; Mu¨ller-Fahrnow, A.; Winterfeldt,
E. Chem. Eur. J. 1999, 5, 1226. HelV. Chim. Acta 2000, 83, 1854 and
references therein.
(6) LaCour, T. G.; Guo, C.; Boyd, M. R.; Fuchs, P. L. Org Lett. 2000,
2, 33.
10.1021/ol034894e CCC: $25.00 © 2003 American Chemical Society
Published on Web 07/08/2003

analogs, such as 3, the 23-deoxy,17R-hydroxy South 1.
Bioactivity evaluation of 23′-deoxy,17′R-hydroxy cepha-
lostatin 1 (4) was held to be highly desirable (Scheme 2).
formation of bisepoxide 7. Although stable on silica gel, as
with other singlet oxygen adducts,10 5 gradually converted
to 7 on standing (t1/2 ∼4 months, 25 °C). Palladium-catalyzed
hydrogenation of 5 furnished 86% of 8 and 10% of 7 after
exploring various solvents and temperatures. In contrast,
treatment of 5 with Zn/AcOH in dichloromethane at reflux
afforded 6 in 98% yield without a trace of 7. Hydrogenation
of 6 gave 8 in almost quantitative yield.
Scheme 2
Compound 6 reacted with hydrogen chloride etherate or
HBr (30% in AcOH) in dichloromethane to give the
corresponding ∆14,16R-halogenated products 9 and 10 in
good to excellent yields (Scheme 4). The structure of 9 was
Scheme 4 a
Herein, we report the results of the D-ring functionalization
of diene 1, the synthesis of 4, and the bioactivity of 2 and 4
compared to the natural cephalostatin 1.
D-Ring Functionalization of Diene 1. Reaction of dienyl
spiroketal 1 with singlet oxygen stereospecifically provided
the desired 4 + 2 adduct 58 in excellent yield9 (Scheme 3).
Scheme 3 a
a Reagents and conditions: (a) HCl‚OEt2, CH2Cl2, 0 °C, 3.5 h;
(b) HBr (30% in AcOH), CH2Cl2, 0 °C, 3.5 h; (c) TMSI, CH2Cl2,
25 °C, 10 min; (d) Ag2O, THF/H2O (4:1), reflux, 2 h; (e) Bu3SnH/
AIBN, benzene, reflux, 12 h.
secured by X-ray analysis.8 However, compound 11 was
obtained when 6 was treated with TMSI in dichloromethane
at room temperature, likely via epoxide 12. No 16â-
halogenated products were observed, presumably because of
steric shielding of the â face of the D ring by the E and F
rings. This observation was consistent with the hydrogena-
tion7 and singlet oxygen reactions of diene 1, in which R-face
products were exclusively formed. Epoxide 12 was isolated
in 74% yield by treating 9 with Ag2O in THF/H2O (4:1)
solution. Compound 10 failed to generate the desired South
1 analog 3 via radical dehalogenation. On the basis of proton
NMR, this reaction gave 50% of epoxide 12 admixed with
other unidentified compounds.
a Reagents and conditions: (a) sun lamp, O2, 0.5 mol % TPP,
CH2Cl2, 0 °C, 40 min; (b) Zn/AcOH, CH2Cl2, reflux, 30 min; (c)
H2, Pd/C, AcOEt, rt, 3 h; (d) H2, Pd/C, AcOEt, rt, 1 h.
In contrast, performing the reaction at 0 °C for extended
reaction time by reducing the amount of sensitizer 5,10,15,-
20-tetraphenyl-21H,23H-porphine (TPP) only led to the
(7) Li, W.; LaCour, T. G.; Fuchs, P. L. J. Am. Chem Soc. 2002, 124,
4548.
(8) The X-ray structures of 5, 9, 3, and 3â-OH of 13 have been submitted
to the Cambridge Crystallographic database.
The quest for an efficient preparation of 3 next turned to
a study of the dehydration of the 14,17-diol 8. Several of
(9) For Diels-Alder reactions of D-ring dienes of other steroidal
compounds, see: (a) Basler, S.; Brunk, A.; Jautelat, R.; Winterfeldt, E.
HelV. Chim. Acta 2000, 83, 1854. (b) Fell, J. D.; Heathcock, C. H. J. Org.
Chem. 2002, 67, 4742.
(10) (a) Fatma Sevin; Mckee, M. L. J. Am. Chem. Soc. 2001, 123, 4591-
4596. (b) Boyd, J. D.; Foote, C. S.; Imagawa, D. K. J. Am. Chem. Soc.
1980, 102, 3641.
2850
Org. Lett., Vol. 5, No. 16, 2003

