Synthesis of α-methyl kainic acid by stereospecific lithiation-dearomatizing cyclization of a chiral benzamide

Article abstract of DOI:10.1016/S0040-4039(03)00570-7

Stereospecific lithiation of N-α-methylbenzyl benzamides gives configurationally stable tertiary benzyllithiums which undergo a stereospecific dearomatizing cyclization with >99% retention of stereochemistry. The products are partially saturated isoindolinones which carry a new fully-substituted stereogenic centre. A ten-step sequence converts one of these products to the α-methyl analogue of kainic acid.

Full text of DOI:10.1016/S0040-4039(03)00570-7

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 3397–3400
Synthesis of a-methyl kainic acid by stereospecific
lithiation–dearomatizing cyclization of a chiral benzamide
Jonathan Clayden,* Faye E. Knowles and Christel J. Menet
Department of Chemistry, University of Manchester, Oxford Road, Manchester M13 9PL, UK
Received 13 January 2003; revised 26 February 2003; accepted 27 February 2003
Abstract—Stereospecific lithiation of N-a-methylbenzyl benzamides gives configurationally stable tertiary benzyllithiums which
undergo a stereospecific dearomatizing cyclization with >99% retention of stereochemistry. The products are partially saturated
isoindolinones which carry a new fully-substituted stereogenic centre. A ten-step sequence converts one of these products to the
a-methyl analogue of kainic acid. © 2003 Elsevier Science Ltd. All rights reserved.
While stereospecific1 nucleophilic substitution at an
electrophilic stereogenic centre is commonplace, reliable
stereospecific electrophilic substitution at a nucleophilic
stereogenic centre is more elusive.2 Some chiral organo-
lithium nucleophiles—most notably those with a-het-
eroatom substituents—can translate stereochemistry at
the CꢀLi bond directly into stereochemistry at a CꢀC
bond. Often, however, stereochemical fidelity is com-
promised either by lack of stereospecificity in the elec-
trophilic substitution step or by lack of configurational
stability at the CꢀLi bond.3 Even when stereospecific
electrophilic substitution is possible, it can be of syn-
thetic value only when the organolithium can be made
in a stereocontrolled manner.
In this paper we describe a stereospecific reaction
sequence which starts with a deprotonation yielding
stereospecifically a configurationally stable tertiary
Scheme 1. Dearomatizing cyclization of an N-a-methylbenzyl benzamide. Reagents and conditions: (i) t-BuLi, THF, −78°C; (ii)
a
warm to 0°C; (iii) p-BrC6H4CH2Br or BnBr or MeI or PhCHO; (iv) HCl, H2O. Yield not determined; bFrom 1.
* Corresponding author.
0040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0040-4039(03)00570-7

