An efficient method for the synthesis of versatile intermediates leading to 13-deoxy- and 9,13-dideoxyphorbols

Article abstract of DOI:10.1016/S0040-4039(99)02103-6

An efficient method for the synthesis of versatile intermediates for biologically interesting 13-deoxy- and 9,13-dideoxyphorbol ester analogs is described. First, more efficient synthetic routes to the bicyclic ketones 7 and 8, which are well-known interm

Full text of DOI:10.1016/S0040-4039(99)02103-6

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 509–513
An efficient method for the synthesis of versatile intermediates
leading to 13-deoxy- and 9,13-dideoxyphorbols
Akihiro Sekine, Naoya Kumagai, Koichiro Uotsu, Takashi Ohshima and
Masakatsu Shibasaki ∗
Graduate School of Pharmaceutical Sciences, The University of Tokyo, Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
Received 12 October 1999; revised 1 November 1999; accepted 2 November 1999
Abstract
An efficient method for the synthesis of versatile intermediates for biologically interesting 13-deoxy- and
9,13-dideoxyphorbol ester analogs is described. First, more efficient synthetic routes to the bicyclic ketones 7
and 8, which are well-known intermediates for the synthesis of 13-deoxyphorbols, than the previous one were
established. Second, 7 and 8 were successfully converted to the intermediates 19, 25 and 27 for the synthesis of
9,13-dideoxyphorbols using Peterson reaction, oxymercuration and nitrile oxide cycloaddition as key steps. © 2000
Elsevier Science Ltd. All rights reserved.
Keywords: PKC; tumor promotors; structure–activity relationships.
The enzyme protein kinase C (PKC) is now thought1 to play a major role in cellular signal transduction
and as a consequence is apparently involved in cell growth regulation, including those diverse growth
patterns which lead to tumor promotion and cancer. Since its first isolation, PKC has been shown to exist
as several isoforms, the relative amounts and activities of which vary amongst cell populations. PKC is
strongly activated in cells by endogenous 1,2-diacyl compounds such as diacylglycerol and other diverse
tumor promotors, the most commonly known of which is phorbol 12-myristate-13-acetate (1: PMA) (Fig.
1). Computer modeling studies have shown2 that compounds which are activators of PKC have common
Fig. 1.
∗
Corresponding author. Tel: +0081 0 3 5684 0651; fax: +0081 0 3 5684 5206; e-mail: mshibasa@mol.f.u-tokyo.ac.jp (M.
Shibasaki)
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(99)02103-6
tetl 16031

