Controlled Regioselective Anilide Formation from Aspartic and Glutamic Acid Anhydrides

Article abstract of DOI:10.1021/jo971375e

The regioselectivity of the reaction of aniline with a series of N-protected aspartic and glutamic acid anhydrides was controlled through the choice of reaction solvent. In benzene, the formation of aspartic or glutamic acid α-anilides was favored, with α

Full text of DOI:10.1021/jo971375e

J . Org. Chem. 1997, 62, 8821-8825
8821
Con tr olled Regioselective An ilid e F or m a tion fr om Asp a r tic a n d
Glu ta m ic Acid An h yd r id es†
Xicai Huang, Xuehong Luo, Yoann Roupioz, and J effrey W. Keillor*
De´partement de chimie, Universite´ de Montre´al, C.P. 6128, Succursale centre-ville,
Montre´al, Que´bec H3C 3J 7, Canada
Received J uly 25, 1997
The regioselectivity of the reaction of aniline with a series of N-protected aspartic and glutamic
acid anhydrides was controlled through the choice of reaction solvent. In benzene, the formation
of aspartic or glutamic acid R-anilides was favored, with R:(â or γ) regioselectivities as high as
100:0. On the other hand, in DMSO, the formation of either aspartic acid â-anilide or glutamic
acid γ-anilide was favored, with R:(â or γ) regioselectivities as high as 0:100. This regioselectivity
was not observed for alkylamines, amino acids, thiols, or alcohols. However, the high yields of the
aniline reactions and the ability to specify their regioselectivities demonstrate the usefulness of
our method for the formation of aspartic and glutamic acid anilide bonds.
In tr od u ction
acid. During our initial attempts to make the p-nitro-
anilide, we found that activation of the γ-carboxylic group
of Cbz-glutamic acid R-esters (through formation of the
acid chloride or reaction with DCC) and subsequent
attack of p-nitroaniline, a poor nucleophile, resulted in
the formation of Cbz-pyroglutamic acid esters rather than
the desired anilides. (The activation of Fmoc-glutamic
acid as the chloride has also been shown9 to give the
pyroglutamate.) Since the reaction of glutamic acid
anhydride (protected with an N-phthalyl group) with
p-nitroaniline has been shown10 to give the desired
γ-isomer with good regioselectivity, we elected to try to
make our target compounds in a similar manner, through
the direct reaction of Cbz-glutamic acid anhydride with
a series of anilines. We discovered that the regioselec-
tivity of this reaction is remarkably dependent upon the
reaction solvent and allows preparation of either the R-
or the γ-isomer in nearly quantitative yield with good to
excellent regioselectivity. Our investigation of this selec-
tive reaction as well as its extension to aspartic acid
anhydride and to other N-protecting groups (Fmoc and
Boc) are reported herein.
N-Protected glutamic and aspartic acid anhydrides are
very useful intermediates in peptide synthesis. Forma-
tion of an intramolecular anhydride from the parent
N-protected amino acid is easily achieved and activates
carboxylate carbonyls toward nucleophilic attack. Fol-
lowing nucleophilic addition, one carboxylate is released
from the parent anhydride in a form that is suitable for
a subsequent coupling reaction. In this way, the use of
anhydrides serves as a more direct alternative to the
protection, activation, and deprotection of carboxylates
required to make aspartate and glutamate derivatives.
Nucleophilic attack that is not completely regiospecific
for one carbonyl results however in a mixture of R and
â/γ product isomers that must be separated and purified,
reducing the yield and ease of synthesis of the desired
derivative. For example, the reactions of Cbz-glutamic
acid anhydride with glycine1 and with lysine2 have been
reported to produce mixture of R and γ coupled dipeptide
products, whereas Cbz-aspartic acid anhydride has been
reported3 to react with phenylalanine with good regiose-
lectivity. Both glutamic and aspartic acid anhydrides,
N-protected with Fmoc, Cbz, or Boc groups, have also
been reported to react with methanol with moderate
regioselectivity.4,5 The reactions of Cbz-6,7 and Boc-
8aspartic acid anhydrides in benzyl alcohol have been
reported to give the R-benzyl ester with good regioselec-
tivity.
Resu lts a n d Discu ssion
For each reaction between an N-protected aspartic or
glutamic acid anhydride with an aniline (Scheme 1),
reaction progress was followed by TLC, and upon disap-
pearance of starting material (usually within 20 min) the
1
Recently we were interested in the preparation of a
reaction mixture was characterized by H NMR in order
series of novel p-substituted γ-anilides of Cbz-glutamic
to determine the product regioisomer ratios. Identifica-
tion of the R-, â-, and γ-isomers was based on the (∼0.5
ppm) difference between the chemical shifts of their
anilido N-H protons. The products were then isolated
and purified through flash column chromatography and
characterized by 1H and 13C NMR and through elemental
analysis. Finally, authentic R-, â-, and γ-anilides were
prepared by independent, unambiguous synthesis11 and
† Abbreviations used: Boc ) tert-butoxycarbonyl, Cbz ) benzoxy-
carbonyl, Fmoc ) 9-fluorenylmethoxycarbonyl, Asp ) L-aspartic acid,
Glu ) L-glutamic acid (the aspartic and glutamic acids used throughout
this study were always the L stereoisomers).
Abstract published in Advance ACS Abstracts, November 15, 1997.
(1) Le Quesne, W. J .; Young, G. T. J . Chem Soc. 1950, 1954-1959.
(2) Manesis, N. J .; Goodman, M. J . Org. Chem. 1987, 52, 5331-
5341.
1
used to confirm the validity of our H NMR assignments.
As can been seen from Tables 1 and 2, the regioselec-
(3) Yang, C. P.; Su, C. S. J . Org. Chem. 1986, 51, 5186-5191.
(4) Chen, F. M. F.; Benoiton, N. L. Int. J . Pept. Protein Res. 1992,
40, 13-18.