3398
J. Clayden et al. / Tetrahedron Letters 44 (2003) 3397–3400
organolithium.4 This organolithium then undergoes a
stereospecific dearomatizing cyclization.5 Overall, the
sequence constructs a fully substituted stereogenic cen-
tre from a tertiary centre by stereospecific replacement
of the CꢀH bond with a CꢀC bond. A further ten
synthetic steps allow the newly formed stereogenic cen-
tre to be incorporated into the a-position of a new
methylated analogue of kainic acid.
quaternary, and simultaneously generates useful reac-
tivity from the benzamide’s aromatic ring. To widen the
scope of the cyclization, we required amides bearing
nitrogen substituents which could be manipulated in the
cyclised products. We chose the diastereoisomeric pairs
of amides meso- (or [R,S]-) and (R,R)-10 because of
their straightforward synthesis from amines 7a and 7b
by the usefully complementary amine/amide syntheses
shown in Scheme 2. Adapting the known synthesis of
(R,R)-10a13 gave (R,R)-10b selectively by diastereose-
lective reduction of the imine 8b, while meso-10a, meso-
10b and (R,S)-10c were made from 12a–c by our
‘meso-selective’ amide a-lithiation–alkylation.14
Chiral and enantiomerically pure amide (+)-1 was made
by acylation of N-isopropyl-a-methylbenzylamine6 and
was lithiated at −78°C with t-BuLi in THF to give a red
intermediate presumed to be 2 (Scheme 1). Warming to
0°C promoted a dearomatizing cyclization to the
extended enolate 3 which reacted with electrophiles to
yield either 4 or a mixture of 4 and 5. Enol ethers 4
were conveniently isolated as the enones 6 after hydroly-
sis with aqueous acid.7 Comparison of the HPLC trace8
of 6e with that of a racemic sample made from ( )-1
showed that 6 had retained >99% ee and that the
lithiation (1“2) and cyclization (2“3) steps therefore
proceeded with complete stereospecificity.9 The X-ray
crystal structure of 4a confirmed that the lithiation–
cyclization sequence (1“3) went with retention of stere-
ochemistry, presumably indicating retention in both
steps.10
All five N-a-methylbenzyl benzamides 10 were deproto-
nated with t-BuLi to give red organolithiums which
cyclized on warming to room temperature. Aqueous
quench and neutral or acidic aqueous work-up gave
good yields of the bicyclic dienes or enones (Scheme 3).
Dienes 13a,b and 15a,b contained no trace of each
other: the cyclizations are still fully stereospecific.
Hydrolysis to the enones 14 and 16 with 0.5 M HCl
gave a single stereo- and regioisomer of 14b.
The ‘meso-selective’ alkylation–cyclization sequence is
devalued by the fact that enantiomerically pure amines
7 necessarily lead to racemic products 15 and 16. We
aimed to overcome this with (R,S)-10c, in which meso
symmetry is broken by the presence of a methoxy
group. Unfortunately, however, the cyclization was not
sufficiently selective with respect to the two possible
cyclizing groups to be synthetically useful, giving a 4:1
mixture of cyclized enones 16c and 16d.15 The major
product 16c arises by cyclization of the methoxy-substi-
tuted a-methylbenzyl group—presumably the p-
methoxy group destabilizes the benzyllithium (and
therefore promotes faster cyclization) more than it dis-
favours deprotonation.
We had previously observed a related stereospecific
cyclization of N-a-methylbenzyl naphthamides,11 but in
that case we showed that the stereospecificity originated
in a chiral memory effect at the stereogenic ArꢀCO
bond.12 In 1 and 2 the ArꢀCO bond cannot be stereo-
genic, so stereospecificity at least in these cyclizations
must be due to stereospecific formation and reaction of
a configurationally stable organolithium intermediate.
The sequence of events running from 1 to 6 makes use
of the single stereogenic centre of 1 to construct the
three or four stereogenic centres of 6, up to two of them
Scheme 2. Diastereoselective amide synthesisa. Reagents and
conditions: (i) Ar2COMe, toluene, reflux; (ii) 1. NaBH4; 2.
crystallise (see Ref. 15); (iii) p-MeOC6H4COCl, Et3N,
Scheme 3. Diastereospecificity in the cyclizationa. Reagents
and conditions: (i) 1. t-BuLi, THF, −78 to +20°C; 2. NH4Cl,
2
,
CH2Cl2; (iv) 1. Ar CHO, MeOH, 4 A sieves, 20°C, 4 h; 2.
a
NaBH4; (v) 1. t-BuLi, THF, −78°C; 2. MeI (see Ref. 16).
a7a–12a Ar1=Ar2=Ph; 7b–12b Ar1=Ar2=p-MeOC6H4; 10c–
12c Ar1=p-MeOC6H4Ph; Ar2=Ph.
(ii) 0.5 M HCl. 13a, 15a Ar1=Ar2=Ph; 13b–16b Ar1=Ar2=
p-MeOC6H4; 16c, 16d Ar1=p-MeOC6H4; Ar2=Ph. bPlus
21% of the b,g-enone regioisomer.