510
structural features. In the case of PMA (1), these include a carbonyl group at C-3, hydroxyl groups at
C-4, C-9, and C-20, and a long-chain hydrophobic portion at C-12, indicating that an acetoxy group at
C-13 is not important for the interaction of PMA (1) with PKC.
However, the biological evaluation of 13-deoxyphorbol esters3 and photoaffinity labeling by 13-
diazoacetyl phorbol ester derivative or 13-trifluorodiazopropionyl phorbol ester derivative4,5 have sug-
gested that the ester group at C-13 in phorbol esters is located in close proximity to PKC and interacts
with PKC in phorbol ester-PKC-phospholipid complexes. Moreover, some recent reports on molecular
modeling have supported the importance of the C-13 ester group.6 It is of interest to assume that the
C-13 acetoxy group would form a hydrogen bond with the C-9 hydroxyl group, not with PKC, based
on X-ray crystallography7 and MM calculation,6 suggesting that these functional groups would interact
with regulatory domain of PKC by hydrophobic interaction through a hydrogen bond between the C-
13 acetoxy group and the C-9 hydroxyl group. Therefore, 9,13-dideoxyphorbol 3 would be quite an
interesting compound, providing further information on structure–activity relationships of phorbols to
propose a more precise pharmacophore model. In this communication we report an efficient method for
the synthesis of versatile intermediates 7, 8, 19, 25 and 27 leading to 13-deoxy- and 9,13-dideoxyphorbols
2 and 3.
Scheme 1. Reagents and conditions: (a) Me2CuLi (2.0 equiv.), THF, 0°C, 30 min, 98% (b) TMSCl (3.0 equiv.), LHMDS (1.2
equiv.), THF, −78°C→rt, 2 h, 78% (c) 36% HCHO aq., Yb (OTf)3 (5.0 mol%) rt, 85 h, 81% (d) TBSCl (1.5 equiv.), imidazole
(3.0 equiv.), DMF, rt, 2 h, 83% (e) LDA (1.3 equiv.), zirconocene dichloride (1.3 equiv.), PhSeBr (1.1 equiv.), THF, −78°C, 2
h, 84% (f) mCPBA (1.5 equiv.), CH2Cl2, rt, 2 h, 85% (g) n-Bu4NF (1.2 equiv.), THF, 0°C, 2 h, quant. (h) (1) 30% H2O2 aq. (4.0
equiv.), NaOH (0.1 equiv.), THF, 0°C, 2 h, (2) pivaloyl chloride (1.5 equiv.), pyridine (3.0 equiv.), CH2Cl2, rt, 10 h, 81% (two
steps) (i) Al–Hg, THF:EtOH:H2O (3:2:1), −45°C, 4 days, 87% (j) TBDPSCl (2.0 equiv.), imidazole (4.0 equiv.) DMF, rt, 6 h,
60% (k) SmI2 (2.0 equiv.), THF, −78→0°C then PhSeBr (2.0 equiv.), 0°C→rt, 15 min, 75% (l) TMSCl (3.0 equiv.), LHMDS
(1.3 equiv.), THF, −78°C, 10 min, 97% (m) 36% HCHO aq., Yb(OTf)3 (10 mol%), THF, rt, 5 h, 64% (n) 30% H2O2 aq. (10
equiv.), THF, rt, 5 min, quant.
Our strategy for synthesis of 3 basically stands on that for the synthesis of 2.3 However, the synthesis
of the intermediate 7 was rather lengthy in our previous attempt (11 steps from (+)-3-carene, 30%
overall yield).3 Therefore, a shorter synthetic route to 7, which is also useful for the medicinal chemistry
of 13-deoxyphorbol esters 2, was first investigated. As a result, we succeeded in synthesizing 7 with
improved and shorter procedures than the previous one (seven steps from (+)-3-carene, 42% overall
yield) (Scheme 1). Highly regioselective enol silylation of 5, which is easily derived from (+)-3-carene
in five steps,8 with LHMDS and TMSCl, and subsequent aldol reaction of the enol silyl ether using 36%
9
aqueous HCHO and 5 mol% of Yb(OTf)3 gave 7 in 63% yield from 5. The stereochemistry of 7 thus
prepared was unequivocally determined by comparison with an authentic sample.3 It is noteworthy that

