Concise synthesis of (+)-allo-kainic acid via MgI2-mediated tandem aziridine ring opening-formal [3 + 2] cycloaddition

Article abstract of DOI:10.1021/ol4020333

3-Methyl vinyl aziridine undergoes a mild MgI₂-promoted S _N2′ ring opening and concomitant cyclization with fumarate Michael acceptors to give trisubstituted pyrrolidines. The process is efficient and highly diastereoselective. This methodology has been applied to a concise asymmetric synthesis of (+)-allo-kainic acid.

Full text of DOI:10.1021/ol4020333

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Concise Synthesis of (þ)-allo-Kainic Acid
via MgI2‑Mediated Tandem Aziridine Ring
OpeningꢀFormal [3 þ 2] Cycloaddition
Giada Arena, C. Chun Chen, Daniele Leonori,* and Varinder K. Aggarwal*
School of Chemistry, University of Bristol, Cantock’s Close, Bristol, BS8 1TS, U.K.
daniele.leonori@bristol.ac.uk; v.aggarwal@bristol.ac.uk
Received July 19, 2013
ABSTRACT
3-Methyl vinyl aziridine undergoes a mild MgI2-promoted SN20 ring opening and concomitant cyclization with fumarate Michael acceptors to give
trisubstituted pyrrolidines. The process is efficient and highly diastereoselective. This methodology has been applied to a concise asymmetric
synthesis of (þ)-allo-kainic acid.
The kainoids are a class of natural, nonproteinogenic
amino acids with interesting structural and biological
properties (Figure 1).1 They are characterized by a densely
functionalized pyrrolidine ring bearing three contiguous
stereogenic centers and two carboxylic acid units. These
structural features make them conformationally restricted
analogues of glutamic acid, and as such they have been
extensively used in studies of several neurological diseases
such as Huntington corea2 and Alzheimer’s disease.3 Their
significant biological activity, coupled with an alarming
global shortage,4 has fueled intense interest in the synthesis
of this class of compounds.5 Among the kainoids, (þ)-allo-
kainic acid 16 has received considerably less attention than
its C3ꢀC4 diastereomer, (ꢀ)-kainic acid.7,8
(1) (a) Kainic acid as a Tool in Neurobiology; McGeer, E. G., Oleny,
J. W., McGeer, P. L., Eds.; Raven: New York, 1978. (b) Excitatory Amino
Acids; Simon, P. R., Ed.; Thieme Medical: New York, 1992. (c) Moloney,
M. G. Nat. Prod. Rep. 1990, 16, 485. (d) Moloney, M. G. Nat. Prod. Rep.
1998, 15, 205. (e) Glutamate: Transmitter in Central Nervous Systems;
Roberts, P. J., Storm-Mathesen, J., Johnston, G. A. R., Eds.; Wiley: Chich-
ester, U.K., 1981. (f) Goldberg, O.; Luini, A.; Teichberg, V. I. J. Med. Chem.
1983, 24, 939. (g) Johnson, R. L.; Koerner, J. F. J. Med. Chem. 1988,
31, 2057.
(2) (a) Coyle, J. T.; Schwarcz, R. Nature 1976, 263, 244. (b) McGeer,
E. G.; McGeer, P. L. Nature 1976, 263, 517.
(3) Sperk, G. Prog. Neurobiol. 1994, 42, 1.
(4) (a) Tremblay, J.-F. Chem. Eng. News 2000, March 6, 131. (b)
Tremblay, J.-F. Chem. Eng. News 2000, Jan 3, 131.
(5) For reviews on kainoid syntheses, see: (a) Parsons, A. F. Tetra-
hedron 1996, 52, 4149. (b) Moloney, M. G. Nat. Prod. Rep. 2002, 19, 597.
(6) For the isolation of 1 from Digenea simplex, see: (a) Murakami,
S.; Takemoto, T.; Shimizu, Z. J. Pharm. Soc. Jpn. 1953, 73, 1026. (b)
Marimoto, H. J. Pharm. Soc. Jpn. 1975, 75, 766. (c) Murakami, S.;
Daigo, K.; Takagi, N.; Takemoto, T.; Tei, Z. J. Pharm. Soc. Jpn. 1955,
75, 1252. (d) Watase, H. Bull. Chem. Soc. Jpn. 1958, 31, 932. (e) Nitta, I.;
Watase, H.; Tomiie, Y. Nature 1958, 181, 761.