J. Clayden et al. / Tetrahedron Letters 44 (2003) 3397–3400
3399
Nevertheless, the cyclization of (R,R)-10b incorporating
two methoxy-substituted a-methylbenzyl groups offers
several advantages. Schemes 4 and 5 illustrate this in a
practical synthetic application of the enantiomerically
pure cyclized product 14b. Firstly, the exocyclic a-
methyl-p-methoxybenzyl group is susceptible to oxida-
tive cleavage with ceric ammonium nitrate.16 Conjugate
addition of Me2CuLi to 14b gave the ketone 17 quanti-
tatively, and treatment with ceric ammonium nitrate
followed by Boc2O gave the carbamate 19.
p-methoxyphenyl group: its enhanced susceptibility to
oxidation with Ru(VIII) to provide carboxyl sub-
stituents.20 However, oxidative removal of the
methoxyphenyl ring of 19 with Ru(VIII) was only
partially successful—steric hindrance from the methyl
group largely derailed the degradation of the ring one
stop before the target carboxylic acid derivative, giving
(after methylation) mainly 20 and little 21. We returned
to the carbamate 19 and subjected it to Baeyer–Villiger
oxidation with m-CPBA, regioselectively rupturing the
cyclohexanone ring to give lactone 22.21 Careful
methanolysis (NaOMe, −78°C) of the lactone avoided
epimerization a to the amide carbonyl group, giving an
alcohol which was protected as its trichloroacetate 23.
Oxidative degradation of the methoxyphenyl ring at
this stage in the synthesis still stopped short of comple-
tion, but basic hydrogen peroxide was sufficient to
further oxidise any a-ketoacid to the carboxylic acid.
Hydrolysis of the trichloroacetate ester took place con-
currently, and the product was isolated after methyla-
tion with trimethylsilyldiazomethane as the ester 24.
The kainoid isopropenyl group of 25 was revealed by
elimination of water from 24 via a selenoxide; reduction
of the lactam carbonyl group with DIBAL and Et3SiH
gave 26 and deprotection yielded 27, an a-methylated
analogue of kainic acid. Alkyl substituted kainoids
have been made before and their properties investi-
gated, but this is the first synthesis of 27.22
The carbamate 19 exhibits the same absolute and rela-
tive stereochemistry as the kainoid series of amino
acids,17 and we have used related cyclization products
in syntheses of (−)-kainic acid itself18 and of other
members of the kainoid family.19 Conversion of 19 to a
kainoid analogue is assisted by a second property of the
Scheme 4. Oxidative deprotection. Reagents and conditions:
(i) Me2CuLi·LiBr, Me3SiCl, THF, −78°C; (ii) 2 M HCl; (iii)
Ce(NH4)2(NO3)6, H2O, MeCN; (iv) Boc2O, Et3N, DMAP,
CH2Cl2, 12 h.
Supplementary material
X-Ray crystallographic data for 4a and experimental
details for the synthesis of 27.
Acknowledgements
We are grateful to the EPSRC, Aventis Cropscience
(Lyon) and GlaxoSmithKline for support, to Dr Dar-
ren Mansfield and Dr Ian Baldwin for helpful discus-
sions, and to Dr Madeleine Helliwell for the X-ray
crystal structure of 4a.
References
1. In the sense of Zimmermann. See: Zimmerman, H. E.;
Singer, L.; Thyagarajan, B. S. J. Am. Chem. Soc. 1959,
81, 108.
2. (a) March, J. Advanced Organic Chemistry; 4th ed.;
Wiley: New York, 1992; (b) Gawley, R. E. Tetrahedron
Lett. 1999, 40, 4297; (c) Gawley, R. E.; Low, E.; Zhang,
Q.; Harrris, R. J. Am. Chem. Soc. 2000, 122, 3344.
3. For discussions, see: (a) Clayden, J. Organolithiums:
Selectivity for Synthesis; Pergamon: Oxford, 2002; (b)
Basu. A.; Thayumanavan, S. Angew. Chem., Int. Ed.
2002, 41, 716.
4. Stereospecific deprotonation (with no external source of
stereocontrol) yielding configurationally stable organo-
lithiums is relatively rare [most examples generate tertiary
N- or O-substituted organolithiums. See: (a) Faibish, N.
C.; Park, Y. S.; Lee, S.; Beak, P. J. Am. Chem. Soc. 1997,
Scheme 5. Synthesis of a-methyl kainic acid. Reagents and
conditions: (i) RuCl3, NaIO4, H2O, MeCN, EtOAc, 4 h; (ii)
Me3SiCHN2; (iii) m-CPBA, CH2Cl2, 48 h (80%); (iv) NaOMe,
MeOH, −78°C, 2 h (82%); (v) Cl3CCOCl, Et3N, CH2Cl2, 0°C
(85%); (vi) H2O2, Na2CO3, dioxane (35% 24 from 23); (vii)
N-PhSe-phthalimide, n-Bu3P, THF, 20°C, 2 h; (viii) H2O2,
py, THF, 20°C, 12 h (64% from 24); (ix) DIBAL, THF,
−78°C, 2 h; (x) Et3SiH, BF3·Et2O, CH2Cl2, −78°C (41% from
25); (xi) LiOH, MeOH, 12 h, 20°C; (xii) CF3CO2H, reflux, 2
h (45% from 26).

