Enantiocontrolled synthesis of (4S,5S,6S)-6-benzyl-4,5-epoxy-6-hydroxy-2-cyclohexen-1-one, a model compound for the epoxycyclohexenone moiety of scyphostatin

Article abstract of DOI:10.1016/S0040-4039(00)01312-5

An efficient synthesis of (4S,5S,6S)-6-benzyl-4,5-epoxy-6-hydroxy-2-cyclohexen-1-one (2) representing a model compound for the cyclohexenone moiety of scyphostatin (1) was accomplished; the method features masking of the enone system in 10 in the form of

Full text of DOI:10.1016/S0040-4039(00)01312-5

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 7651–7655
Enantiocontrolled synthesis of
(4S,5S,6S)-6-benzyl-4,5-epoxy-6-hydroxy-2-cyclohexen-1-one,
a model compound for the epoxycyclohexenone moiety of
scyphostatin
Takashi Izuhara† and Tadashi Katoh*
Sagami Chemical Research Center, Nishi-Ohnuma, Sagamihara, Kanagawa 229-0012, Japan
Received 22 June 2000; revised 26 July 2000; accepted 28 July 2000
Abstract
An efficient synthesis of (4S,5S,6S)-6-benzyl-4,5-epoxy-6-hydroxy-2-cyclohexen-1-one (2) representing a
model compound for the cyclohexenone moiety of scyphostatin (1) was accomplished; the method features
masking of the enone system in 10 in the form of the bromo ether 13 (10“11“12“13), aldol coupling
of 13 with benzaldehyde to construct the requisite asymmetric quaternary carbon center at the C-6
position (13“14), and epoxide ring formation (21“2) as the key steps. The key intermediate 10 was
prepared from commercially available (−)-quinic acid (3). © 2000 Elsevier Science Ltd. All rights reserved.
Keywords: scyphostatin; antibiotics; aldol reactions; enantiocontrol; cyclohexenones; epoxycyclohexenones.
Scyphostatin (1), isolated from the culture broth of Dasyscyphus mollissimus SANK-13892 by
the Sankyo research group in 1997, has been shown to be a powerful and specific inhibitor of
neutral sphingomyelinase (N-SMase).1–4 The use of N-SMase inhibitors can regulate the level of
ceramide, the product of sphingomyelin hydrolysis by N-SMase, in a wide variety of cells.5
Therefore, 1 is anticipated to be a promising agent for the treatment of ceramide-mediated
pathogenic states such as inflammation and immunological and neurological disorders.2,3,5 The
gross structure of 1 was revealed by extensive spectroscopic studies to have a novel, highly
oxygenated cyclohexane ring incorporated with a C-20 unsaturated fatty acid-substituted amino
propanol side chain.1,6–8 Its remarkable biological properties as well as its unique structural
features make 1 an exceptionally intriguing and timely target for total synthesis (Fig. 1).
* Corresponding author. Fax: +81-427-49-7631; e-mail: takatoh@alles.or.jp
† Graduate student from Department of Electronic Chemistry, Tokyo Institute of Technology, Yokohama 226-8502,
Japan.
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)01312-5

7652
Figure 1. Structures of scyphostatin (1) and the model compound (2) for the cyclohexenone moiety of 1
We embarked on a project directed at the total synthesis of 1 and its congeners with the aim
of exploring the structure–activity relationships. In this communication, we wish to report an
efficient and facile method for the synthesis of (4S,5S,6S)-6-benzyl-4,5-epoxy-6-hydroxy-2-cyclo-
hexen-1-one (2) which represents a model compound for the cyclohexenone moiety of 1. To the
best of our knowledge, this is the first approach toward the highly functionalized cyclohexenone
moiety of 1. The method for the synthesis of 2 involves the masking of the highly reactive enone
system in 10 in the form of the bromo ether 13 (10“11“12“13), the aldol coupling of 13 with
benzaldehyde to construct the requisite asymmetric quaternary carbon center at the C-6 position
(13“14), and the sequential epoxide ring formation (19“20“21“2) as the key steps.
At first, we pursued the synthesis of the key intermediate 10 possessing the requisite enone
system and three contiguous oxygen functionalities with correct stereochemistries at the C-4,
C-5, and C-6 positions (2-cyclohexen-1-one numbering) (Scheme 1). The starting material 4 was
prepared from commercially available (−)-quinic acid (3) in three steps according to the reported
procedure.9 After protection of the hydroxy group in 4 as its tert-butyldimethylsilyl (TBDMS)
ether (98%), the carbonyl group in the resulting TBDMS ether 5 was reduced with sodium
borohydride to provide the alcohols 6a (53%) and 6b (44%) as an epimeric mixture, which could
be separated by column chromatography on silica gel. Dehydration of 6b by treatment with
diethyl azodicarboxylate (DEAD) and triphenylphosphine occurred regioselectively to afford the
Scheme 1. Synthesis of the key intermediate 10 (a) TBDMSCl, imidazole, DMF, rt, 98%; (b) NaBH4, THF–H2O, rt,
53% for 6a, 44% for 6b; (c) DEAD, Ph3P, benzoic acid, THF, rt, 98%; (d) KOH, MeOH, rt, 100%; (e) DEAD, Ph3P,
THF, rt, 67%; (f) mCPBA, NaHCO3, CH2Cl2, rt, 92%; (g) Se2Ph2, NaBH4, EtOH, 0°C“reflux; H2O2, THF,
0°C“reflux, 78%; (h) Dess–Martin periodinane, CH2Cl2, rt, 95%