Figure 1. Structures of kainoids. The red color highlights the
analogy with L-glutamic acid.
(7) Ene reaction: (a) Oppolzer, W.; Robbiani, C. Helv. Chim. Acta
1980, 63, 2010. (b) Oppolzer, W.; Robbiani, C.; Battig, K. Helv. Chim.
Acta 1980, 63, 2015. (c) Oppolzer, W.; Robbiani, C.; Battig, K. Tetra-
hedron 1984, 40, 1391. Ni-catalyzed cyclization: (d) Montgomery, J.;
Chevliakov, M. V. Angew. Chem., Int. Ed. 1998, 37, 3144. Allyl silane
cyclization: (e) Mooiweer, H. H.; Hiemstra, H.; Speckamp, W. N.
Tetrahedron 1991, 47, 3451. Isonitrile cyclization: (f) Murakami, M.;
Hasegawa, N.; Hayashi, M.; Ito, Y. J. Org. Chem. 1991, 56, 7356. Allyl-Pd:
(g) Cook, G. R.; Sun, L. Org. Lett. 2004, 6, 2481.
(8) [3 þ 2] Cycloaddition: (a) Kraus, G. A.; Nagy, J. O. Tetrahedron
1985, 41, 3537. (b) DeShong, P.; Kell, D. A. Tetrahedron Lett. 1986, 27,
3979. Aza-Cope: (c) Agami, C.; Cases, M.; County, F. J. Org. Chem.
1994, 59, 7937. (d) Agami, C.; County, F.; Lin, J.; Mikaeloff, A.;
Poursoulis, M. Tetrahedron 1993, 49, 7239.
r
10.1021/ol4020333
XXXX American Chemical Society

As outlined in Scheme 1A, these synthetic efforts can be
divided into “single bond disconnections” and “multiple
bond disconnections”. The former approaches are based
on the formation of the pyrrolidine ring and control of the
three contiguous stereocenters while forming a single CꢀC
bond on an advanced intermediate.7 The latter, and more
direct, approaches assemble the highly substituted pyrro-
lidine ring by forming two CꢀC bonds simultaneously.8
Upon inspection of these methods we were intrigued that
attention has been almost exclusively given to the discon-
nection across the C2ꢀC3 and C4ꢀC5 bonds.9
in achieving a short, stereocontrolled total synthesis of 1
using this strategy.
In our retrosynthetic analysis we envisioned the discon-
nection of 1 via a single key step (Scheme 1C). We reasoned
that opening of vinyl aziridine 312 and concomitant annu-
lation with a suitable fumarate derivative 4 would lead to
the pyrrolidine ring with the functionality required for the
kainic acids and the correct stereochemistry for (þ)-allo-
kainic acid 1. From pyrrolidine 5, a single functional group
manipulation, namely the ArndtꢀEistert homologation, and
subsequent deprotection would complete the total synthesis.
Initial efforts at promoting the reaction between aziridine
3and Michael acceptors 4a (R=H) and4b (R=OEt) using
our previously optimized conditions {[Pd2(dba)3 CHCl3],13
3
Scheme 1
P(furyl)3, and TBAC14 in THF},9 however, were fruitless
(Table 1, entries 1ꢀ2). The reactions were usually charac-
terized by complete decomposition of 3 and quantitative
recovery of the acceptor. These observations indicated that
the activation of the aziridine was indeed occurring but that
the Michael acceptor was not reactive enough to undergo
the addition process.15 We reasoned that the use of a more
activating group such as a thioester or an oxazolidinone
would promote the initial nucleophilic attack. Pleasingly,
when thioester 4c was used, the reaction gave a mixture
of diastereoisomers 5cAꢀC in good yield but with
poor diastereocontrol (entry 3). The diastereoisomers were
(12) Aziridine 3 can be obtained in four steps from serine methyl ester
hydrochloride; see: (a) Smith, A. B., III; Kim, D.-S. J. Org. Chem. 2006, 71,
2547. Alternatively, we have developed a one-step, regioselective procedure
from isoprene (see Supporting Information). For some relevant work on
the aziridination of unsymmetrical dienes, see: (b) Ho, C.-M.; Lau, T.-C.;
Kwong, H.-L.; Wong, W.-T. J. Chem. Soc., Dalton Trans. 1999, 2411. (c)
Minakata, S.; Ando, T.; Nishimura, M.; Ryu, I.; Komatsu, M. Angew.
