Synthesis of hydroxy-α-sanshool

Article abstract of DOI:10.1055/s-0032-1317172

The amide hydroxy-α-sanshool is responsible for the mild numbing sensation experienced when Sichuan (or Szechuan) peppercorns (huajiao) are eaten. It has been synthesized in six steps from simple commercially available starting materials using Wittig reactions as the key transformations for construction of the carbon skeleton. The penultimate synthetic intermediate was 2E,6Z,8E,10E-dodecatetraenoic acid, and its crystalline nature allowed it, and thus hydroxy-α-sanshool, to be purified to a very high level of stereochemical homogeneity.

Full text of DOI:10.1055/s-0032-1317172

2564▌
LETTER
lSetyternthesis of Hydroxy-α-sanshool
Synthesis of Hydroxy-α-sanshool
Bo Wu, Kun Li, Patrick H. Toy*
Department of Chemistry, The University of Hong Kong, Pokfulam Road, Hong Kong, P. R. of China
Fax +852(2)8571586; E-mail: phtoy@hku.hk
Received: 03.07.2012; Accepted after revision: 03.08.2012
of stereoselective Wittig reactions as the key carbon-con-
necting steps.
Abstract: The amide hydroxy-α-sanshool is responsible for the
mild numbing sensation experienced when Sichuan (or Szechuan)
peppercorns (huajiao) are eaten. It has been synthesized in six steps
from simple commercially available starting materials using Wittig
reactions as the key transformations for construction of the carbon
skeleton. The penultimate synthetic intermediate was
2E,6Z,8E,10E-dodecatetraenoic acid, and its crystalline nature al-
lowed it, and thus hydroxy-α-sanshool, to be purified to a very high
level of stereochemical homogeneity.
O
N
H
α-sanshool
O
Key words: natural products, isomers, alkenes, Wittig reaction, al-
dehydes
N
H
β-sanshool
O
The sanshools are a group of polyunsaturated amides
found in various zanthoxylum species (Figure 1).1 Nota-
bly, hydroxy-α-sanshool (1; Scheme 1) is responsible for
the mild numbing/tingling sensation one experiences
when eating Sichuan (or Szechuan) peppercorns known as
huajiao.2a Recently its mechanism of action has been
studied, and it has been found to excite sensory neurons
through inhibition of two-pore potassium channels.2 This
observation has generated interest in synthesizing and
studying analogues of 1,3 since its isolation is tedious and
low-yielding.4 In light of the importance of such biologi-
cal activity,5 the interesting physiological effects of 1,6
and its potential utility in the cosmetics7 and food8 indus-
tries, we sought to develop an efficient synthesis that
would allow for easy access to not only 1, but also ana-
logues for structure–activity relationship studies. Herein,
we report the realization of this objective and describe
what is, to the best of our knowledge, the first reported de
novo synthesis of 1.
N
H
γ-sanshool
O
N
H
δ-sanshool
Figure 1 Sanhool compounds
O
N
H
OH
hydroxy-α-sanshool (1)
O
O
H2N
OH
OH
+
Our retrosynthetic analysis for 1 is outlined in Scheme 1.
Carboxylic acid 29 was targeted as the key intermediate
since it could be used to prepare not only 1, but also a
library of amide analogues. Thus, the final step would in-
volve coupling acid 2 with amine 3, which can be pre-
pared by reaction of isobutylene oxide (4) with
benzylamine followed by debenzylation,3 to afford 1. It
was envisioned that a Wittig reaction could be used to
synthesize an ester of 2 by reaction of a phosphorane de-
rived from 5 with sorbaldehyde (6), and that ester 5 could,
in turn, be prepared in two steps from commercially avail-
able building blocks 4-bromobutanol (7) and phosphorane
8. Thus, our plan was to synthesize 2 from 6–8 using a pair
4
3
2
O
O
Br
+
OMe
H
sorbaldehyde (6)
5
7
O
Br
Ph3P
OH
+
OMe
8
Scheme 1 Retrosynthetic analysis for hydroxy-α-sanshool (1)
The successful realization of our synthetic plan is present-
ed in Scheme 2.10 Oxidation of alcohol 7 using PCC af-
forded aldehyde 9 in 76% yield, and subsequent reaction
SYNLETT 2012, 23, 2564–2566
Advanced online publication: 13.09.2012
DOI: 10.1055/s-0032-1317172; Art ID: ST-2012-W0564-L
© Georg Thieme Verlag Stuttgart · New York
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6