511
in this route all the intermediates can be purified by simple distillation under reduced pressure. Then,
according to the established procedure,3 7 was transformed into 8 as shown in Scheme 1. Furthermore,
we were pleased to find that 8, also a intermediate for 13-deoxyphorbols, could be synthesized in shorter
10
steps. The acetoxy ketone 4 was treated with SmI2 in THF, and the resulting samarium enolate was
directly reacted with PhSeBr to give 12 in 75% yield. Compound 12 was converted to enol silyl ether in
a highly regiocontrolled manner (97%), which was subjected to aldol reaction using aqueous HCHO
and Yb(OTf)3, affording 13 (64%) exclusively. Treatment of 13 with aqueous H2O2 furnished 8 in
quantitative yield. The enone 8 was then converted to 11 according to the established procedure3 (Scheme
1).
With optically pure 11 in large quantities, we then attempted crucial stereocontrolled conversion of C-9
ketone in 11 to a vinyl group. After many attempts, we finally found that the strategy using homologation
of ketone to give aldehyde followed by Wittig olefination was the most effective (Scheme 2). Treatment of
11 with cerium reagent prepared from Me3SiCH2OCH3, sec-BuLi and CeCl311,12 gave 14 in 70% yield as
a mixture of diastereomers. Direct conversion of 14 into aldehyde 17 with formic acid was not successful
because of the sensitivity of 14 to acid. Enol methyl ether 15 obtained in a highly regiocontrolled manner
from 14 on treatment with KH as single stereoisomer was also sensitive to strongly acidic conditions, such
as H2O:H2SO4:acetone (1:1:2), required to hydrolyze hindered trisubstituted enol methyl ether. Then, as
an alternative process, oxymercuration of 15 with Hg(OAc)2 and MeOH followed by NaBH4 reduction
was examined, giving dimethyl acetal 16 as a desired single isomer.13 The stereochemistry of 16 was
unequivocally determined by NOE experiments as shown in Scheme 2. Acetal 16 was treated with TsOH
in acetone to give aldehyde 17 in quantitative yield. However, a small amount of aldehyde was epimerized
to the undesired isomer in a ratio of 1:10. Finally, Wittig olefination of 17 gave vinyl compound 18.
Then, C-12 oxygen of 18 was inverted to give the desired isomer 19 exclusively by oxidation followed
by reduction.
Scheme 2. Reagents and conditions: (a) Me3SiCH2OCH3 (2.5 equiv.), sec-BuLi (2.0 equiv.), CeCl3 (3.0 equiv.), THF,
−40→0°C, 2 h, 70% (b) KH (8.0 equiv.), THF, rt, 2 h, 73% (c) Hg(OAc)2 (1.1 equiv.), MeOH:CH2Cl2 (1:1), 0°C, 3 h then
NaBH4 (2.0 mol equiv.), 0°C, 10 min, 70% (d) p-toluenesulfonic acid (3.0 equiv.), acetone:H2O (10:1), rt, 10 h (e) Ph3PCH3Br
(4.0 equiv.), n-BuLi (3.0 equiv.), THF, 0°C, 2 h, 98% (two steps), (9β:9α, 10:1) (f) n-Bu4NF (2.0 equiv.), THF, rt, 11 h, 88%,
(g) SO3·Py (10 equiv.), Et3N (20 equiv.), DMSO, rt, 1 h, 86% (h) NaBH4 (8.0 mol equiv.), MeOH, 0°C, 30 min, 86% (i) TBSCl
(5.0 equiv.) imidazole (10 equiv.), DMF, rt, 22 h, 96%
Having synthesized optically pure 19, B-ring construction was attempted. First, as shown in Scheme
3, 19 was transformed into 24 according to the established procedure.3
Unfortunately, however, the intramolecular nitrile oxide cycloaddition with p-chlorophenyl isocyanate
and triethylamine gave 25 in only poor yield, together with the recovery of the starting material 24 (80%)
(Scheme 4). The stereochemistry of 25 was unequivocally determined by NOE experiments. We assumed
that 24 was conformationally unfavorable for intramolecular nitrile oxide cycloaddition, and in fact this