3400
J. Clayden et al. / Tetrahedron Letters 44 (2003) 3397–3400
119, 11561; (b) Hara, O.; Ho, M.; Hamada, Y. Tetra-
dearomatizing cyclization of a secondary organolithium.
See: Clayden, J.; Menet, C. J.; Mansfield, D. J. Chem.
Commun. 2002, 38.
hedron Lett. 1998, 39, 5537; (c) Kra¨mer, T.; Hoppe, D.
Tetrahedron Lett. 1987, 28, 5149; (d) Hoppe, D.;
Carstens, A.; Kra¨mer, T. Angew. Chem., Int. Ed. Engl.
1990, 29, 1424; (e) Carstens, A.; Hoppe, D. Tetrahedron
1994, 50, 6097; (f) Derwing, C.; Hoppe, D. Synthesis
1996, 149; (g) Hammerschmidt, F.; Hanninger, A. Chem.
Ber. 1995, 128, 1069 but also (h) Hoppe, D.; Paetow, M.;
Hintze, F. Angew. Chem., Int. Ed. Engl. 1993, 32, 394; (i)
Kaiser, B.; Hoppe, D. Angew. Chem., Int. Ed. Engl. 1995,
34, 323] because, as Hoppe has pointed out (Hoppe, D.;
Kra¨mer, T. Angew. Chem., Int. Ed. Engl. 1986, 25, 160)
factors promoting deprotonation are also conducive to
configurational instability.
10. Invertive electrophilic substitution at benzyllithiums is
precedented, but not invertive deprotonation (see Refs. 2
and 3): Ockham’s razor (Hoffmann, R. Bull. Soc. Chim.
Fr. 1996, 133, 117) leads us to discount a double inver-
sion mechanism.
11. Bragg, R. A.; Clayden, J. Tetrahedron Lett. 1999, 40,
8323.
12. Bragg, R. A.; Clayden, J. Tetrahedron Lett. 1999, 40,
8327.
13. (a) Marshall, J. A.; Lebreton, J. J. Am. Chem. Soc. 1988,
110, 2925; (b) Overberger, C. G.; Marullo, N. P.; Hiskey,
R. G. J. Am. Chem. Soc. 1959, 81, 1517.
5. Lithiated amides [(a) Ahmed, A.; Clayden, J.; Rowley,
M. J. Chem. Soc., Chem. Commun. 1998, 297; (b) Ahmed,
A.; Clayden, J.; Yasin, S. A. Chem. Commun. 1999, 231;
(c) Clayden, J.; Menet, C. J.; Mansfield, D. J. Org. Lett.
2000, 2, 4229; (d) Clayden, J.; Tchabanenko, K.; Yasin,
S. A.; Turnbull, M. D. Synlett 2001, 302; (e) Clayden, J.;
Purewal, S.; Helliwell, M.; Mantell, S. J. Angew. Chem.
Int. Ed. 2002, 41, 1091], sulfones [(f) Crandall, J. K.;
Ayers, T. A. J. Org. Chem. 1992, 57, 2993], sulfonamides
[(g) Aggarwal, V. K.; Ferrara, M. Org. Lett. 2000, 2,
4107; (h) Aggarwal, V. K.; Alonso, E.; Ferrara, M.; Spey,
S. E. J. Org. Chem. 2002, 67, 2335; (i) Breternitz, H.-J.;
Schaumann, E.; Adiwidjaja, G. Tetrahedron Lett. 1991,
32, 1299] and phosophonamides [(j) Ferna´ndez, I.; Ortiz,
F. L.; Tejerina, B.; Granda, S. G. Org. Lett. 2001, 3,
1339; (k) Go´mez, G. R.; Ortiz, F. L. Synlett 2002, 781]
undergo anionic cyclizations with loss of aromaticity. For
a review, see: Ref. 3a.
14. Bragg, R. A.; Clayden, J.; Menet, C. J. Tetrahedron Lett.
2002, 43, 1955.
15. The regiochemistry of the major product was determined
by isotopically labelling (R,S)-14c by alkylation of 16c
with CD3I.
16. See Ref. 14 and Bull, S. D.; Davies, S. G.; Kelly, P. M.;
Gianotti, M.; Smith, A. D. J. Chem. Soc., Perkin Trans.
1 2001, 3106.
17. Parsons, A. F. Tetrahedron 1996, 52, 4149.
18. (a) Clayden, J.; Tchabanenko, K. Chem. Commun. 2000,
317; (b) Clayden, J.; Menet, C. J.; Tchabanenko, K.
Tetrahedron 2002, 58, 4727.
19. (a) Ahmed, A.; Bragg, R. A.; Clayden, J.; Tchabanenko,
K. Tetrahedron Lett. 2001, 42, 3407; (b) Bragg, R. A.;
Clayden, J.; Bladon, M.; Ichihara, O. Tetrahedron Lett.
2001, 42, 3411.
20. (a) Carlsen, P. H.; Katsuki, T.; Mart´ın, V. S.; Sharpless,
K. B. J. Org. Chem. 1981, 46, 3936; (b) Shioiri, T.;
Matsuura, F.; Hamada, Y. Pure Appl. Chem. 1994, 66,
2151. In our synthesis of (−)-kainic acid (Ref. 18b) we
used a meta-methoxy substituent to facilitate oxidation.
21. The regioselectivity appears to be a conformational effect
involving the nearby methyl group. See: Ref. 18.
22. Collado, I.; Ezquerra, J.; Mateo, A. I.; Rubio, A. J. Org.
Chem. 1998, 63, 1995.
6. Hammerschmidt, F.; Hanninger, A. Chem. Ber. 1995,
128, 823.
7. Aldol products 4d and 6d were formed as single
diastereoisomers: relative stereochemistry at the hydroxyl
bearing centre was assigned by analogy with a similar
reaction. See: Ref. 5a.
8. HPLC on the Regis Whelk-O1® chiral stationary phase.
9. We previously demonstrated stereospecificity in the