LETTER
Synthesis of Hydroxy-α-sanshool
2565
of 9 with phosphorane 8 afforded enoate 5 in 91% yield synthesis of 1, and envision that easy access to large quan-
with predominate E-stereoselectivity (92:8). This was fol- tities of 2 will allow the preparation of a broad range of
lowed by treatment of 5 with Ph3P to afford the corre- amide analogues of 1 for use in structure–activity relation-
sponding phosphonium salt 10 in 96% yield. Salt 10 was ship studies. Results of this ongoing research program
then mixed with K2CO3 and aldehyde 6 to effect the sec- will be reported shortly.
ond Wittig reaction, which afforded 11 in 80% yield with
approximate 2:1 Z/E-stereoselectivity.11 Ester 11 was then
Acknowledgment
converted into 2 in 43% yield after recrystallization.
This research was supported financially by the University of Hong
Kong and the Research Grants Council of the Hong Kong S. A. R.,
P. R. of China (Project No. HKU 705209P).
Finally, coupling of 2 with 3 (prepared according to the
literature procedure from 43) with HBTU and Et3N afford-
ed 1 in 92% yield. Thus, 1 was synthesized expediently in
21% overall yield in a process that required six reactions
in the longest linear sequence.12 It should be noted that
Supporting Information for this article is available online at
commercially available 6 is labeled as being ‘predomi- http://www.thieme-connect.com/ejournals/toc/synlett.SupIpnoifgoit
rmnartSnpufIormigiotnat
nantly trans,trans’, and is approximately a 5:1 mixture of
isomers according to 1H NMR analysis. This is one of the
major reasons for the relatively modest yield of 2. Consid-
ering this fact and the amounts of undesired alkene iso-
mers formed in the two Wittig reactions, our achieved
yield of 2 is quite respectable. Fortunately, nearly all of
the undesired stereochemical isomers due to the nature of
6 used as the starting material, and formed in the Wittig
reactions, could be removed by the recrystallization of 2
(ca. 96% isomerically pure).
References and Notes
(1) (a) Yasuda, I.; Takeya, K.; Itokawa, H. Phytochemistry
1982, 21, 1295; and references cited therein. (b) Mitzutani,
K.; Fukunaga, Y.; Tanaka, O.; Takasugi, N.; Saruwatari, Y.-
I.; Fuwa, T.; Yamauchi, T.; Wang, J.; Jia, M.-R.; Li, F.-Y.;
Ling, Y.-K. Chem. Pharm. Bull. 1988, 36, 2362.
(c) Kashiwada, Y.; Ito, C.; Katagiri, H.; Mase, I.; Komatsu,
K.; Namba, T.; Ikeshiro, Y. Phytochemistry 1997, 44, 1125.
(d) Xiong, Q.; Shi, D.; Yamamoto, H.; Mizuno, M.
Phytochemistry 1997, 46, 1123. (e) Chen, I.-H.; Chen, T.-L.;
Lin, W.-Y.; Tsai, I.-L.; Chen, Y.-C. Phytochemistry 1999,
52, 357. (f) Jang, K. H.; Chang, Y. H.; Kim, D.-D.; Oh, K.-
B.; Oh, U.; Shin, J. Arch. Pharm. Res. 2008, 31, 569.
(2) (a) Sugai, E.; Morimitsu, Y.; Iwasaki, Y.; Morita, A.;
Watanabe, T.; Kubota, K. Biosci. Biotechnol. Biochem.
2005, 69, 1951. (b) Koo, J. Y.; Jang, Y.; Cho, H.; Lee, C.-H.;
Jang, K. H.; Chang, Y. H.; Shin, J.; Oh, U. Eur. J. Neurosci.
2007, 26, 1139. (c) Bautista, D. M.; Sigal, Y. M.; Milstein,
A. D.; Garrison, J. L.; Zorn, J. A.; Tsuruda, P. R.; Nicoll, R.
A.; Julius, D. Nature Neurosci. 2008, 11, 772. (d) Riera, C.
E.; Menozzi-Smarrito, C.; Affolter, M.; Michlig, S.; Munari,
C.; Robert, F.; Vogel, H.; Simon, S. A.; le Coutre, J. Br. J.
Pharmacol. 2009, 157, 1398.
O
PCC, CH2Cl2, r.t.
Br
7
H
9
8, CH2Cl2, r.t.
Ph3P, MeCN, reflux
5
O
6, K2CO3, PhMe, 70 °C
Br
Ph3P
OMe
10
O
1. NaOH, H2O, 70 °C
2. 1 M HCl, r.t.
2
(3) Menozzi-Smarrito, C.; Riera, C. E.; Munari, C.; le Coutre, J.;
Robert, F. J. Agric. Food Chem. 2009, 57, 1982.
OMe
(4) According to Bautista et al.,2c 50 g of dried seeds from
Zanthoxylum piperitum afforded 55.2 mg of crude 1 after
preparative HPLC. Repetitive chromatographic separation
was required to further purify 1 to homogeneity
11
3, HBTU, Et3N
1
CH2Cl2, r.t.
(5) (a) Baraldi, P. G.; Preti, D.; Materazzi, S.; Geppetti, P. J.
Med. Chem. 2010, 53, 5085. (b) Mathie, A. J. Pharm.
Pharmacol. 2010, 62, 1089. (c) Es-Salah-Lamoureux, Z.;
Steele, D. F.; Fedida, D. Trends Pharmacol. Sci. 2010, 31,
587.
1. BnNH2, Et3N, H2O, r.t.
2. H2, Pd/C, MeOH, r.t.
4
3
Scheme 2 Synthesis of hydroxy-α-sanshool (1)
(6) (a) Sawyer, C. M.; Carstens, M. I.; Simons, C. T.; Slack, J.;
McCluskey, T. S.; Furrer, S.; Carstens, E. J. Neurophysiol.
2009, 101, 1742. (b) Lennertz, R. C.; Tsunozaki, M.;
Bautista, D. M.; Stucky, C. L. J. Neurosci. 2010, 30, 4353.
(c) Albin, K. C.; Simons, C. T. PLoS One 2010, 5, e9520.
(d) Klein, A. H.; Sawyer, C. M.; Zanotto, K. L.; Ivanov, M.
A.; Cheung, S.; Carstens, M. I.; Furrer, S.; Simons, C. T.;
Slack, J. P.; Carstens, E. J. Neurophysiol. 2011, 105, 1701.
(7) Artaria, C.; Maramaldi, G.; Bonfigli, A.; Rigano, L.;
Appendino, G. Int. J. Cosmetic Sci. 2011, 33, 328.
(8) Starkenmann, C.; Cayeux, I.; Birkbeck, A. A. Chimia 2011,
65, 407.
In conclusion, we report a short, high-yielding synthesis
of 1 from four simple and inexpensive building blocks
that involves two stereoselective Wittig reactions as the
key transformations for construction of the carbon skele-
ton. Importantly, penultimate intermediate 2 can be re-
crystallized to
a
high level of stereochemical
homogeneity, and thus the undesired diastereomers result-
ing from starting material 6 and formed in the Wittig reac-
tions can be readily removed. We are currently
investigating the application of our previously reported
methodology for facilitating Wittig reactions13 to the
(9) For what is, to our knowledge, the only previously reported
synthesis of 2 (and α-sanshool) by a different route, see:
Sonnet, P. E. J. Org. Chem. 1969, 34, 1147.
© Georg Thieme Verlag Stuttgart · New York
Synlett 2012, 23, 2564–2566