Chem., Int. Ed. 1998, 37, 3392. (d) Nishimura, M.; Minakata, S.; Takahashi,
T.; Oderaotoshi, Y.; Komatsu, M. J. Org. Chem. 2002, 67, 2101. (e)
Nishimura, M.; Minakata, S.; Thongchant, S.; Ryu, T.; Komatsu, M.
Tetrahedron Lett. 2000, 41, 7089. (f) Knight, J. G.; Muldowney, M. P.
Synlett 1995, 949. (g) Piangiolino, C.; Gallo, E.; Caselli, A.; Fantauzzi, S.;
Ragaini, F.; Cenini, S. Eur. J. Org. Chem. 2007, 743.
(13) For Pd°-mediated opening of unactivated vinyl aziridines, see:
(a) Butler, D. C. D.; Inman, G. A.; Alper, H. J. Org. Chem. 2000, 65,
5887. (b) Trost, B. M.; Fandrick, D. R. J. Am. Chem. Soc. 2003, 125,
11836. For the Pd°-mediated opening of activated aziridines, see: (c)
Hedley, S. J.; Moran, W. J.; Prenzel, A. H. G. P.; Price, D. A.; Harrity,
J. P. A. Synlett 2001, 1596.
(14) The addition of TBAC is aided to prevent ion pairing of the
amide anion with the cationic Pd complex. For discussions on the effects
of halide additives in palladium chemistry, see: (a) Reference 7. (b)
Reference 10. For a review, see: (c) Fagnou, K.; Lautens, M. Angew.
Chem., Int. Ed. 2002, 41, 26.
We have recently reported the Pd°-mediated annulation
of vinyl aziridines 2 with Michael acceptors in the stereo-
controlled synthesis of substituted pyrrolidines, and we
have applied this methodology to a short formal synthesis
of (ꢀ)-kainic acid (Scheme 1B).10,11 While the methodol-
ogy resulted in rapid construction of the pyrrolidine ring, it
required additional functional group interconversion and
redox chemistry to reach the target. We recognized that a
similar reaction manifold, but with the correct juxtaposi-
tion of functional groups in the two reacting components,
could result in formation of the kainoids with minimal
downstream manipulation. Here, we describe our success
(15) For a tandem process of vinyl aziridines with doubly activated
Michael acceptors, see: (a) Shim, J. G.; Yamamoto, Y. J. Org. Chem.
1998, 63, 3067. For a review, see: (b) Wolfe, J. P.; Hay, M. B.
Tetrahedron 2007, 63, 261. For related reactions on hydroxy allylic
carbonates, see: (c) Sekido, M.; Aoyagi, K.; Nakamura, H.; Kabuto, C.;
Yamamoto, Y. J. Org. Chem. 2001, 66, 7142. (d) Aoyagi, K.; Nakamura,
H.; Yamamoto, Y. J. Org. Chem. 2002, 67, 5977. For reviews, see: (e)
Patil, N. T.; Yamamoto, Y. Synlett 2007, 1994. (f) Patil, N. T.; Yamamoto,
T. Top. Organomet. Chem. 2006, 19, 91. For related reactions of
oxazolidinones and oxazinanones, see: (g) Knight, J. G.; Tchabanenko,
K.; Stoker, P. A.; Harwood, S. J. Tetrahedron Lett. 2005, 46, 6261. (h)
Knight, J. G.; Stoker, P. A.; Tchabanenko, K.; Harwood, S. J.; Lawrie,
K. W. M. Tetrahedron 2008, 64, 3744. (j) Wang, C.; Tunge, J. A. Org.
Lett. 2006, 8, 3211. (k) Wang, C.; Tunge, J. A. J. Am. Chem. Soc. 2008,
130, 8118. See also: (i) Patil, N. T.; Huo, Z.; Yamamoto, Y. J. Org.
Chem. 2006, 71, 6991. (l) Martinon, L.; Azoulay, S.; Monteiro, N. E.;
(9) For a disconnection across NꢀC2 and C3ꢀC4 bonds, see: Barco,
A.; Benetti, S.; Casolari, A.; Pollini, G. P.; Spalluto, G. Tetrahedron Lett.