Table 1. Dehydration of 14,17-Diol 8
Table 2. Epoxidation of Diene 1
resultsa
ratio by 1H NMRa
run
conditions
3:13:1:8
run
conditions
12:15:16
1
2
3
4
5
CeCl3‚7H2O/NaI, CH3CN, reflux, 21 h
TMSCl/NaI, CH3CN, rt, 10 min
H2SO4 (3 equiv), AcOH,rt, 8 h
H2SO4 (3 equiv), CH3CN, rt, 20 min
H2SO4 (20 equiv), CH3CN, 0 °C, 5 min
21:31:1:27
15:62:13:0
49:21:3:2
40:5:4:21
51:7:1:12
1
2
3
mCPBA, Na2HPO4, CH2Cl2, rt, 47 min
mCPBA, CH2Cl2, -45 to 0 °C, 2 h
DMDO, CH2Cl2,0 °C 40 min, rt 2.5 h
10:3:1
3.7:1.2:1
3.9:1:1.7
a 5-10% of diene 1 remains unreacted under these conditions. In run 1,
12 and 16 were isolated in 53% yield in a 10:1 ratio.
a All data are based on 1H NMR except in run 2, which are isolated
yields.
Table 2. Only three products were formed. The 16,17-â
epoxide was not formed because the â face was blocked by
the E and F rings. The structures of 15 and 16 were assigned
by reasoning that the epoxidations of 12 and 15 only give
R,â-bisepoxide 16 instead of the authentic R,R-bisepoxide
7.
We next explored epoxidation via dioxiranes derived from
more sterically demanding trifluoroacetophenone analogs.12
We were pleased to find that the ratio of 12 and (15 + 16)
was increased to 6-7:1 and desired epoxide 12 was obtained
in 72% yield (79% borsm) by using ketone 17 (Scheme 6).
the conditions explored are listed in Table 1. Compound 13
was formed as the major product when E2 elimination
dominated (runs 1 and 2). Under strongly acidic conditions,
the desired homoallylic alcohol 3 became major (runs 3-5),
presumably via E1 elimination. The structures of both 3 and
13 were secured by X-ray analysis.8 It was found that an
unidentified compound was formed in less than 10% yield
in runs 4 and 5. Although pure 3 could be obtained in 40%
yield by silica gel chromotography, the unknown was always
contaminated with 3 in a 1:1 ratio. Although only observed
when using acetonitrile and strong protic acids, the unknown
was not a product of Ritter reaction.11
Scheme 6 a
An alternative approach for generation of the desired
∆14,17R-OH moiety is via regioselective reduction of epoxide
12. As shown in Scheme 5, reduction of 12 using DIBAL-H
Scheme 5 a
a Conditions: ketone 17 (0.2 equiv), Oxone (0.6 equiv, 1.2 equiv
of oxidant), and NaHCO3 (2.4 equiv) in (1.5:1) CH3CN-0.05 M
Na2B4O7 in 4 × 10-4 M aqueous EDTA, 0 °C, 60 min, rt, 80 min.
The comparison of the three methods to prepare 14 is shown
in Scheme 7.
a Reagents and conditions: (a1) DIBAl-H, CH2Cl2, -78 °C, 3
h; (a2) TPAP/NMO/molecular sieves, CH2Cl2, rt, 3 h.
Scheme 7
gave a mixture of triols. Without purification, this mixture
was oxidized to 14, the 3-ketone product of 3, in 61% yield.
It was observed that SN2 hydride reduction using LAH failed,
presumably because the â face of the epoxide was blocked.
This approach is potentially superior to the dehydration of
14,17-diol 8 if preparation of 12 from diene 1 can be
improved.
Epoxidation of diene 1 using mCPBA or DMDO gave
epoxide 12 with poor selectivity. The results are shown in
Completion of Synthesis of 23′-Deoxy,17′r-hydroxy
Cephalostatin 1 (4). With 1413 in hand, bromide 18 and
(11) Krimen, L. I.; Cota, D. J. Org. React. 1969, 17, 213.
Org. Lett., Vol. 5, No. 16, 2003
2851

Table 3. Cytotoxicities versus Representative Tissue Types in NCI 10 Cell Lines (GI50 nM)
leukemia
HL-60
lung
colon
CNS
breast
ovary
melanoma
M-14
renal
RXF-393
prostate
PC-3
compound
A-549
HT-29
SF-295
MCF-7
IGR OV1
A-498
cephalostatin 1
2
4
0.11
2.4
8.3
0.30
3.5
3.6
3.9
21
33
<0.10
0.18
0.12
0.71
12
3.4
1.8
15
9.3
4.3
60
14
<0.10
0.52
0.23
0.32
6.6
9.3
0.12
13
azidoketone 19 were easily obtained in two steps, respec-
tively (Scheme 8). Thus, protected cephalostatin 1 analog
methodology developed in our laboratory.14 Deprotection
furnished the cephalostatin 1 analog 4.
Biological Activity. Testing of analogs 2 and 4 against
natural cephalostatin 1 in the National Cancer Institute (NCI)
10 cell lines revealed that 2 and 4 displayed very high
activity. The GI50 values are shown in Table 3.
Scheme 8 a
In conclusion, studies toward the D-ring functionalization
of diene 1 led to our efficient synthesis of 23-deoxy,17R-
hydroxy South 1 analog and the subsequent 23′-deoxy,17′R-
hydroxy cephalostatin 1 (4). Analog 4 featured the only
difference of the site of the hydroxyl group comparing to
the natural cephalostatin 1. DI50 values of 2 and 4 on NCI
human cancer 10 cell lines are in the nanomolar range.
Acknowledgment. We thank the National Institute of
Health (CA 60548) for funding this project. Special thanks
are due to Dr. John A. Beutler of the Molecular Targets Drug
Discovery Program at the NCI, Frederick, MD.
Supporting Information Available: Experimental pro-
cedures and copies of 1H and 13C NMR spectra. This material
is available free of charge via the Internet at http://pubs.acs.org.
a Reagents and conditions: (a) PTAB/HBr (cat. 30% in AcOH),
THF, 0 °C, 30 min; (b) TMGA, nitromethane (freshly distilled),
rt, 10 h; (c) 0.7 equiv of 21, PVP, Bu2SnCl2, benzene, reflux, 3 h;
(d) TBAF, THF, 65 °C, 2 h, then K2CO3, MeOH/H2O (12:1), reflux,
0.5 h.
OL034894E
(12) For more extensive studies, see the following paper.
(13) Compound 14 also could be obtained from 3 after deprotection and
oxidation, see Supporting Information.
(14) Guo, C.; Bhandaru, S.; Fuchs, P. L. J. Am. Chem. Soc. 1996, 118,
10672.
20 was prepared by coupling the azidoketone 19 with North
1 aminomethoxime 21 using the unsymmetrical coupling
2852
Org. Lett., Vol. 5, No. 16, 2003