512
Scheme 3. Reagents and conditions: (a) (1) LiAlH4 (2.0 mol equiv.), Et2O, 0°C, 30 min (2) SO3·Py (2.0 equiv.), Et3N (4.0
equiv.), DMSO, rt, 2 h, 95% (b) 21 (6.0 equiv.), KHMDS (4.0 equiv.), THF, rt, 18 h, 73% (c) (1) DIBAL-H (3.0 equiv.),
toluene, −78°C, 3 h (2) TBDPSCl (3.0 equiv.), imidazole (15 equiv.), DMF, rt, 5 h, 74% (two steps) (d) (1) DDQ (1.5 equiv.),
CH2Cl2:H2O (10:1), rt, 2 h (2) MeSO2Cl (5.0 equiv.), Et3N (10 equiv.), CH2Cl2, 0°C, 30 min (3) Nal (40 equiv.), 2-butanone,
50°C, 9 h, 63%, (three steps) (e) AgNO2 (5.0 equiv.), Et2O, reflux, 48 h, 67%
appeared to be supported by our MM calculation. On the other hand, based on MM calculation, 26 was
expected to give 27 more efficiently. To our delight, intramolecular nitrile oxide cycloaddition of 26
derived from 18 proceeded smoothly to give 27 in 75% yield.
Scheme 4.
In conclusion, we have succeeded in synthesizing 27 efficiently starting from optically pure (+)-3-
carene. The tetracyclic isoxazolines 25 and 27 are already interesting compounds for biological evalua-
tion. This is the first example of a synthetic approach to 9-deoxyphorbols. Therefore, the synthetic inter-
mediates such as 19 and 27 would be quite useful for research groups working in the structure–activity
relationships of phorbols. Details of all the results will be reported in due course.
References
1. (a) Nishizuka, Y. Nature 1984, 308, 693. (b) Farago, A.; Nishizuka, Y. FEBS Lett. 1990, 268, 350 and references cited
therein.
2. (a) Castagna, M.; Takai, Y.; Kaibuchi, K.; Sano, K.; Kikkawa, U.; Nishizuka, Y. J. Biol. Chem. 1982, 257, 7847. (b) Jeffery,
A. M.; Liskamp, R. M. J. Proc. Natl. Acad. Sci. USA 1986, 83, 241. (c) Wender, P. A.; Koehler, K. F.; Sharkey, N. A.;
Dell’Aquila, M. L.; Blumberg, P. M. J. Proc. Natl. Acad. Sci. USA 1986, 83, 4214. (d) Wender, P. A.; Cribbs, C. M.;
Koehler, K. F.; Sharkey, N. A.; Herald, C. L.; Kamano, Y.; Pettit, G. R.; Blumberg, P. M. J. Proc. Natl. Acad. Sci. USA 1988,
85, 7197. (e) Itai, A.; Kato, Y.; Tomioka, N.; Iitaka, N.; Endo, Y.; Hasegawa, M.; Shudo, K.; Fujiki, H.; Sakai, S. J. Proc.
Natl. Acad. Sci. USA 1988, 85, 3688. (f) Nakamura, H.; Kishi, Y.; Pajares, M. A.; Rando, R. R. J. Proc. Natl. Acad. Sci. USA
1989, 86, 9672. (g) Rando, R. R.; Kishi, Y. Biochemistry 1992, 31, 2211.

513
3. (a) Sugita, K.; Shigeno, K.; Neville, C. F.; Sasai, H.; Shibasaki, M. Synlett 1994, 325. (b) Sugita, K.; Neville, C. F.; Sodeoka,
M.; Sasai, H.; Shibasaki, M. Tetrahedron Lett. 1995, 36, 1067.
4. Sodeoka, M.; Uotsu, K.; Shibasaki, M. Tetrahedron Lett. 1995, 36, 8795.
5. Uotsu, K.; Sodeoka, M.; Shibasaki, M. Bioorg. Med. Chem. 1998, 6, 1117.
6. Wang, S.; Kazanietz, M. G.; Blumberg, P. M.; Marquez, V. E.; Milne, G. W. A. J. Med. Chem. 1996, 39, 2541.
7. Zhang, G.; Kazanietz, M. G.; Blumberg, P. M.; Hurley, J. H. Cell 1995, 81, 917.
8. Paquette, L. A.; Ross, R. J.; Shi, Y. J. J. Org. Chem. 1990, 55, 1589.
9. Kobayashi, S.; Hachiya, I. J. Org. Chem. 1994, 59, 3590.
10. Girard, P.; Namy, J. L.; Kagan, H. B. J. Am. Chem. Soc. 1980, 102, 2693.
11. Magnus, P.; Roy, G. Organometallics 1982, 1, 553.
12. Imamoto, T.; Kusumoto, Y.; Tawarayama, Y.; Sugiura, T.; Mita, Y.; Hatanaka, Y.; Yokoyama, M. J. Org. Chem. 1984, 49,
3904.
13. Radical intermediate is formed by NaBH4 reduction of alkylmercury compound, and this is readily equilibrated to equatorial
acetal.