(5) J ouin, P.; Castro, B.; Zeggaf, C.; Pantaloni, A.; Senet, J . P.;
Lecolier, S.; Sennyey, G. Tetrahedron Lett. 1987, 28, 1665-1668.
(6) Walczyna, R.; Sokolowski, J . Carbohydr. Res. 1988, 180, 147-
151.
(7) Furuta, T.; Katayama, M.; Shibasaki, H.; Kasuya, Y. J . Chem.
Soc., Perkin Trans. 1 1992, 1643-1648.
(8) Pietta, P. G.; Cavallo, P.; Marshall, G. R. J . Org. Chem. 1971,
36, 3966-3970.
(9) Benoiton, N. L.; Chen, F. M. F. Int. J . Pept. Protein Res. 1994,
43, 321-324.
(10) Lindsay, H.; Whitaker, J . F. Org. Prep. Proc. Int. 1974, 7, 89-
91 and unpublished observations from this laboratory.
(11) Independent synthetic route (not shown here) involved selective
protection and deprotection of each carboxylate according to standard
literature techniques.
S0022-3263(97)01375-3 CCC: $14.00 © 1997 American Chemical Society

8822 J . Org. Chem., Vol. 62, No. 25, 1997
Huang et al.
Sch em e 1
in the case of the Boc-glutamic acid anhydride, a small
amount of γ-anilide was also formed, although the
R-anilide remains the major product.13 When the same
reactions were performed with N-protected aspartic
rather than glutamic acid anhydrides, the regioselectivi-
ties were good, but not excellent, as might be expected
of a more reactive anhydride.14
In sharp contrast to the results in benzene, reactions
in a polar aprotic solvent such as DMSO gave predomi-
nantly the γ- or â-isomers (3b-13b) in excellent yields.
For example, the reaction of Cbz- or Fmoc-glutamic acid
anhydride with anilines produced almost exclusively the
γ-anilide isomer (Table 2). The more reactive aspartic
acid anhydride analogues gave somewhat lower regiose-
lectivity, with the â-isomer as the major product. The
regioselectivities of the reactions of the aspartic and
glutamic acid anhydrides protected with a Boc group
were found once again to be lower than those of their
Cbz- and Fmoc-protected analogues; the regioselectivity
of the reaction of Boc-glutamic acid anhydride was
moderate, and that of Boc-aspartic acid anhydride was
poor.
The possibility of exploiting this method through the
use of other nucleophiles was also explored. However,
reaction of N-protected aspartic and glutamic acid an-
hydrides with thiols such as thiophenol and benzyl
mercaptan (chosen for their resemblance to aniline)
showed no regioselectivity. Phenol and cyclohexanol
were unreactive, whereas alcohols such as benzyl alcohol,
allyl alcohol, and methanol15 reacted with poor regiose-
lectivity. In reactions of the respective alkoxides with
N-protected glutamic acid anhydride, the alkoxides ap-
peared to react as bases rather than as nucleophiles,
catalyzing the formation of pyroglutamic acid. In addi-
tion, reactions of primary alkylamines and R-amino acids
proceed without regioselectivity. Earlier work16 has
suggested that there is no relationship between the steric
and electronic nature of primary amines and the R/γ
product distribution from their nucleophilic attack on
Cbz-glutamic acid anhydride. The greater regioselectiv-
ity of anilines compared to primary alkylamines may be
due to their reduced basicity (nucleophilicity) and greater
steric bulk.
Ta ble 1. Yield s a n d Regioselectivities of th e Rea ction s
of N-P r otected Glu ta m ic a n d Asp a r tic An h yd r id es w ith
Va r iou s An ilin es in Ben zen e
amino acid N-protecting p-aniline
major
regio-
anhydride
group
(P)
substituent product yielda selectivity
(n)
(X)
anilide (%)
(R:â/γ)b
Glu (2)
Glu (2)
Glu (2)
Glu (2)
Glu (2)
Glu (2)
Glu (2)
Glu (2)
Asp (1)
Asp (1)
Asp (1)
Asp (1)
Cbz
Cbz
Cbz
Cbz
Cbz
Cbz
Fmoc
Boc
Cbz
Cbz
Fmoc
Boc
H
CH3
Cl
OCH3
NO2
NO2
H
H
H
CH3
H
H
3a
4a
5a
6a
7a
7a
8a
9a
10a
11a
12a
13a
98
95
94
99
95
85
84
95
87
95
96
100
100:0
100:0
100:0
100:0
8:1
100:0c
100:0
11:1
5:1
6:1
3:1
2:1
a
Refers to combined masses of both isolated and purified anilide
isomers. As determined by 1H NMR through integration of
b
anilido N-H peaks. c In the presence of 2% acetic acid.
Ta ble 2. Yield s a n d Regioselectivities of th e Rea ction s
of N-P r otected Glu ta m ic a n d Asp a r tic An h yd r id es w ith
Va r iou s An ilin es in DMSO
amino acid N-protecting p-aniline
major
regio-
anhydride
group
(P)
substituent product yielda selectivity
(n)
(X)
anilide (%)
(R:â/γ)b
Glu (2)
Glu (2)
Glu (2)
Glu (2)
Glu (2)
Glu (2)
Asp (1)
Asp (1)
Asp (1)
Asp (1)
Cbz
Cbz
Cbz
Cbz
Fmoc
Boc
Cbz
Cbz
Fmoc
Boc
H
CH3
Cl
OCH3
H
H
H
CH3
H
3b
4b
5b
6b
8b
93
99
99
99
97
95
93
92
95
1:18
1:12
1:11
1:11
0:100
1:7
1:4
1:4
1:6
1:2
Other workers have also observed similar solvent
effects on the regioselectivity of N-protected aspartic or
glutamic acid anhydride ring opening. Yang reported3
that the reaction of Cbz-aspartic acid anhydride with
phenylalanine in DMSO gave predominantly the â-iso-
mer dipeptide (â:R ) 25:1). However, in toluene the
regioselectivity of the same reaction was reversed, and
the major product was the R-isomer (R:â ) 3:1). Benoiton
9b
10b
11b
12b
13b
H
100
a
Refers to combined masses of both isolated and purified anilide
isomers. As determined by 1H NMR through integration of
b
anilido N-H peaks.
