Asymmetric synthesis of the key intermediate of an SP antagonist RPR107880 using a base-catalyzed Diels-Alder reaction

Article abstract of DOI:10.1016/S0040-4039(00)00555-4

An asymmetric synthesis of the key intermediate of RPR 107880, a substance P antagonist, using a base-catalyzed Diels-Alder reaction of 3- hydroxy-2-pyrone and N-benzylmaleimide is described. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00555-4

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 4147–4150
Asymmetric synthesis of the key intermediate of an SP antagonist
RPR107880 using a base-catalyzed Diels–Alder reaction
Hiroaki Okamura,∗ Hideki Shimizu, Yasuko Nakamura, Tetsuo Iwagawa and
Munehiro Nakatani
Department of Chemistry and Bioscience, Faculty of Science, Kagoshima University, 1-21-35 Korimoto, Kagoshima
890-0065, Japan
Received 28 February 2000; revised 16 March 2000; accepted 31 March 2000
Abstract
An asymmetric synthesis of the key intermediate of RPR 107880, a substance P antagonist, using a base-
catalyzed Diels–Alder reaction of 3-hydroxy-2-pyrone and N-benzylmaleimide is described. © 2000 Elsevier
Science Ltd. All rights reserved.
Keywords: base-catalyzed Diels–Alder reaction; asymmetric synthesis; RPR 107880; substance P.
Substance P (SP) is a small peptide classified in the tachykinine family, and it behaves as a
neurotransmitter.1 Recent investigations into SP have disclosed its various physiological functions, such
as pain transmission, vasodilatation, smooth muscle contraction, plasmatic extravasations, inflammation,
and the regulation of immune response.2 Therefore, such antagonists of SP are expected to be excellent
drugs. Over the last 10 years vigorous research to find strong and stable antagonists was undertaken, and
some compounds were reported as antagonists.3 Particularly, the perhydrorisoindole derivatives (ex. 1)
were recently recognized as efficient antagonists (Scheme 1).3a–c
Scheme 1.
In the preceding paper, we reported the base-catalyzed Diels–Alder (DA) reaction of 3-hydroxy-
2-pyrone (3) and N-methylmaleimide (4a),4a and its asymmetric reaction using a cinchona alkaloid
∗
Corresponding author. Fax: +81 99 285 8117; e-mail: okam@sci.kagoshima-u.ac.jp (H. Okamura)
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00555-4
tetl 16834

4148
catalyst (Scheme 2).4b Since the resulting adduct 5a has various functionalities and a rigid cyclic
structure, it seems to be an excellent building block for the synthesis of various molecules, in particular,
perhydroisoindoles. Thus, we have examined the synthesis of SP antagonists RPR 107880 (1) and related
compounds.3 Although the asymmetric syntheses of these compounds were already achieved by using the
chiral pool method3a or optical resolution of racemic material,3b,c no direct methods using an asymmetric
reaction have been reported. In this paper, we disclose an efficient method to construct perhydroisoindole
derivative using the asymmetric base-catalyzed DA reaction.
Scheme 2.
As a concrete synthetic target, we chose compound 2 which was originally derived from (R)-(+)-
pulegone as the key intermediate of 1 (Scheme 1). Since this compound was anticipated to be synthesized
from 5b, we examined the asymmetric base-catalyzed DA reaction of 3 and N-benzylmaleimide (4b)
prior to the synthesis (Table 1).
Table 1
Asymmetric base-catalyzed DA reaction of 4 and 5ba
As well as the asymmetric reaction of 3 and 4a,4b cinchona alkaloid catalyst, CH2Cl2 solvent,
and low temperature (−78∼−20°C) were again effective to give optically active 5b, and the highest
enantioselectivity was obtained from the reaction with 1 equivalent of quinine (entry 2).5 Although
decreasing the amount of catalyst (entry 3) lowered the enantioselectivity, a slight improvement was
achieved by the slow-addition method (condition B, entry 4). Based on several examinations, over 7 g of
(−)-5b in 59% ee was obtained from one reaction using 0.3 equivalents of quinine (entry 4).
This insufficient optical purity of the resulting adduct prompted us to examine the purification of the
enantiomer. Since the resulting product was obtained as a crystalline material, we initially attempted the
purification by recrystallization, and surprisingly, only one recrystallization from hexane/ethyl acetate
afforded the complete optically pure mother liquor and racemic crystalline material. Indeed, recrystal-
lization of 7.0 g of (−)-5b (59% ee) from ca. 20 ml of hexane:ethyl acetate (1:9) afforded 2.9 g of
racemate as a crystalline material and 4.1 g of the homochiral (−)-enantiomer from the mother liquor.