7653
desired olefin 7 in 67% yield.10 In this reaction, the other regioisomeric olefin was not obtained
at all. On the contrary, treatment of the epimeric alcohol 6a under the same conditions gave
none of the desired dehydration product 7, and resulted in complete recovery of the starting
material 6a.11 Therefore, 6a was converted to 6b (98%, two steps) by using the Mitsunobu
inversion procedure.12 To forward the synthesis, the olefin 7 was treated with 3-chloroperoxy-
benzoic acid (mCPBA) to furnish the epoxide 813 (92%) as a single stereoisomer, which was then
transformed to the allylic alcohol 9 (78%) by employing the Sharpless protocol.14 Finally,
oxidation of 9 by the use of Dess–Martin periodinane15 provided the key intermediate 1016 in
95% yield.
Having obtained the key intermediate 10, our next efforts were devoted to the crucial aldol
coupling of 10 with benzaldehyde. We envisaged that the introduction of the benzyl group at the
C-6 position in 10 would occur exclusively from the less hindered a-face of the enolate generated
from 10 under the influence of the b-oriented O-isopropylidenedioxy moiety. However, initial
attempts to achieve aldol coupling of 10 with benzaldehyde were quite fruitless because the
enone olefin moiety involved in 10 was extremely susceptible to nucleophilic attack of the
enolate generated from 10 itself.17 Consequently, we decided to mask the highly reactive enone
system of 10 in the form of the bromo ether 13 during the aldol coupling. Toward this end, as
shown in Scheme 2, 10 was converted to 13 in 63% overall yield via a three-step sequence
involving Diels–Alder reaction with cyclopentadiene in the presence of diethylaluminum
chloride,18 deprotection of the TBDMS group in the endo-adduct 11, and bromo etherification
of the resulting alcohol 12 with N-bromosuccinimide (NBS).19 We were delighted to find that the
crucial coupling of 13 with benzaldehyde under standard conditions [LiN(TMS)2 (2.5 equiv.),
THF, −78°C] proceeded smoothly with concomitant formation of a cyclopropane ring, affording
a remarkable yield (98%) of the coupling product 14 as an almost inseparable mixture of the
1
epimeric alcohols (ca. 6:1 by 500 MHz H NMR). Removal of the hydroxy group in 14 was
Scheme 2. Synthesis of the model compound 2 (a) Cyclopentadiene, Et2AlCl, CH2Cl2, −78“0°C, 98%; (b) TBAF,
THF, 0°C, 75%; (c) NBS, CH2Cl2, 0°C, 86%; (d) LiN(TMS)2, THF, −78°C; benzaldehyde, −78°C, 98%; (e) phenyl
chlorothionoformate, DMAP, MeCN, rt, 92%; (f) n-Bu3SnH, AIBN, toluene, 110°C, 79%; (g) TMSI, CCl4, −10°C,
89%; (h) Zn, AcOH, MeOH, 60°C, 91%; (i) Ph2O, reflux, 93%; (j) MsCl, pyridine, DMAP, CH2Cl2, 0°C“rt, 89%; (k)
TFA, H2O, 0°C, 85%; (l) 0.2 M NaOH, Et2O, 0°C, 90%

7654
effected by using Robins’ modification20 of the Barton method. Thus, treatment of 14 with
phenyl thionochloroformate in acetonitrile in the presence of 4-dimethylaminopyridine (DMAP)
provided the corresponding phenoxythionocarbonyl ester 15 (92%), which was then allowed to
react with tri-n-butyltinhydride and a catalytic amount of 2,2-azobisisobutyronitrile (AIBN) in
toluene at 110°C, furnishing the desired deoxygenated product 1621,22 in 79% yield.
Having introduced the requisite benzyl substituent with the correct stereochemistry at the C-6
position, we then focused our attention on regeneration of the cyclohexenone olefin moiety.
After regioselective cleavage of the cyclopropane ring in 16 by reaction with iodotrimethylsilane
(TMSI)23 (89%), the resulting g-iodo ketone 17 was effectively converted to the requisite
cyclohexenone 19 by applying the conditions of Ogasawara.19 Thus, 17 was treated with zinc
powder in methanol containing acetic acid to give the endo-alcohol 18 (91%), which was then
subjected to a retro-Diels–Alder reaction19 by thermolysis at reflux in diphenyl ether, providing
19 in 93% yield. The final phase remaining to complete the synthesis of 2 was the critical epoxide
ring formation utilizing the two oxygen functionalities present at the C-4 and C-5 positions in
19. Mesylation of the hydroxy group in 19 under the standard conditions followed by acid
hydrolysis of the acetonide moiety in the mesylate 20 afforded the corresponding diol 21 in 76%
yield for the two steps. Finally, brief exposure of 21 to aqueous sodium hydroxide in ether at
0°C led to the formation of the target model compound 224 in 90% yield.
In summary, we have succeeded in developing a facile synthetic pathway to (4S,5S,6S)-6-
benzyl-4,5-epoxy-6-hydroxy-2-cyclohexen-1-one (2) which is the first entry to the highly func-
tionalized cyclohexenone moiety of scyphostatin (1). The explored synthetic method features the
masking of the highly reactive enone system in 10 in the form of the bromo ether 13, the aldol
coupling of 13 with benzaldehyde to construct the requisite asymmetric quaternary carbon
center at the C-6 position, and the final epoxide ring formation as the key steps. Based on this
strategy, work on the total synthesis of 1 is in progress and will be reported shortly.
Acknowledgements
We are grateful to Drs. H. Kogen and T. Ogita, Sankyo Co., Ltd., for many valuable
suggestions and encouragement.
References
1. Tanaka, M.; Nara, F.; Suzuki-Konagai, K.; Hosoya, T.; Ogita, T. J. Am. Chem. Soc. 1997, 119, 7871–7872.
2. Nara, F.; Tanaka, M.; Hosoya, T.; Suzuki-Konagai, K.; Ogita, T. J. Antibiot. 1999, 52, 525–530.
3. Nara, F.; Tanaka, M.; Masuda-Inoue, S.; Yamamoto, Y.; Doi-Yoshioka, H.; Suzuki-Konagai, K.; Kumakura,
S.; Ogita, T. J. Antibiot. 1999, 52, 531–535.
4. It is reported that scyphostatin (1) exhibits N-SMase and acidic SMase (A-SMase) with IC50 values of 1.0 and
49.3 mM, respectively.1–3 This natural product is the most potent of the few known small molecule inhibitors for
N-SMase.
5. Ceramide, the primary sphingomyelin catabolite, has been recognized to be a lipid second messenger in cell
membranes and plays key roles in the regulation of cell proliferation, differentiation, and apoptosis. For a recent
excellent review on the catabolites of sphingolipids as novel therapeutic targets, see: Kolter, T.; Sandhoff, K.
Angew. Chem., Int. Ed. 1999, 38, 1532–1568.
6. This initial structure elucidation1 only determined the relative and absolute stereochemistry of the cyclohexenone
moiety in 1. Quite recently, the Sankyo group elucidated and reported the relative and absolute configurations of

7655
the three stereocenters within the C-20 unsaturated fatty acid moiety by chemical degradation of 1 followed by
extensive chemical correlation to the known chiral compounds.7 Subsequently, Hoye et al. disclosed the
enantioselective synthesis of the C-20 unsaturated fatty acid moiety, leading to alternative proof of its stereostructure
including absolute configuration.8
7. Saito, S.; Tanaka, N.; Fujimoto, K.; Kogen, H. Org. Lett. 2000, 2, 505–506.
8. Hoye, T. R.; Tennakoon, M. A. Org. Lett. 2000, 2, 1481–1483.
9. Wang, Z.-X.; Miller, S. M.; Anderson, O. P.; Shi, Y. J. Org. Chem. 1999, 64, 6443–6458 and references cited therein.
10. The NOESY experiment of 6b supported that the cyclohexane ring adopts a boat-form which places the hydroxy
group in an axial position; this conformation would facilitate E2 elimination leading to the formation of 7.
11. The NOESY experiment of 6a indicated that the cyclohexane ring takes a boat-form and the hydroxy group is
equatorial orientation; this conformation may preclude any possibility of E2 elimination.
12. Mitsunobu, O. Synthesis 1981, 1–28.
13. The stereochemistry of the newly produced epoxide ring in 8 was determined by an NOE experiment.
14. Sharpless, K. B.; Lauer, R. F. J. Org. Chem. 1973, 38, 2697–2699.
15. (a) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155–4156. (b) Dess, D. B.; Martin, J. C. J. Am. Chem.
Soc. 1991, 113, 7277–7287. (c) Ireland, R. E.; Liu, L. J. Org. Chem. 1993, 58, 2899.
16. Compound 10: Colorless prisms; mp 55–56°C; [h]2D0 −84.7° (c 1.02, CHCl3); IR (KBr): 3545, 3368, 2990, 2934, 2859,
1
1696, 1464, 1383, 1252, 1167, 1076, 1005, 891, 839, 779, 727, 669, 517 cm−1; H NMR (500 MHz, CDCl3) l: 0.14
(3H, s), 0.17 (3H, s), 0.92 (9H, s), 1.40 (3H, s), 1.43 (3H, s), 4.39–4.42 (1H, m), 4.44 (1H, d, J=5.9 Hz), 4.52–4.55
(1H, m), 6.08 (1H, d, J=10.3 Hz), 6.76 (1H, ddd, J=0.9, 3.8, 10.3 Hz); 13C NMR (125 MHz, CDCl3) l: 194.53,
148.47, 127.87, 110.17, 79.64, 74.36, 67.07, 27.43, 25.88, 25.70 (three carbons), 18.08, −4.73, −4.74; EIMS m/z:
298 (M+), 283 [(M−Me)+].
17. Reaction of the enolate of 10, generated by treatment with LiN(TMS)2, with benzaldehyde in THF at −78°C resulted
in the predominant formation of the dimerized product i (46%) along with a small amount of the desired coupling
product ii (12%).
18. It is noteworthy that this Diels–Alder reaction proceeded smoothly in a completely diastereofacial- and
endo-selective manner to give the corresponding adduct 11 as a single isomer in almost quantitative yield (98%).
19. Ogasawara, K. J. Synth. Org. Chem. Jpn. 1999, 57, 957–968 and references cited therein.
20. Robins, M. J.; Wilson, J. S.; Hansske, F. J. Am. Chem. Soc. 1983, 105, 4059–4065.
21. Compound 16: Colorless prisms; mp 103–104.5°C; [h]2D0 −69.6° (c 0.97, CHCl3); IR (KBr): 2986, 2934, 1711, 1494,
1
1454, 1381, 1296, 1238, 1167, 1062, 939, 877, 831, 767, 702, 513 cm−1; H NMR (500 MHz, CDCl3) l: 0.91 (3H,
s), 1.36 (3H, s), 1.67 (1H, d, J=1.4, 6.4 Hz), 1.79 (1H, d, J=11.0 Hz), 1.85 ( 1H, d, J=11.0 Hz), 2.33 (1H, d,
J=6.4 Hz), 2.46 (1H, s), 2.90 (1H, t, J=2.3 Hz), 2.99 (1H, d, J=13.8 Hz), 3.22 (1H, d, J=13.8 Hz), 4.38 (1H,
d, J=3.0 Hz), 4.53 (1H, t, J=2.8 Hz), 4.57 (1H, t, J=2.3 Hz), 7.20–7.35 (5H, m); 13C NMR (125 MHz, CDCl3)
l: 207.48, 135.20, 131.90 (two carbons), 127.88 (two carbons), 126.96, 109.22, 85.73, 83.20, 78.54, 76.83, 46.16,
43.29, 41.20, 31.91, 31.85, 30.30, 28.12, 26.32, 20.55; EIMS m/z: 338 (M+), 280 [(M−Me2CO)+], 247 [(M−PhCH2)+].
22. At this stage, the stereochemistry at the newly formed C-6 position of the aldol coupling product 14 was confirmed
by NOE experiment of 16; thus, NOE interactions between the signals due to the benzylic methylene protons and
the C-5 proton were observed.
23. Miller, R. D.; Mckean, D. R. J. Org. Chem. 1981, 46, 2412–2414.
24. Compound 2: Colorless caramel; [h]2D0 +45.6° (c 0.80, CHCl3); IR (neat): 3481, 3030, 2920, 1690, 1495, 1454, 1379,
1
1254, 1238, 1196, 1146, 1094, 958, 860, 843, 750, 704, 627, 579, 544, 503 cm−1; H NMR (500 MHz, CDCl3) l:
2.93 (1H, d, J=13.6 Hz), 3.01 (1H, d, J=13.6 Hz), 3.60 (1H, dt, J=1.6, 3.9 Hz), 3.65 (1H, s), 3.77 (1H, d, J=3.9
Hz), 6.16 (1H, dd, J=1.6, 9.9 Hz), 7.09–7.15 (3H, m), 7.22–7.32 (3H, m); 13C NMR (125 MHz, CDCl3) l: 197.48,
145.14, 133.55, 130.31 (two carbons), 130.15, 128.37 (two carbons), 127.33, 77.66, 56.01, 47.94, 44.37; EIMS m/z:
216 (M+), 199 [(M−OH)+].
.