(12) A noteworthy exception is the case of p-nitroaniline, where the
regioselectivity of 8:1 (R:γ) was improved to exclusive R-isomer through
the addition of a catalytic amount of acetic acid, probably due to
activation of the R-carbonyl through selective protonation. It was also
observed that many of the R-anilide products actually crystallized out
of solution as the reaction progressed, further facilitating their
regioselective synthesis through this modified procedure.
(13) The addition of a catalytic amount of acetic acid to the reaction
of aniline with Boc-glutamic acid anhydride in boiling benzene did not
enhance regioselectivity as it did in the case of p-nitroaniline attack
on Cbz-glutamic acid anhydride (see ref 12), probably owing to the
greater nucleophilicity of aniline compared to p-nitroaniline.
(14) Note however, that for the reaction of Cbz-aspartic acid
anhydride with aniline, pure R-anilide isomer can be obtained from
the reaction mixture by selective precipitation through the addition
of Et2O.
tivities thus determined for the reactions examined are
excellent and attest to the usefulness and economy of our
method for making either R- or â-/γ-anilides of aspartic
or glutamic acids.
The reaction of N-protected glutamic (1a -c) or aspartic
(1d -f) acid anhydrides with anilines 2a -e in a nonpolar
solvent such as benzene affords the R-anilide isomers
(3a -13a ) with high regioselectivity in excellent yields,
as shown in Table 1. Notably, the reaction of Cbz-
glutamic acid anhydride with anilines gave the R-isomer
anilide exclusively.12 When the protecting group was
changed from Cbz to Fmoc, the ring opening in benzene
was still found to give exclusively R-anilides. However,
(15) See also refs 4 and 5.
(16) Antonjuk, D. J .; Boadle, D. K.; Cheung, H. T. A.; Tran, T. Q. J .
Chem. Soc., Perkin Trans. 1 1984, 1989-2002.

Regioselective Anilide Formation
J . Org. Chem., Vol. 62, No. 25, 1997 8823
Sch em e 2
glutamic acid anhydrides than aspartic acid anhydrides
because of the greater distance between the R-amino
group and the ω-carbonyl.
Con clu sion s
We have shown that the regioselectivity of the reaction
of anilines with Cbz- and Fmoc-aspartic and glutamic
acid anhydrides can be controlled through choice of
reaction solvent to give regioselectively R- or ω-anilides.
In benzene, R-anilides are obtained, whereas in DMSO,
â/γ-anilides are produced. Our method provides a direct
and high-yielding alternative to multistep processes
involving the selective protection, activation, substitution,
and deprotection of aspartic and glutamic carboxylic
acids. This reaction is therefore very useful for the
formation of aspartic and glutamic anilides and should
find its place among the repertoire of peptide chemists.
Sch em e 3
Exp er im en ta l Section
Ma ter ia ls. L-Aspartic acid, L-glutamic acid, and N,N′-
dicyclohexylcarbodiimide (DCC) were purchased from Sigma-
Aldrich; the acylating reagents Cbz-Cl, Fmoc-Cl, and di-tert-
butyl dicarbonate (“Boc anhydride”) were purchased from
Aldrich Chemical Co. Anilines were either purchased from
Aldrich or prepared in our laboratory via a modified Hofmann
rearrangement.20,21 Benzene and DMSO were dried over CaH2
prior to use. Other commercially available products were used
without further purification. Column chromatography was
carried out by using silica gel (200-430 mesh) obtained from
A & C American Chemicals, Ltd. Analytical silica gel 60 F254
aluminum-backed TLC plates were purchased from EM Sci-
ence and were developed using UV light (254 nm).
has also reported4 a similar solvent effect and modest
regioselectivity for the attack of phenylalanine on Cbz-
aspartic acid anhydride. In DMF, the â-dipeptide was
the predominant product (â:R ) 5:1) whereas in dichlo-
romethane, the R-isomer, was the major product (R:â )
2:1). All of these examples are consistent with our
findings in that the use of a nonpolar reaction solvent
favors formation of the R-isomer, whereas the use of a
polar aprotic reaction solvent gives rise to the â- or
γ-isomer.
One possible reason for this effect on regioselectivity
may be the involvement of an intramolecular hydrogen
bond proposed3 to exist between the hydrogen on the
R-amino nitrogen and the oxygen of the R-carbonyl
(Scheme 2). In nonpolar solvents, such an intramolecular
interaction would be strong and render the R-carbonyl
more electrophilic and more susceptible to nucleophilic
attack. Table 1 shows that in benzene the R-anilide is
indeed the predominant isomer. In polar aprotic solvents
such as DMSO, the putative intramolecular hydrogen
bond would be replaced by an intermolecular hydrogen
bond between the R-amino hydrogen and a solvent
molecule (Scheme 3). In the absence of intramolecular
activation, attack is favored on the less sterically hin-
dered â/γ, as illustrated in Table 2. For example,
N-(phthalyl)glutamic acid anhydride, which lacks any
hydrogen on the R-amino nitrogen and is thus incapable
of forming an intramolecular hydrogen bond, reacts with
anilines to give exclusively the γ-anilide isomers.10
The role of the steric environments of the R and â/γ
carbonyls in the regioselectivity of the reaction is revealed
in the influence of the size of the N-protecting group on
regioselectivity. For example, anhydrides substituted
with large N-protecting groups more capable of sterically
hindering the R-carbonyl, such as Cbz and Fmoc, reacted
with greater regioselectivity than those protected with
the smaller Boc group (Tables 1 and 2). Very low
regioselectivity has also been reported for the reaction
of aniline with N-(acetyl)glutamic acid anhydride17 and
for the reaction of p-cyanoaniline with N-(trifluoroacetyl)
aspartic18,19 and glutamic19 acid anhydrides. This effect
of steric hindrance is more pronounced in the case of
Meth od s. Reported melting points are uncorrected. El-
emental analyses were performed on-site in the Laboratoire
d′Analyse EÄ le´mentaire of the Universite´ de Montre´al.
L-N-Cbz-glu ta m ic Acid r-An ilid e (3a ). Gen er a l P r o-
ced u r e. Aniline (46 µL, 0.5 mmol) and L-N-Cbz-glutamic acid
anhydride2 (1a ) (138 mg, 0.5 mmol) were heated in boiling
benzene for 20 min. The pure product was collected by
filtration as a white solid (175 mg, 98%) and washed with
benzene. A sample was recrystallized for microanalysis from
1
CHCl3: mp 196-197 °C, H NMR (300 MHz, CDCl3) δ 2.00-
2.14 (m, 1H), 2.18-2.22 (m, 1H), 2.42-2.57 (m, 1H), 2.62-
2.74 (m, 1H), 4.58-4.70 (m, 1H), 5.15 (s, 2H), 5.70 (d, 1H, J )
6 Hz), 7.05-7.12 (m, 1H), 7.22-7.32 (m, 2H), 7.38 (s, 5H),
7.50-7.59 (m, 2H), 8.88 (s, 1H); 13C NMR (75 MHz, CD3OD) δ
176.6, 172.8, 158.7, 139.5, 138.2, 129.9, 129.6, 129.2, 129.0,
125.6, 121.7, 68.0, 56.6, 31.4, 28.8. Anal. Calcd for
C19H20N2O5: C, 64.0; H, 5.7; N, 7.9. Found: C, 64.4; H, 5.8;
N, 7.8.
L-N-Cbz-glu ta m ic a cid r-p-m eth yla n ilid e (4a ): yield
95%; mp 178-180 °C; 1H NMR (300 MHz, CDCl3) δ 1.98-2.12
(m, 1H), 2.16-2.28 9m, 1H), 2.30 (s, 3H), 2.42-2.58 (m, 1H),
2.60-2.74 (m, 1H), 4.56-4.69 (m, 1H), 5.12 (s, 2H), 5.73 (d,
1H, J ) 6 Hz), 7.00-7.12 (m, 2H), 7.30-7.48 (m, 7H), 8.72 (s,
1H); 13C NMR (75 MHz, CD3OD) δ 176.5, 172.5, 158.5, 138.1,
136.8, 135.3, 130.3, 129.5, 129.0, 128.9, 121.7, 67.8, 56.5, 31.3,
28.7, 20.9. Anal. Calcd for C20H22N2O5: C, 64.9; H, 6.0; N,
7.6. Found: C, 65.2; H, 6.2; N, 7.5.
L-N-Cbz-glu ta m ic a cid r-p-ch lor oa n ilid e (5a ): yield
94%; mp 183-184 °C; 1H NMR (300 MHz, CDCl3) δ 2.00-2.12
(m, 1H), 2.15-2.28 (m, 1H), 2.40-2.54 (m, 1H), 2.60-2.74 (m,
1H), 4.55-4.68 (m, 1H), 5.15 (s, 2H), 5.68 (d, 1H, J ) 6 Hz),
7.28 (s, 5H), 7.32-7.40 (m, 2H), 7.42-7.53 (m, 2H), 8.99 (s,
1H); 13C NMR (75 MHz, CD3OD) δ 176.4, 172.7, 158.7, 138.3,
138.1, 130.3, 129.8, 129.5, 129.1, 128.9, 122.8, 67.9, 56.5, 31.2,
28.6. Anal. Calcd for C19H19ClN2O5: C, 58.4; H, 4.9; N, 7.2.
Found: C, 58.5; H, 5.0; N, 7.0.
(17) Nicolet, B. H. J . Am. Chem. Soc. 1930, 52, 1192-1195.
(18) Lapidus, M.; Sweeney, M. J . Med. Chem. 1973, 16, 163-166.
(19) Kawai, M.; Nyfeler, R.; Berman, J . M.; Goodman, M. J . Med.
Chem. 1982, 25, 397-402.
(20) Huang, X.; Keillor, J . W. Tetrahedron Lett. 1997, 38, 313-316.
(21) Huang, X.; Seid, M.; Keillor, J . W. J . Org. Chem. 1997, 62,
7495-7496.

8824 J . Org. Chem., Vol. 62, No. 25, 1997
Huang et al.
L-N-Cbz-glu ta m ic a cid r-p-m eth oxya n ilid e (6a ): yield
99%; mp 184-185 °C; 1H NMR (300 MHz, CDCl3) δ 2.00-2.14
(m, 1H), 2.18-2.30 (m, 1H), 2.42-2.55 (m, 1H), 2.59-2.62 (m,
1H), 3.78 (s, 3H), 4.55-4.68 (m, 1H), 5.13 (s, 2H), 5.76 (d, 1H,
J ) 6 Hz), 6.77-6.86 (m, 2H), 7.34 (s, 5H), 7.40-7.48 (m, 2H),
8.77 (s, 1H); 13C NMR (75 MHz, CD3OD) δ 176.5, 172.4, 158.5,
158.2, 138.1, 132.3, 129.5, 129.0, 128.9, 123.4, 145.0, 67.8, 56.4,
55.9, 31.3, 28.7. Anal. Calcd for C20H22N2O6: C, 62.2; H, 5.7;
N, 7.2. Found: C, 62.1; H, 5.8; N, 7.2.
L-N-Cbz-glu ta m ic Acid r-p-Nitr oa n ilid e (7a ). After 2
h of heating in the presence of acetic acid (0.2 mL), a pale
yellow solid (0.171 g, 85%) was obtained: mp 162-164 °C; 1H
NMR (300 MHz, CDCl3) δ 2.00-2.14 (m, 1H), 2.20-2.34 (m,
1H), 2.48-2.60 (m, 1H), 2.64-2.80 (m, 1H), 4.66-4.70 (m, 1H),
5.15 (s, 2H), 5.56 (d, 1H, J ) 6 Hz), 7.38 (s, 5H), 7.66-7.74
(m, 2H), 8.10-8.20 (m, 2H), 9.52 (s, 1H); 13C NMR (75 MHz,
CD3OD) δ 176.4, 173.2, 158.5, 145.7, 144.8, 138.0, 129.5, 129.0,
128.9, 125.7, 120.7, 67.9, 56.6, 31.2, 28.4. Anal. Calcd for
C19H20N3O7: C, 56.7; H, 5.0; N, 10.4. Found: C, 57.0; H, 4.9;
N, 10.3. In the absence of acetic acid, a mixture of R and γ
isomers was obtained (95% yield, R:γ ) 8:1).
L-N-Cbz-glu ta m ic Acid γ-An ilid e (3b). Gen er a l P r o-
ced u r e. L-N-Cbz-glutamic acid anhydride2 (1a ) (66, mg, 0.25
mmol) and aniline (27 µL, 0.30 mmol) were stirred in DMSO
(2.0 mL) for 20 min at rt. Sodium bicarbonate (saturated, 10
mL) and water (10 mL) were then added, and the aqueous
solution was extracted twice with CH2Cl2. The aqueous
solution was then acidified to pH 1 and the product extracted
with EtOAc (3 × 30 mL). The combined EtOAc layers were
washed with water and dried over MgSO4, and the solvent was
removed to give a white solid (83 mg, 93%, γ:R ) 18:1).
Separation by flash column chromatography (eluent EtOAc,
then 5% MeOH in EtOAc) afforded the pure γ-anilide as a
white solid. A sample was recrystallized for microanalysis
from EtOH/water: mp 157-158 °C; 1H NMR (300 MHz,
CDCl3) δ 2.02-2.16 (m, 1H), 2.28-2.40 (m, 1H), 2.48-2.73 (m,
2H), 4.36-4.48 (m, 1H), 5.12 (s, 2H), 5.91 (d, 1H, J ) 6 Hz),
7.08-7.20 (m, 1H), 7.26-7.40 (m, 7H), 7.48-7.58 (m, 2H), 7.98
(s, 1H); 13C NMR (75 MHz, CD3OD) δ 175.5, 173.3, 158.8,
139.9, 138.2, 129.9, 129.6, 129.1, 129.0, 125.3, 121.5, 67.9, 55.1,
34.3, 28.7. Anal. Calcd for C19H20N2O5: C, 64.0; H, 5.7; N,
7.9. Found: C, 64.3; H, 6.1; N, 8.1.
L-N-Cbz-glu ta m ic a cid γ-p-m eth yla n ilid e (4b): yield
99% (γ:R ) 11.6:1); mp 135-137 °C; 1H NMR (300 MHz,
CDCl3) δ 2.02-2.20 (m, 1H), 2.26-2.42 (m, 1H), 2.32 (s, 3H),
2.48-2.72 (m, 2H), 4.37-4.50 (m, 1H), 5.12 (s, 2H), 5.92 (d,
1H, J ) 6 Hz), 7.05-7.20 (m, 2H), 7.26-7.49 (m, 7H), 7.94 (s,
1H). 13C NMR (75 MHz, CD3OD) δ 175.5, 173.2, 158.8, 138.2,
137.3, 135.0, 130.3, 129.6, 129.1, 129.0, 121.5, 67.8, 55.1, 34.3,
28.7, 21.0. Anal. Calcd for C20H22N2O5: C, 64.9; H, 6.0; N,
7.6. Found: C, 64.5; H, 6.2; N, 7.6.
MHz, DMSO-d6) δ 1.80-2.05 (m, 2H), 2.24-2.45 (m, 2H),
4.12-4.35 (m, 4H), 7.00-7.12 (m, 1H), 7.20-7.39 (m, 4H),
7.39-7.46 (m, 2H), 7.55-7.65 (m, 2H), 7.65-7.79 (m, 3H),
7.84-7.92 (m, 2H), 10.03 (s, 1H); 13C NMR (100 MHz, DMSO-
d6) δ 173.8, 170.6, 156.1, 143.9, 140.8, 138.9, 128.7, 127.7,
127.1, 125.4, 123.4, 120.1, 119.4, 65.8, 54.8, 46.7, 30.4, 27.2.
Anal. Calcd for C26H24N2O5: C, 70.3; H, 5.4; N, 6.3. Found:
C, 69.9; H, 5.5; N, 6.1.
L-N-F m oc-glu t a m ic Acid γ-An ilid e (8b ). L-N-Fmoc-
glutamic acid anhydride4 (1b) (88.8 mg, 0.25 mmol) and aniline
(27.3 µL, 0.30 mmol) in DMSO (3 mL) were stirred for 20 min
at rt. Ethyl acetate was then added, and the solution was
washed with water, then dilute HCl, and then dried over
MgSO4. Evaporation of the solvent gave the pure γ product
(103.2 mg, 97%). A sample was recrystallized for microanaly-
sis from EtOH/H2O: mp 218-219 °C; 1H NMR (400 MHz,
DMSO-d6) δ 1.80-1.95 (m, 1H), 2.05-2.20 (m, 1H), 2.39-2.55
(m, 2H), 3.96-4.06 (m, 1H), 4.14-4.35 (m, 3H), 6.98-7.05 (m,
1H), 7.24-7.38 (m, 4H), 7.38-7.46 (m, 2H), 7.50-7.62 (m, 2H),
7.65-7.79 (m, 3H), 7.84-7.92 (m, 2H); 13C NMR (100 MHz,
DMSO-d6) δ 173.7, 170.4, 156.2, 143.9, 140.8, 139.3, 128.7,
127.7, 127.1, 125.3, 123.0, 120.2, 119.1, 65.7, 53.5, 46.7, 32.8,
26.5. Anal. Calcd for C26H24N2O5: C, 70.3; H, 5.4; N, 6.3.
Found: C, 70.2; H, 5.5; N, 6.3.
L-N-Boc-glu ta m ic Acid r-An ilid e (9a ). L-N-Boc-glutamic
acid anhydride22 (1c) was allowed to react with aniline
according to the same procedure used for L-N-Cbz-glutamic
acid R-anilide (3a ). A mixture of products (R:γ ) 11:1) was
obtained (180 mg, 95%). Separation by flash column chroma-
tography (eluent EtOAc/CH2Cl2 1:1, then EtOAc, and then 5%
EtOH in EtOAc) yielded both the pure R-isomer (0.165 g, 86%)
and the pure γ-isomer (15 mg, 8%). A sample of the major
product was recrystallized for microanalysis from EtOAc/
1
hexane: mp 73-75 °C; H NMR (300 MHz, CDCl3) δ 1.44 (s,
9H), 1.98-2.12 (m, 1H), 2.18-2.30 (m, 1H), 2.42-2.58 (m, 1H),
2.58-2.70 (m, 1H), 4.56-4.68 (m, 1H), 5.72 (d, 1H, J ) 6 Hz),
7.00-7.12 (m, 1H), 7.18-7.32 (m, 2H), 7.50-7.62 (m, 2H), 9.26
(s, 1H); 13C NMR (100 MHz, DMSO-d6) δ 173.8, 170.8, 155.7,
138.9, 128.7, 123.4, 119.4, 78.3, 54.5, 30.4, 28.2, 27.2. Anal.
Calcd for C16H22N2O5: C, 59.6; H, 6.9; N, 8.7. Found: C, 59.8;
H, 7.2; N, 8.5.
L-N-Boc-glu ta m ic Acid γ-An ilid e (9b). L-N-Boc-glutamic
acid anhydride22 (1c) was allowed to react with aniline
according to the procedure used for L-N-Cbz-glutamic acid
γ-anilide (3b). Yield: 95% (γ:R ) 7.2:1). The products were
separated through flash column chromatography (eluent EtOAc/
CH2Cl2 1:1, then EtOAc, and then 5% EtOH in EtOAc).
A
sample of the major product was recrystallized for microanaly-
sis from EtOAc/hexane: mp 182-184 °C; 1H NMR (300 MHz,
CDCl3) δ 1.44 (s, 9H), 2.00-2.12 (m, 1H), 2.22-2.40 (m, 1H),
2.42-2.70 (m, 2H), 4.30-4.45(m, 1H), 5.60 (d, 1H, J ) 6 Hz),
7.08-7.20 (m, 1H), 7.25-7.32 (m, 2H), 7.50-7.64 (m, 2H), 8.49
(s, 1H); 13C NMR (100 MHz, DMSO-d6) δ 173.2, 168.2, 155.3,
139.1, 128.7, 123.2, 119.1, 78.2, 50.2, 38.2, 28.2; 13C NMR (100
MHz, CD3OD) δ 177.3, 175.0, 159.8, 141.5, 131.4, 126.8, 123.0,
82.3, 56.2, 36.0, 30.4. Anal. Calcd for C16H22N2O5: C, 59.6;
H, 6.9; N, 8.7. Found: C, 59.3; H, 7.1; N, 8.5.
L-N-Cbz-Asp a r tic Acid r-An ilid e (10a ). Gen er a l P r o-
ced u r e. (a ) L-N-Cbz aspartic acid anhydride23 (1d ) (124.5 mg,
0.5 mmol) and aniline (55 mL, 0.6 mmol) in benzene (5.0 mL)
and AcOH (0.1 mL) were heated to reflux for 10 min. The
mixture was then cooled to rt and Et2O (40 mL) was added.
The precipitate (148 mg, 58%, a only) was collected by
filtration. A sample of the product was recrystallized for
microanalysis from CHCl3: mp 177-179 °C; 1H NMR (300
MHz, CDCl3) δ 2.84 (dd, 2H, J 1 ) 17.5, J 2 ) 6.4 Hz), 3.05-
3.20 (m, 2H), 4.68-4.70 (m, 1H), 5.19 (s, 2H), 6.09 (d, 1H, J )
6 Hz), 7.10-7.20 (m, 1H), 7.28-7.42 (m, 7H), 7.42-7.52 (m,
2H), 8.43 (s, 1H); 13C NMR (75 MHz, CD3OD) δ 173.9, 171.7,
158.4, 139.4, 138.0, 129.8, 129.5, 129.0, 128.9, 125.5, 121.7,
67.9, 53.7, 37.2. Anal. Calcd for C18H18N2O5: C, 63.2; H, 5.3;
N, 8.2. Found: C, 63.2; H, 5.6; N, 8.2.
L-N-Cbz-glu ta m ic a cid γ-p-ch lor oa n ilid e (5b): yield 99%
(γ:R ) 11.2:1); mp 162-164 °C; 1H NMR (300 MHz, CDCl3) δ
2.02-2.16 (m, 1H), 2.28-2.40 (m, 1H), 2.48-2.74 (m, 2H),
4.37-4.48 (m, 1H), 5.24 (s, 1H), 5.85 (d, 1H, J ) 6 Hz), 7.25-
7.45 (m, 7H), 7.45-7.58 (m, 2H), 8.12 (s, 1H); 13C NMR (75
MHz, CD3OD) δ 175.4, 173.2, 158.7, 138.7, 138.0, 129.9, 129.8,
129.5, 129.0, 128.9, 122.6, 67.8, 55.0, 34.2, 28.5. Anal. Calcd
for C19H19ClN2O5: C, 58.4; H, 4.9; N, 7.2. Found: C, 58.4; H,
5.0; N, 7.1.
L-N-Cbz-glu ta m ic a cid γ-p-m eth oxya n ilid e (6b): yield
98% (γ:R ) 11.1:1); mp 118-120 °C; 1H NMR (300 MHz,
CDCl3) δ 2.02-2.20 (m, 1H), 2.20-2.42 (m, 1H), 2.50-2.72 (m,
2H), 3.80 (s, 3H), 4.38-4.48 (m, 1H), 5.12 (s, 2H), 5.92 (d, 1H,
J ) 6 Hz), 6.48-6.92 (m, 2H), 7.32 (s, 5H), 7.34-7.45 (m, 2H),
7.84 (s, 1H); 13C NMR (75 MHz, CD3OD) δ 175.5, 173.1, 158.8,
158.1, 138.3, 132.9, 129.6, 129.1, 129.0, 123.3, 115.1, 67.8, 56.0,
55.2, 34.2, 28.8. Anal. Calcd for C20H22N2O6: C, 62.2; H, 5.7;
N, 7.2. Found: C, 62.1; H, 5.7; N, 7.2.
L-N-F m oc-glu t a m ic Acid r-An ilid e (8a ). L-N-Fmoc-
glutamic acid anhydride4 (1b) (176 mg, 0.5 mmol) and aniline
(55 µL, 0.6 mmol) were heated in boiling benzene (20 mL) for
20 min. Hexane (40 mL) was then added, and the precipitate
was collected by filtration to give the pure product as a white
solid (186 mg, 84%). A sample was recrystallized for mi-
croanalysis from EtOH/H2O: mp 213-215 °C; 1H NMR (400
(22) Schro¨der, E.; Klieger, E. Ann. 1964, 673, 197-207.
(23) Munegumi, T.; Meng, Y. Q.; Harada, K. Bull. Chem. Soc. J pn
1989, 62, 2748-2750.

Regioselective Anilide Formation
J . Org. Chem., Vol. 62, No. 25, 1997 8825
(b) If the precipitate in benzene was collected without
addition of Et2O and washed with benzene (10 mL), a mixture
(171 mg, 87%, R:â ) 3:1) was obtained.
according to the procedure used for L-N-Fmoc-glutamic acid
γ-anilide (8b). Yield: 95% (â:R ) 5.5:1). Purification of the
major product through flash column chromatography (eluent
EtOAc/CH2Cl2 1:1 with 0.5% AcOH) gave pure â-isomer as a
white solid. A sample of the product was recrystallized for
microanalysis from EtOH/H2O: mp 174-176 °C; 1H NMR (400
MHz, DMSO-d6) δ 2.70 (dd, 1H, J 1 ) 16, J 2 ) 8 Hz), 2.86 (dd,
1H, J 1 ) 16 Hz, J 2 ) 5 Hz), 4.20-4.35 (m, 3H), 4.40-4.53 (m,
1H), 7.00-7.10 (m, 1H), 7.20-7.36 (m, 4H), 7.36-7.46 (m, 2H),
7.52-7.65 (m, 2H), 7.65-7.76 (m, 3H), 7.85-7.92 (m, 2H),
10.00 (s, 1H); 13C NMR (100 MHz, DMSO-d6) δ 173.1, 168.1,
155.9, 143.8, 140.7, 139.2, 128.7, 127.7, 127.1, 125.3, 123.2,
120.2, 119.1, 65.7, 50.5, 46.6, 38.2. Anal. Calcd for
L-N-Cbz-a sp a r tic a cid r-p-m eth yla n ilid e (11a ): (a ) yield
67% (R only) by Et2O precipitation; mp 175-177 °C; 1H NMR
(300 MHz, CDCl3) δ 2.32(s, 3H), 2.82 (dd, 2H, J 1 ) 17.3, J 2
)
6.4 Hz), 3.04-3.18 (m, 2H), 4.65-4.78 (m, 1H), 5.16 (s, 2H),
6.08 (d, 1H, J ) 6 Hz), 7.05-7.18 (m, 2H), 7.30-7.45 (m, 7H),
8.48 (s, 1H); 13C NMR (75 MHz, CD3OD) δ 173.9, 171.5, 158.3,
138.0, 136.7, 135.3, 130.2, 129.5, 129.0, 128.9, 121.8, 67.9, 53.7,
37.3, 20.9. Anal. Calcd for C19H20N2O5: C, 64.0; H, 5.7; N,
7.9. Found: C, 64.1; H, 5.7; N, 7.8.
(b) Without addition of Et2O, a mixture was obtained (95%,
R:â ) 3:1).
C
25H22N2O5: C, 69.8; H, 5.2; N, 6.5. Found: C, 70.0; H, 5.5;
N, 6.6.
L-N-Boc-a sp a r tic Acid r-An ilid e (13a ). L-N-Boc-aspartic
L-N-Cbz-a sp a r tic Acid â-An ilid e (10b). Gen er a l P r o-
ced u r e. L-N-Cbz-aspartic acid anhydride23 (1d ) was allowed
to react with aniline according to the procedure used for L-N-
Cbz-glutamic acid γ-anilide (3b). A mixture of products was
obtained (93%, â:R ) 4:1), and the major product was purified
by flash column chromatography (eluent EtOAc, then 5%
MeOH in EtOAc). A sample of the product was recrystallized
for microanalysis from CHCl3: mp 170-172 °C; 1H NMR (300
MHz, CDCl3) δ 2.98 (dd, 2H, J 1 ) 16.1, J 2 ) 7.7 Hz), 3.08-
3.22 (m, 2H), 4.55-4.64 (m, 1H), 5.14 (s, 2H), 6.15 (d, 1H, J )
acid anhydride24 (1f) was allowed to react with aniline accord-
ing to the procedure used for L-N-Cbz-glutamic acid R-anilide
(3a ). Yield: 100% (R:â ) 2:1). Isolation of the pure R-isomer
by flash column chromatography (eluent EtOAc/CH2Cl2 1:1,
then EtOAc, and then 5% MeOH in EtOAc) gave a white solid.
A sample of the product was recrystallized for microanalysis
from EtOAc/hexane: mp 160-161 °C; 1H NMR (300 MHz,
CDCl3) δ 1.50 (s, 9H), 2.84 (dd, 1H, J 1 ) 17 Hz, J 2 ) 6 Hz),
3.00-3.16 (m, 1H), 4.60-4.78 (m, 1H), 5.75-5.90 (m, 1H),
7.07-7.16 (m, 1H), 7.20-7.38 (m, 2H), 7.46-7.55 (m, 2H), 8.62
(s, 1H); 13C NMR (100 MHz, CD3OD) δ 177.7, 168.8, 159.8,
141.1, 131.5, 127.2, 123.3, 82.7, 55.0, 39.0, 30.3. Anal. Calcd
for C15H20N2O5: C, 58.4; H, 6.5; N, 9.1. Found: C, 58.4; H,
6.7; N, 9.0.
6 Hz), 7.32-7.42 (m, 7H), 7.45-7.52 (m, 2H), 7.66 (s, 1H); 13
C
NMR (75 MHz, CD3OD) δ 174.7, 170.6, 158.4, 139.7, 138.1,
129.8, 129.4, 129.0, 129.7, 125.3, 121.3, 67.7, 52.2, 39.6. Anal.
Calcd for C18H18N2O5: C, 63.2; H, 5.3; N, 8.2. Found: C, 63.2;
H, 5.6; N, 8.2.
L-N-Cbz-a sp a r tic a cid â-p-m eth yla n ilid e (11b): yield
92% (â:R ) 4.3:1). mp 177-179 °C; 1H NMR (300 MHz, CDCl3)
δ 2.32 (s, 3H), 2.93 (dd, 2H, J 1 ) 15.6, J 2 ) 7.7 Hz), 3.08-3.22
(m, 2H), 4.54-4.64 (m, 1H), 5.14 (s, 2H), 6.17 (d, 1H, J ) 6
Hz), 7.10-7.20 (m, 2H), 7.30-7.40 (m, 7H), 7.72 (s, 1H); 13C
NMR (75 MHz, CD3OD) δ 174.8, 170.5, 158.4, 138.1, 137.0,
135.1, 130.3, 129.5, 129.0, 128.7, 121.5, 67.7, 52.3, 39.5, 21.0.
Anal. Calcd for C19H20N2O5: C, 64.0; H, 5.7; N, 7.9. Found:
C, 64.2, H, 6.0; N, 7.7.
L-N-F m oc-a sp a r tic Acid r-An ilid e (12a ). L-N-Fmoc-
aspartic acid anhydride4 (1e) was allowed to react with aniline
according to the procedure used for L-N-Fmoc-glutamic acid
R-anilide (8a ). Yield: 96% (R:â ) 3:1). Purification of the
major product through flash column chromatography (eluent
EtOAc/CH2Cl2 1:1 with 0.5% AcOH) gave pure R-isomer as a
white solid. A sample of the product was recrystallized for
microanalysis from EtOH/H2O: mp 191-193 °C; 1H NMR (400
MHz, DMSO-d6) δ 2.60 (dd, 1H, J 1 ) 16, J 2 ) 8 Hz), 2.74 (dd,
1H, J 1 ) 16, J 2 ) 5 Hz), 4.18-4.38 (m, 3H), 4.48-4.55 (m,
1H), 7.00-7.10 (m, 1H), 7.20-7.38 (m, 4H), 7.38-7.48 (m, 2H),
7.55-7.64 (m, 2H), 7.80-7.88 (m, 2H), 7.88-7.83 (m, 1H),
7.83-7.94 (m, 2H), 10.09 (s, 1H); 13C NMR (100 MHz, DMSO-
d6) δ 171.6, 169.7, 155.9, 143.8, 140.8, 139.0, 128.7, 127.7,
127.1, 125.4, 123.4, 120.2, 119.4, 65.8, 52.2, 46.7, 36.3. Anal.
Calcd for C25H22N2O5: C, 69.8; H, 5.2; N, 6.5. Found: C, 69.6;
H, 5.2; N, 6.4.
L-N-Boc-a sp a r tic Acid â-An ilid e (13b). L-N-Boc-aspartic
acid anhydride24 (1f) was allowed to react with aniline accord-
ing to the procedure used for L-N-Cbz-glutamic acid γ-anilide
(3b). Yield: 100% (â:R ) 2.2:1). Isolation of the pure â-isomer
by flash column chromatography (eluent EtOAc/CH2Cl2 1:1,
then EtOAc, and then 5% MeOH in EtOAc) gave a white solid.
A sample of the product was recrystallized for microanalysis
from EtOAc/hexane: mp 153-155 °C; 1H NMR (300 MHz,
CDCl3) δ 1.48 (s, 9H), 2.93 (dd, 1H, J 1 ) 16, J 2 ) 8 Hz), 3.08-
3.20 (m, 1H), 4.47-4.58 (m, 1H), 5.90-5.99 (m, 1H), 7.15-
7.22 (m, 1H), 7.30-7.40 (m, 2H), 7.44-7.54 (m, 1H), 7.92 (s,
1H); 13C NMR (100 MHz, CD3OD) δ 176.6, 172.4, 159.5, 141.4,
131.5, 126.9, 122.9, 82.4, 53.5, 41.4, 30.4. Anal. Calcd for
C
15H20N2O5: C, 58.4; H, 6.5; N, 9.1. Found: C, 58.8; H, 6.8;
N, 9.1.
Ack n ow led gm en t. The authors acknowledge Denis
Bilodeau for the elemental analyses reported herein.
The authors are also grateful for the financial support
of the Universite´ de Montre´al, the Natural Sciences and
Engineering Research Council of Canada (NSERC), and
the Fonds pour la formation de chercheurs et l′aide a`
la recherche (FCAR) of the province of Que´bec. One of
the authors (Y.R.) also thanks the Re´gion Rhoˆne-Alpes
of France for an exchange program bursary.
L-N-F m oc-a sp a r tic Acid â-An ilid e (12b). L-N-Fmoc-
aspartic acid anhydride4 (1e) was allowed to react with aniline
(24) Schro¨der, E.; Klieger, E. Ann. 1964, 673, 208-229.
J O971375E