4149
These weights were quite consistent with calculated quantities from the amount and purity of the starting
5b.
With quantities of the optically pure starting material in hand, we started the synthesis of the
intermediate 2 (Scheme 3). The reduction to give trihydroxylamine (7) was quite difficult because of
the instability of 5b under basic condition. From several examinations, compound 7 was obtained in fair
yield by TMS protection of 5b and subsequent AlH3 reduction. The oxidative cleavage of 7 followed by
methylation using MeMgBr exclusively afforded the desired isomer 8. The stereochemistry of this methyl
group was presumed to be controlled by coordination between MeMgBr and the neighboring oxygen
atom at the hydroxyl group. After the methylation, the hydroxyl group was removed by thiocarbonylation
followed by n-Bu3SnH reduction to give the desired intermediate 2 in good yield. To confirm its structure
and stereochemistry, 2 was further converted to another intermediate 10 using a known process.3a Since
the 1H NMR and [α]D value of 10 showed good agreement with those of previously reported data,3a,6,7
the structure including absolute stereochemistry were confirmed as shown in Scheme 3.
Scheme 3. Reagents and conditions: (a) TMSCl, imidazole, DMF, 60°C, 12 h, 93%; (b) AlH3, THF, −78°C–rt, 2 days, 62%; (c)
NaIO4, aq THF, 0°C, 12 h, 74%; (d) MeMgBr, THF, −78°C, 2 h, 48%; (e) thiocarbonyldiimidazole, THF, reflux, 2 days, 46%
(26% of 10 was recovered); (f) n-Bu3SnH, AIBN, toluene, reflux, 2 h, 93%; (g) o-anisyllithium, LiBr, −78°C, 2 h, 62%
In conclusion, we have established an efficient method to obtain a homochiral perhydroisoindole
derivative using an asymmetric DA reaction. Since perhydroisoindoles are supposed to be prospects of
the strong SP antagonists and potential drugs, further synthetic applications using this method are under
way.
Furthermore, quite efficient recrystallization to give optically pure 5b was also noteworthy. Detailed
features of this recrystallization are also under investigation.
References
1. Guard, S.; Watson, S. P. Neurochem. Int. 1991, 18, 149–165.
2. (a) Otsuka, M.; Yanagisawa, M. Trends Pharmacol. Sci. 1987, 8, 506–510. (b) Maggio, J. E. Annu. Rev. Neurosci. 1988,
11, 13–28.
3. (a) Mutti, S.; Daubie, C.; Decalogne, F.; Fournier, R.; Rossi, P. Tetrahedron Lett. 1996, 37, 3125–3128. (b) Mutti, S.;
Daubie, C.; Malpart, J.; Radisson, X. Tetrahedron Lett. 1996, 37, 8743–8746. (c) Peyronel, J.-F.; Truchon, A.; Moutonnier,
C.; Garret, C. Bioorg. Med. Chem. Lett. 1992, 2, 37–40. (d) Lowe III, J. A.; Drozda, S. E.; Snider, M.; Longo, K. P.; Zorn,
S. H.; Morrone, J.; Jackson, E. R.; McLean, S.; Bryce, D. K.; Bordner, J.; Nagahisa, A.; Kanai, Y.; Suga, O.; Tsuchiya, M.
J. Med. Chem. 1992, 35, 2591–2560.
4. (a) Okamura, H.; Iwagawa, T.; Nakatani, M. Tetrahedron Lett. 1995, 36, 5939–5942. (b) Okamura, H.; Nakamura, Y.;
Iwagawa, T.; Nakatani, M. Chem. Lett. 1996, 193–194. (c) Okamura, H.; Morishige, K.; Iwagawa, T.; Nakatani, M.
Tetrahedron Lett. 1998, 39, 1211–1214.
5. All the reactions summarized in Table 1 afforded endo-isomer 5b exclusively. The enantiomeric excess of 5b was
determined on the basis of the 1H NMR signals corresponding to H-6 of the (−)- and (+)-enantiomers, which separately
appeared at 3.62 and 3.54 ppm, respectively, by measurement in a saturated CDCl3 solution of (S)-(−)-binaphthol.

4150
22
6. Spectra data for 10: [α]D −51 (c 0.59, EtOH), lit.3a [α]D −51.7 (c 0.5, EtOH); IR (neat): 3283, 2951, 2913, 1597,
1582,1483, 1454, 1235, 1028, 754, 700 cm−1: 1H NMR (400 MHz, d6-DMSO with a few drops of CD3CO2D): δ 0.88 (d,
J=6.2 Hz, 3H), 1.30 (m, 1H), 1.54 (d, J =12.1 Hz, 1H), 1.75 (d, J=13.6 Hz, 1H), 2.17 (t, J=12.1 Hz, 2H), 2.65 (m, 1H),
2.73 (d, J=9.9 Hz, 1H), 2.80–2.88 (m, 2H), 2.93 (t, J=9.5 Hz, 1H), 3.18–3.23 (m, 1H), 3.80 (s, 3H), 4.10 (m, 2H), 6.95 (m,
2H), 7.23 (br t, J=7.7 Hz, 1H), 7.35–7.55 (m, 5H), 7.65 (br d, J=7.7 Hz, 1H); 13C NMR (100 MHz): δ 156.1, 128.4, 128.2,
127.6, 127.1, 120.5, 111.1, 74.5, 59.5, 56.9, 56.1, 55.0, 44.6, 41.5, 36.6, 34.5, 22.1, 21.6; HRMS m/z 352.2269 ([M+H]+
calcd for C23H29O2NH+: 352.2277).
7. In the 1H NMR spectrum of 10, the chemical shifts of the signals corresponding to the N–CH group were varied with the
amount of CD3CO2D.