1990, 31, 4917.
(10) Lowe, M. A.; Ostovar, M.; Ferrini, S.; Chun Chen, C.; Lawrence,
P. G.; Fontana, F.; Calabrese, A. A.; Aggarwal, V. K. Angew. Chem., Int.
Ed. 2011, 50, 6370.
(11) For similar work from our group, see: (a) Fontana, F.; Chun
Chen, C.; Aggarwal, V. K. Org. Lett. 2011, 13, 3454. (b) Fontana, F.;
Tron, G. C.; Barbero, N.; Ferrini, S.; Thomas, S.; Aggarwal, V. K.
Chem. Commun. 2010, 46, 267.
€
Kundig, P.; Balme, G. J. Organomet. Chem. 2004, 689, 3831. For a
recent report on the ring expansion of vinyl aziridines to 3-pyrrolines,
see: (m) Brichacek, M.; Villalobos, M. N.; Plichta, A.; Njardarson, J. T.
Org. Lett. 2011, 13, 1110.
B
Org. Lett., Vol. XX, No. XX, XXXX

Lewis acids was however unsuccessful, and no product was
obtained (entry 7).20
Table 1. Optimization of the Tandem Aziridine
OpeningꢀFormal [3 þ 2] Cycloaddition Sequence
These unproductive avenues prompted us to investigate
the use of a different mechanistic pathway for the implementa-
tion of this tandem process. We envisaged the use of a
bifunctional activator of generic structure MꢀX where M
(metal) would display sufficient Lewis acidity to activate the
carbonyl group (via coordination) and X would display appro-
priate nucleophilicity to open the aziridine regioselectively
(via an SN20 process). We were particulary intrigued by early
reports from Carreira21 and Lautens22 where MgI2 has been
found to be a competent electrophilic/nucleophilic promoter in
the opening of cyclopropanes.23 We thus speculated that
addition of equimolar amounts of MgI2 to our reaction would
both open the aziridine and activate the fumarate derivative.
Our plan was not without potential problems since MꢀX salts
(M = Mg, Li, In; X = I, Br, Cl) have been reported to open
(vinyl)-aziridines at the less substituted position.24
Scheme 2
a Yield after column cromatography for the mixture of diastereomers.
b Determined by 1H NMR spectroscopy of the crude mixture. c Determined
by chiral SFC. d Determined after removal of the chiral auxiliary. e 6 was
used instead of P(furyl)3. f Yield for the 5dA. g 0.1 equiv of MgI2 was used.
Scheme 3
separated, and their structures were elucidated on the basis
of their characteristic J values and NOE experiments.16 At
this point the reaction parameters were further explored.
The use of alternative halides was not effective (entry 4).17
Attempts to control both the relative and absolute stereo-
chemistry were explored initially using chiral ligands. How-
ever, the Trost ligand 6gave the product in 52% yield but with
similar dr and without any enantioinduction (entry 5).18 We
then explored chiral auxiliaries. When oxazolidinone-based
acceptor 4d was used, the formation of diastereomer 5dC
was suppressed but a 1:1 mixture of 5dA and 5dB isomers
was obtained but again with no control from the chiral
auxiliary (i.e., after removal of the auxiliary the product was
racemic) (entry 6). We believed that the poor diastereocon-
trol from the auxiliary was due to two competing isoener-
getic pathways arising from the fast interconversion of the
syn-s-cis and anti-s-cis conformations around the auxiliary
(Scheme 3A).19 Attempts to control this using chelating
While reactions with fumarate derivatives 4a and 4c
were not successful, they did give the aziridine ring-opened
product 7 showing that the SN20 with MgI2 had occurred in
(20) Lewis acids screened: Et2AlCl, Sc(OTf)3, AgSbF6, Mg(OEt)2.
(21) Alper, P. B.; Meyers, C.; Lerchner, A.; Siegel, D. R.; Carreira,
E. M. Angew. Chem., Int. Ed. 1999, 38, 3186.
(22) (a) Lautens, M.; Han, W. J. Am. Chem. Soc. 2002, 124, 6312. (b)
Scott, M. E.; Han, W.; Lautens, M. Org. Lett. 2004, 6, 3309. (c) Scott,
M. E.; Lautens, M. Org. Lett. 2005, 7, 3045.
(23) (a) Wales, S. M.; Walker, M. M.; Johnson, J. S. Org. Lett. 2013,
15, 2558. (b) Parsons, A. T.; Smith, A. G.; Neel, A. J.; Johnson, J. S.