2566
B. Wu et al.
LETTER
(10) See the Supporting Information for details.
6.03 (dd, J = 10.8, 10.8 Hz, 1 H), 6.19–6.06 (m, 2 H), 6.27–
(11) This stereoselectivity is typical of such Wittig reactions
involving unstabilized ylides, and we are currently exploring
the purification of the E-isomer so that it can be used in the
synthesis of β-sanshool derivatives.
6.33 (m, 2 H), 6.80–6.87 (dt, J = 13.2, 6.4 Hz, 1 H); 13
C
NMR (100 MHz, CDCl3): δ = 18.4, 26.5, 27.2, 32.1, 50.6,
70.8, 124.0, 125.2, 129.4, 129.7, 130.2, 131.8, 133.6, 144.1,
167.2. This matches the data previously reported for material
isolated from natural sources.1a
(12) Characterization data for synthetic 1: 1H NMR (400 MHz,
CDCl3): δ = 1.21 (s, 6 H), 1.75–1.77 (d, J = 7.2 Hz, 3 H),
2.23–2.35 (m, 4 H), 3.13 (s, 1 H), 3.29–3.31 (d, J = 6.0 Hz,
2 H), 5.31–5.37 (dt, J = 10.4, 7.4 Hz, 1 H), 5.67–5.75 (dq, J
= 14.0, 7.2 Hz, 1 H), 5.83–5.87 (d, J = 15.2 Hz, 1 H), 5.97–
(13) (a) Leung, P. S.-W.; Teng, Y.; Toy, P. H. Synlett 2010, 1997.
(b) Leung, P. S.-W.; Teng, Y.; Toy, P. H. Org. Lett. 2010,
12, 4996. (c) Teng, Y.; Lu, J.; Toy, P. H. Chem. Asian J.
2012, 7, 351.
Synlett 2012, 23, 2564–2566
© Georg Thieme Verlag Stuttgart · New York

Copyright of Synlett is the property of Georg Thieme Verlag Stuttgart and its content may not be copied or
emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.