J. Am. Chem. Soc. 2010, 132, 9688.
(24) (a) Righi, G.; Potini, C.; Bovicelli, P. Tetrahedron Lett. 2002,
43, 5867. (b) Ray, C. A.; Risberg, E.; Somfai, P. Tetrahedron Lett. 2001,
42, 9289. (c) Yadav, J. S.; Reddy, B. V. S.; Kumar, G. M. Synlett 2001,
1417. For a review, see: (d) Hu, X. E. Tetrahedron 2004, 60, 2701.
(16) 5A: J(HAꢀHB) ≈ 8.2 Hz; J(HBꢀHC) ≈ 10.5 Hz. 5B: J(HAꢀHB) ≈
1.3 Hz; J(HBꢀHC) ≈ 7.0 Hz. 5C: J(HAꢀHB) ≈ 8.2 Hz; J(HBꢀHC) ≈ 11.5
Hz. See the Supporting Information for HPLC conditions.
(17) The use of TBAT did not provide any product.
(18) Several commercially available ligands have been screened, but
they were generally ineffective in controlling the diastereoselectivity.
(19) For a discussion on chiral oxazolidinones’ conformations in
similar systems, see: Davies, S. G.; Herman, G. J.; Sweet, M. J.; Smith,
A. D. Chem. Commun. 2004, 1128.
Org. Lett., Vol. XX, No. XX, XXXX
C

conformation (minimizing nonbonded interactions) results
in the observed stereochemistry.
Scheme 4
The complete total synthesis of (þ)-allo-kaininc acid
(þ)-1 is shown in Scheme 4. Having found the appropriate
conditions to access pyrrolidine 5dA, the auxiliary was
removed by base hydrolysis to afford acid 8. Treatment of
8 with oxalyl chloride and TMS-diazomethane gave the
corresponding diazoketone in 63% yield. This compound
proved to be stable under silica gel chromatography and
could be easily purified. Sonication of this intermediate in
the presence of silver benzoate promoted the required
ArndtꢀEistert homologation. The resulting diester did
not require any further purification, and the two final
deprotections could be performed on the crude material.
Thus, saponification of the ethyl esters and removal of the
tosyl group gave (þ)-allo-kaininc acid (þ)-1 {[R]D þ7.1
(c 0.1, H2O); Lit. [R]D þ7.7 (c 0.2, H2O)}.7a
In summary, we have developed a powerful annulation
method for combining readily accessible vinyl aziridines with
Evans’ fumarates for the stereocontrolled synthesis of densely
functionalized pyrrolidines. The methodology has been ap-
plied to a concise asymmetric synthesis of (þ)-allo-kainic acid.
the desired manner (entries 8ꢀ10 and Scheme 2). Pleas-
ingly oxazolidinone 4d showed the right charge-affinity
pattern, and the 2,3,4-trisubstituted pyrrolidine product
was obtained in 63% yield, 9:1 dr, (favoring 5dA) and
essentially complete enantioselectivity (entry 11). Oxazo-
lidinone 4e has been reported to be superior to 4d in
asymmetric 1,4-nucleophilic additions,25 but in this case
only a low diastereoselectivity was obtained (entry 12).
The stereochemical model for this new annulation reac-
tion is provided in Scheme 3B and is based on attack of the
ring-opened aziridine on the more accessible Si face of the
chelated syn-s-cis conformation of 4d.26 Following SN20-
type cyclization of the intermediate enolate in the all-anti
Acknowledgment. We thank EPSRC and the European
Research Council (FP7/2007-2013, ERC Grant No.
ꢀ
246785) for financial support. GA thanks “Universita
degli studi di Siena” for funding. C.C.C. thanks the Uni-
versity of Bristol for a Centenary Scholarship.
Supporting Information Available. Synthesis and char-
acterization of all new compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
(25) Williams, D. R.; Kissel, W. S.; Li, J. J. Tetrahedron Lett. 1998,
39, 8593.
(26) Martinelli, M. J. J. Org. Chem. 1990, 55, 5065.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX