Cyclomaltooligosaccharide-assisted spectroscopic discrimination of phthalimido-derived amino acids through the formation of molecular aggregates

Article abstract of DOI:10.1016/j.carres.2006.09.013

Spectroscopic evidence was used to demonstrate the formation of molecular associates in an aqueous solution of phthalimido tryptophan. These molecular associates are loosely formed through π-π aromatic stacking, properties that are not sufficient to cause NMR spectroscopic enantiomeric discrimination. A cyclomaltooligosaccharide with a larger cavity, such as cyclomaltooctaose (γ-cyclodextrin), is capable of forming a ternary complex with these molecular associates and enhances π-π aromatic stacking interactions, resulting in NMR enantiomeric discrimination. Electrospray-ionization mass spectroscopy (ESIMS) and NOESY two-dimensional NMR spectroscopic methods were used to study these complexes. Association constants and thermodynamic data for these cyclomaltooligosaccharide complexes were also estimated.

Full text of DOI:10.1016/j.carres.2006.09.013

Carbohydrate Research 341 (2006) 2858–2866
Cyclomaltooligosaccharide-assisted spectroscopic discrimination
of phthalimido-derived amino acids through the formation
of molecular aggregates
Branko S. Jursic* and Paresh K. Patel
Department of Chemistry, University of New Orleans, New Orleans, LA 70148, USA
Received 22 July 2005; accepted 19 September 2006
Available online 10 October 2006
Abstract—Spectroscopic evidence was used to demonstrate the formation of molecular associates in an aqueous solution of phtha-
limido tryptophan. These molecular associates are loosely formed through p–p aromatic stacking, properties that are not sufficient
to cause NMR spectroscopic enantiomeric discrimination. A cyclomaltooligosaccharide with a larger cavity, such as cyclomalto-
octaose (c-cyclodextrin), is capable of forming a ternary complex with these molecular associates and enhances p–p aromatic stack-
ing interactions, resulting in NMR enantiomeric discrimination. Electrospray-ionization mass spectroscopy (ESIMS) and NOESY
two-dimensional NMR spectroscopic methods were used to study these complexes. Association constants and thermodynamic data
for these cyclomaltooligosaccharide complexes were also estimated.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Keywords: Cyclomaltooligosaccharide; Enantiomeric discrimination; Molecular aggregation; ESIMS
1. Introduction
meric recognition in comparison with their unmodified
counterparts.2 Even so, non-functionalized CDs immo-
bilized on polymers have been successfully used as a
stationary phase for chiral chromatography.3
Since the first report of chiral recognition by a cyclo-
maltooligosaccharide almost a half a century ago,1 there
has been a substantial amount of research done that
shows that cyclomaltooligosaccharides (commonly
known as cyclodextrins and herein referred to as CDs)
can function as chiral environments that can be used
for enantiomeric recognition.2 However, this research
has proved to be of limited success. It has been postu-
lated that the major reason for the somewhat lackluster
results associated with CDs in enantiomeric recognition
was due to the higher symmetry of CDs and their limited
interactions (mainly hydrophobic) between the guest
molecule and the CD skeleton, due to the fact that the
hydroxyl groups of CD mostly interact with the
surrounding aqueous environment. For these reasons,
functionalized (modified) CDs show superior enantio-
Considering that the cavity of non-functionalized
CDs cannot provide ‘classical’ three-point interactions
that are so often used in the explanation of enantiomeric
recognition4 and enantiomeric discrimination associated
with modified CDs, these essential three-point interac-
tions should come from the guest molecule. One possi-
bility is that enantiomeric guest molecules form
molecular aggregates in an aqueous solution through
several non-bonding interactions. If this occurs then
the small molecular aggregates that form, which are
primarily dimers, should bind into a large CD cavity
and enforce the formation of molecular associates and
cause enantiomeric spectroscopic non-equivalence. Our
preliminary results suggest that these interactions occur
between properly derived amino acids in an aqueous
solution. The presence of certain cyclomaltooligo-
saccharides (notably cyclomaltooctaose (c-CD)) makes
these interactions stronger, as demonstrated by NMR
*
Corresponding author. E-mail addresses: bsjcm@uno.edu; bjursic1@
uno.edu
0008-6215/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2006.09.013

B. S. Jursic, P. K. Patel / Carbohydrate Research 341 (2006) 2858–2866
2859
spectroscopic study.5 Herein we explore the CD-assisted
spectroscopic discrimination of enantiomers of amino
acid derivatives.
non-bonding interactions between enantiomers in bulk
water media.
Electrospray-ionization mass spectroscopy (ESIMS)
has been used extensively for the detection of the forma-
tion of molecular associates in aqueous solutions.5,10
Negative-ion ESIMS spectroscopy (sodium salt) of 1aS
in aqueous solution clearly indicates the formation of
molecular associates, and notably, the formation of
the molecular dimer with MW = 667.1 (Fig. 2). One
can argue that the failure to observe the NMR enantio-
mer non-equivalence is due to a weak enantiomer asso-
ciation constant.
2. Results and discussion
It is well documented that aromatic compounds bind
into the CD cavity.6 On the other hand, amino acids
are an abundant source of chirality in organic chemistry,
yet the majority of them do not form strong inclusion
complexes with CDs. Therefore, introducing an
aromatic ring through derivatizations of amino acids
seemed to be an appropriate method to employ in our
proposed studies. To further reinforce the logic of using
molecular systems derived from aromatic compounds
and amino acids, we found it significant that it has been
well established that aromatic compounds tend to bind
into peptide cavities that have aromatic amino acids
with complementary electronic properties.7 To explore
the importance of aromatic–aromatic interactions in
the formation of chiral molecular aggregates and their
binding into CDs, three amino acids (alanine, phenyl-
alanine, and tryptophan) and two carboxylic acid
anhydrides (phthalic anhydride and cyclohexane-1,2-
dicarboxylic anhydride) were selected in the preparation
of six R-isomeric and six S-isomeric imino derivatives.
The first group of chiral molecules used in our study
of molecular associates consisted of molecular systems
with a contribution from p–p aromatic interactions
and are shown in Scheme 1.
Our reasoning for preparing N-phthalimidotrypto-
phan (1a) was simple: these molecules should form
molecular aggregates in aqueous media, and the aro-
matic moieties of the molecular aggregates should bind
into the CD cavity to form a strong inclusion complex.
There are many structural possibilities to consider in
the formation of molecular complexes between amino
acid derivatives 1 and CDs, and a few are presented in
Figure 1. It is reasonable to assume that all of these
molecular complexes are in dynamic equilibrium in
aqueous media. It is also known that an NMR spectro-
scopic study would give an average chemical shift
(signals) for all molecular aggregates in equilibrium.8
All NMR spectra of 1a with the enantiomeric compo-
sition of R:S = 1:4 (enantiomer non-equivalence) and in
concentration range from 10ꢀ4 to 10ꢀ1 are identical. We
demonstrated in some of our earlier studies of chiral
amides in chloroform solution that NMR non-equiva-
lency of enantiomers can be observed if one of the enan-
tiomers is present in excess.9 This was explained by the
formation of weak diastereomeric complexes between
molecules of the excess enantiomer acting as a resolving
agent.9 According to our NMR study, this is not the
case with compound 1a, presumably due to very weak
If the association constant is very small, then the for-
mation of 1a molecular associate can be enforced by CD
encapsulation (binding) as proposed in Figure 1. The
CD cavity should be of sufficient size to accommodate
two aromatic rings and in this way form a ternary com-
plex with a molecular associate of 1a. Therefore, NMR
enantiomeric non-equivalence should be observed in
c-CD, the cyclomaltooligosaccharide with the largest
1
cavity studied here. Our H NMR spectroscopic studies
fully agree with this reasoning (Fig. 3). Although there
are some chemical shift changes (noticeable between
7.0 and 7.2 ppm), indicating interactions between
a-CD and racemic 1a, it is unlikely that cyclomaltohexa-
ose (a-CD) can form a ternary (three molecules) or
higher degree cyclomaltooligosaccharide complex with
a 1a dimer, as indicated in Figure 1. On the other hand,
enantiomeric discrimination is observed in cyclomalto-
heptaose (b-CD), notably a doubling of some signals
between 7.0 and 7.2 ppm, but not to a high degree.
We believe that this observation is due to the formation
of a ‘normal’ 1:1 complex between racemic 1a and
b-CD. The NMR spectra of racemic 1a in c-CD clearly
shows the NMR enantiomer non-equivalence of
aromatic protons (Fig. 3).
Not only does the presence of c-CD cause enantio-
meric NMR spectroscopic non-equivalence (discrimina-
tion), but it is also possible to assign the signals to the
appropriate enantiomer, as well as determine the enan-
tiomeric composition. For example, there is a little influ-
ence of c-CD on the chemical shift of 1aS (compare the
1aRS spectra from Fig. 3 with 1aS spectra from Fig. 4);
however, there is a substantial upfield chemical shift of
indole hydrogens (6.8–7.5 ppm) and downfield chemical
shift (7.7 ppm) for phthalimido hydrogens (Fig. 3).
Based on this observation, one can suggest a preferential
formation of the c-CD complex with 1aR in comparison
with 1aS.
Considering that the best results of NMR spectro-
scopic non-equivalence (chiral recognition) were ob-
served in an aqueous c-CD of racemic 1a, we turned
This point will be addressed in a separate manuscript that will
extensively study the formation of these molecular associates in
micellar media.

2860
B. S. Jursic, P. K. Patel / Carbohydrate Research 341 (2006) 2858–2866
O
N
O
HO2C
O
ii
N
1b
O
N
i
H
HO2C
O
1a
iii
O
O
H3C
i = Tryptophan/pyridine
ii = Phenylalanine/pyridine
iii = Alanine/pyridine
O
N
Phthalic anhydride
HO2C
1c
O
Scheme 1. Phthalimido amino acid derivatives.
CO2Na
CO2Na
NaO2C
1a
NaO2C
2:2 CD-complex
O
NaO2C
n:n CD-complex
1:1 CD-complex
N
N
H
HO2C
O
NaO2C
NaO2C
HO2C
NaO2C
1:2 CD-complex
NaO2C
2:3 CD-complex
Figure 1. Possible p–p aromatic molecular complexes between CD and 1a.
Figure 2. Negative-ion ESIMS of the sodium salt of 1a in an aqueous solution.
our attention to two-dimensional NMR spectroscopy.
In a pivotal experimental study by Amato et al. on
the application of 1H NMR, two-dimensional, mole-
cular modeling, and molecular dynamic simulation on
CD enantiomeric discrimination,11 it was found that
both R and S enantiomers of the racemic drug CI-
933 were deeply inserted into the host cavity. The dia-
stereomeric complexes thus formed exhibited a different
distribution of intermolecular hydrogen bonds. In our
case, the molecular associates that formed bind in the

B. S. Jursic, P. K. Patel / Carbohydrate Research 341 (2006) 2858–2866
2861
Figure 3. Aromatic portion of the 1H NMR spectra of racemic 1a in aqueous CDs.
Figure 4. Comparison of c-CD induced NMR non-equivalency of 1a.
largest CD cavity to enhance diastereomeric
discrimination.
To better understand the binding of guest compound
1a with host CDs, the association constants (Ka, molꢀ1
)
1
This was perfectly demonstrated on the NOESY spec-
tra of 1aS in water. There are strong space coupling
interactions between the phthalimido (Ha–Hd) moiety
of 1aS and indole (He–Hi) moiety of 1aS. This indicates
that p–p aromatic stacking interactions form molecular
aggregates in aqueous media. When c-CD was added to
the aqueous solution, additional cross coupling signals
were observed between the aromatic moieties of 1aS
and the CD hydrogens. The major cross coupling
interactions were observed between the phthalimido
moiety of 1aS and methylene groups of the c-CD ring
(Fig. 5).
The formation of a ternary complex should also be
observed in negative-ion ESIMS spectra. In basic solu-
tion compound 1a bears at least one negative charge
(carboxylic salt). Considering the c-CD cavity is of
sufficient sizeà to form an inclusive complex with a 1a
dimer, it should not be possible to detect a ‘free’ 1a
dimer (m/z 667) in the solution, but rather it should be
complexed with one c-CD (m/z 981.9), two c-CD (m/z
1630), and three c-CD (m/z 2278.6) molecules, respec-
tively. This is perfectly demonstrated in Figure 6.
were measured by H NMR spectroscopy (500 MHz).
The solution of 1a (10ꢀ3 M) in aqueous NaHCO3
(3 · 10ꢀ3 M) was titrated with the host solution at a dif-
ferent temperature, and each time the change in chemi-
cal shift was measured. Nonlinear regression analysis
using Origin 6.1 (Aston Scientific Ltd) was used to gen-
erate the association constant Ka according to the
equation:12
DmaxKa½Hꢁ
D ¼
1 þ Ka½Hꢁ
where D is the peak shift in parts per million, Dmax the
maximum peak shift in parts per million, and [H] the
concentration of host. At least two experiments were
performed for each system. Using the above equation,
the binding isotherm for guest 1aS to all three CDs
(a, b, and c-CD) were used to calculate the association
constant and standard free energy (DG0 = ꢀRTlnKa;
where R = 8.3144 J Kꢀ1 molꢀ1).
The highest association constant was measured for c-
CD and 1aS (Table 1). These results are in agreement
with the above-mentioned requirements for spectro-
scopic discrimination of enantiomers in aqueous CD
media: mainly that the CD cavity should be sufficiently
large to accommodate the formation of a ternary com-
plex with two guest molecules in the cavity. On the other
à
3
˚
The CD cavity sizes are 174, 262, and 427 A for a-CD, b-CD, and
c-CD, respectively. For further information see Ref. 6.

2862
B. S. Jursic, P. K. Patel / Carbohydrate Research 341 (2006) 2858–2866
Figure 5. The NOESY spectra of 10ꢀ2 M 1aS in D2O (A) and 10ꢀ2 M c-CD (B).
Figure 6. Negative-ion ESIMS of 1aS (10ꢀ3 M) in c-CD (10ꢀ2 M).
Table 1. Association constant and standard free energy DG0 for the
binding of guest molecule 1aS to host a, b and c-CD
and standard enthalpy DH0 of guest 1aS with c-CD at
different temperatures. Due to a higher molecular
motion (molecular dynamics) at a elevated temperature,
the formation of molecular associates at elevated
temperatures decreased, which is reflected in the value
of the association constant.
From the given data, it appears that 1a is a good
molecular system to form molecular associates in
aqueous media and that a CD with large cavity, such
as c-CD will enforce the formation of these molecular
aggregates by forming ternary inclusion complexes with
them. If we assume that molecular associates between 1a
are formed through p–p stacking in aqueous media, then
by replacing the indole ring with a phenyl ring or a
hydrogen atom, these interactions should be diminished
or removed. Indeed, after experimental analysis, we
CD
Association
constant (Ka/molꢀ1
Standard free
energy (DG0/kJ molꢀ1
)
)
a-CD
b-CD
c-CD
66.7 13
105.1 14
226.2 13
ꢀ10.41 0.5
ꢀ11.54 0.3
ꢀ13.44 0.2
hand, it was demonstrated that the R isomer of 1a forms
a slightly stronger ternary complex with c-CD than does
its S isomer. The estimated association constant for 1aR
is approximately 4.64 times higher than that of 1aS with
c-CD.
Listed in Table 2 are the association constant Ka,
standard free energy DG0, standard entropy term TDS0

B. S. Jursic, P. K. Patel / Carbohydrate Research 341 (2006) 2858–2866
Table 2. Thermodynamic parameters for the c-CD binding of 1aS
2863
Temperature (K)
Ka (molꢀ1
)
DG0 (kJ molꢀ1
)
DH0 (kJ molꢀ1
)
TDS0 (kJ molꢀ1
)
298
313
333
226.2 13
ꢀ13.44 0.2
ꢀ12.96 0.1
ꢀ10.67 0.2
ꢀ24.4
ꢀ24.4
ꢀ24.4
2
2
2
ꢀ10.96 1.8
ꢀ11.44 1.9
ꢀ13.73 1.8
186.7
74.1
8
6
Table 3. Association constants and free energies for the CD binding
with 1cS
ternary complex with two molecules of 1c and one CD
is not likely due to the lack of aromatic (electronic)
complementarity observed in 1a.
Our thermodynamic data are in agreement with our
NMR studies. Only small but noticeable enantiomeric
discrimination of racemic 1c was observed in aqueous
b-CD (Fig. 7).
CD
Association
constant Ka (molꢀ1
Standard free
energy DG0 (kJ molꢀ1
)
)
a-CD
b-CD
c-CD
63.5 10
227.1 17
ꢀ10.29 0.4
ꢀ13.45 0.2
ꢀ10.93 0.2
82.1
6
A similar effect should be observed if the phthal-
imido group of 1a is reduced to a cyclohexane ring,
in which case the formation of p–p aromatic stacking
is not possible, and enantiomeric spectroscopic non-
equivalence must come only from a 1:1 CD complex
with the indole moiety in the CD cavity. Again, due
to the size of the CD cavity, only b-CD can induce
spectroscopic enantiomeric non-equivalence. To per-
form this study, the required cyclohexanedicarboimides
2 were prepared by following the reaction outlined in
Scheme 2.
As expected, we were not able to observe any spectro-
scopic non-equivalence with racemic 2b and 2c in the
presence of any of the three CDs. Partial enantiomer sig-
nal separation was observed for the indole protons of
racemic 2a in aqueous b-CD (Fig. 8).
could discern by NMR spectroscopy no enantiomer dis-
crimination in any of the three CD aqueous solutions of
racemic 1b. On the other hand, compound 1c has only
one aromatic group; therefore, the formation of p–p
stacking that is present in 1a (phthalimido–indole aro-
matic interactions) are not possible. Considering the size
of the phthalimide ring, 1c should form the strongest
CD inclusion complex with b-CD.2 This was perfectly
demonstrated with the estimated association constants
between 1c and a-, b-, and c-CD (Table 3). The phthal-
imido ring of 1c is too big to enter the small a-CD
cavity. On the other hand, it fits relatively well into
the b-CD cavity, while the c-CD cavity is too large for
the phthalimido group of 1c. The formation of the
Figure 7. The NMR methyl signal for 1cRS in aqueous CDs.
O
N
H
N
O
HO2C
ii
O
2b
i
HO2C
O
O
iii
O
N
O
O
H3C
Hexahydroisobenzo-
furan-1,3-dione
N
2a
HO2C
2c
O
i = Tryptophan/pyridine; ii = Phenylalanine/pyridine; iii = Alanine/pyridine
Scheme 2. Preparation of cyclohexane-1,2-dicarboxylimido derivatives of amino acids.

2864
B. S. Jursic, P. K. Patel / Carbohydrate Research 341 (2006) 2858–2866
Figure 8. 1H NMR spectrum of the aromatic portion of racemic 2a in aqueous CDs.
3. Conclusions
refluxed for 12 h. The volume of the reaction mixture
was reduced to ꢂ5 mL, and the hot reaction mixture
was added to ice-cooled HCl (150 mL water and
50 mL concd HCl). The solid precipitate that formed
was separated by filtration, washed with water
(3 · 20 mL), and dried in the oven at 60 ꢁC for a few
hours to afford pure 1aS (2.8 g, 84%): mp 180–182 ꢁC.
1H NMR (500 MHz, DMSO-d6): dH 10.74 (1H, s, pyr-
role ring NH) 7.79 (4H, s, benzene ring), 7.48 (1H, d,
J 7.5 Hz), 7.26 (1H, d, J 7.5 Hz) 7.03 (1H, s, pyrrole
ring), 7.00 (1H, t, J 8 Hz) 6.90 (1H, t, J 7.5 Hz) 5.12
(1H, dd, J1 9.5 Hz; J2 6.5 Hz) 3.58 (2H, ddd). 13C
NMR (125 MHz; DMSO-d6): dC 170.5, 167.3 (carbo-
nyls), 136.1, 134.9, 130.9, 127.0, 123.5, 123.4, 121.1,
118.5, 118.0, 111.6, 109.8 (11 aromatic carbons), 52.7
(chiral carbon CH), and 24.2 (methylene carbon).
ESIMSꢀ (CH3OH): m/z 333.1 (MꢀH+,100%), 667.1
(2MꢀH+, 70). Anal. Calcd for C19H14N2O4: C, 68.26;
H, 4.22; N, 8.38. Found: C, 68.31; H, 4.22; N, 8.35.
In conclusion, we can state that, although in general,
aqueous CDs produce a poor chiral environment for
spectroscopic discrimination of racemic mixtures in
comparison with functionalized CDs, this might not be
true if racemic molecules can form molecular associates.
In this case, the binding of these molecular associates
into the CD cavities enhance spectroscopic enantiomeric
discrimination. This was perfectly demonstrated in the
case of 1a and c-CD. If the requirements for the forma-
tion of a molecular associate are eliminated, the spectro-
scopic discrimination is diminished or non-existing. The
evidence of the formation of a ternary CD complex is
provided by electrospray-ionization mass spectroscopy
(ESIMS).
4. Experimental
4.1. General
4.2.2. Preparation of (R)-2-(1,3-dioxo-1,3-dihydroiso-
indol-2-yl)-3-(1H-indol-3-yl)propionic acid (1aR). The
R stereoisomer was prepared in a 79% isolated yield
by following the procedure described for the S stereoiso-
mer. The NMR spectroscopic characteristics for this
compound were identical to those of the S stereoisomer.
Melting points were taken on an Electrothermal IA 9000
Digital Melting Point Apparatus and are uncorrected.
1
The H and 13C NMR spectra were run on a Varian
Unity 400 and a Varian INOVA 500-MHz spectro-
photometer with DMSO-d6 as the solvent and internal
1
standard (2.50 and 35.91 ppm for H and 13C NMR,
4.2.3. Preparation of (S)-2-(1,3-dioxo-1,3-dihydroiso-
indol-2-yl)-3-phenylpropionic acid (1bS). Phthalimide
1bS was prepared by following the procedure for the
preparation of 1aS. The isolated yield of 1bS is 95%
respectively). NOESY spectra were recorded on
500 MHz-spectrophotometer with D2O as the solvent
and the mixing time was kept 0.3 s. The mass spectra
were recorded on a Micromass Quattro 2 Triple Qua-
dropole Mass Spectrometer.
1
(2.8 g): mp 186–188 ꢁC. H NMR (500 MHz, DMSO-
d6): dH 7.82 (4H, s) 7.15 (5H, m), 5.13 (1H, dd, J1
11.5 Hz; J2 4.5 Hz), 3.49 (1H, dd, J1 15.5 Hz; J2
4.5 Hz), 3.35 (1H, dd, J1 14 Hz; J2 11.5 Hz). 13C NMR
(125 MHz, DMSO-d6): dC 170.1, 167.1 (carbonyls),
137.3, 134.9, 130.7, 128.7, 128.3, 126.6, 123.4 (7 aro-
matic carbons), 52.9 (chiral carbon CH), and 33.9 ppm
(methylene carbon). ESIMS (CH3OH): m/z 250.3
(MꢀCO2ꢀH+, 15%), 294.2 (MꢀH+, 100), 589.4
(2MꢀH+, 25).
4.2. General procedure for the preparation of amino acid
imides 1 and 2
4.2.1. Preparation of 2-(1,3-dioxo-1,3-dihydroisoindol-2-
yl)-3-(1H-indol-3-yl)propionic acid (1aS). A pyridine
solution (600 mL) of phthalic anhydride (1.48 g;
0.01 mol) and L-tryptophan (2.04 g; 0.01 mol) was

B. S. Jursic, P. K. Patel / Carbohydrate Research 341 (2006) 2858–2866
2865
4.2.4. Preparation of (R)-2-(1,3-dioxo-1,3-dihydroiso-
indol-2-yl)-3-phenylpropionic acid (1bR). The R stereo-
isomer was prepared in 93% isolated yield by following
the procedure for the preparation of 1aS. The NMR
spectroscopic characteristics for this compound were
identical to those of the 1bS stereoisomer.
was further purified by crystallization from water: mp
145–147 ꢁC. H NMR (500 MHz, DMSO-d6): dH 7.23
1
(2H, t, J 3 Hz), 7.17 (1H, t, J 7 Hz), 7.13 (2H, d, J
6.5 Hz), 4.93 (1H, dd, J1 12 Hz; J2 5 Hz), 3.36 (1H,
dd, J1 14 Hz; J2 5 Hz), 3.26 (1H, dd, J1 13.5 Hz; J2
12 Hz), 2.75 (2H, m), 1.56–0.95 (8H, m). 13C NMR
(125 MHz, DMSO-d6): dC 178.8, 178.6, 169.9 (carbon-
yls), 137.1, 128.9, 128.2, 126.6 (4 aromatic carbons),
52.9 (chiral carbon), 52.2, 38.6, 38.5, 33.3, 23.4, 22.4,
21.2, 21.1 ppm (aliphatic carbon). ESIMSꢀ (CH3OH):
m/z 300.1 (MꢀH+, 100%) 601.2 (2MꢀH+, 20).
4.2.5. Preparation of (S)-2-(1,3-dioxo-1,3-dihydroiso-
indol-2-yl)propionic acid (1cS). Phthalimide 1cS was
prepared by following the procedure for the preparation
of 1aS. Obtained product was further purified by crys-
tallization from water to afford 1cS (2.23 g, 91%): mp
148.5–150 ꢁC. 1H NMR (500 MHz, DMSO-d6): dH
7.85–7.90 (4H, m) 4.87 (1H, q, J 7 Hz), 1.54 (3H, d,
J2 7.5 Hz). 13C NMR (125 MHz, DMSO-d6): dC 171.1,
167.2 (carbonyls), 134.8, 131.3, 123.3 (3 aromatic carb-
ons), 46.9 (chiral carbon CH), and 14.8 ppm (methyl
carbon). ESIMS (CH3OH): m/z 218.2 (MꢀH+, 100%).
4.2.10. Preparation of (R,S)-2-(1,3-dioxo-1H-isoindol-
2(3H,3aH,4H,5H,6H,7H,7aH)-yl)propionic acid (2aRS).
Racemic 2c was prepared from racemic alanine in a
90% isolated yield by the modification of the procedure
for the preparation of 1aS. The pyridine solution
(500 mL) of cyclohexane-1,2-dicarboxylic anhydride
(3.2 g, 0.021 mol) and DL alanine (1.78 g, 0.02 mol)
was refluxed for 10 h. The volume of the reaction mix-
ture was reduced to ꢂ5 mL and the hot reaction mixture
was added to mixture of ethyl acetate (250 mL) and aq
HCl (250 mL water and 50 mL concd HCl). The EtOAc
layer was separated and washed with water (3 · 20 mL)
and then removed under reduced pressure and then
evaporated. The product was crystallized from water
and dried in the air to afford 2cRS (4.05 g, 90%): mp
4.2.6. Preparation of (R)-2-(1,3-dioxo-1,3-dihydroiso-
indol-2-yl)propionic acid (1cR). The R stereoisomer
was prepared in 88% isolated yield by following the pro-
cedure described for the S stereoisomer. The spectro-
scopic characteristics for this compound were identical
to those of the S stereoisomer.
4.2.7. Preparation of (S)-2-(1,3-dioxo-1H-isoindol-
2(3H,3aH,4H,5H,6H,7H,7aH)-yl)-3-(1H-indol-3-yl)pro-
pionic acid (2aS). This imide was prepared by following
the procedure for the preparation of 1aS in 93% (2.1 g
isolated yield): mp 200–201 ꢁC. 1H NMR (500 MHz,
DMSO-d6): dH 10.83 (1H, s, pyrrole ring NH) 7.44
(1H, d, J 8 Hz), 7.30 (1H, d, J 8 Hz) 7.04 (1H, s, pyrrole
ring) 7.03 (1H, t) 6.95 (1H, t, J 7 Hz) 4.91 (2H, dd, J1
11.5 Hz; J2 10.5 Hz) 3.45 (2H, ddd), 2.72 (1H, q, J1
7 Hz), 2.67 (1H, q, J1 7 Hz), 1.50–0.95 (8H, m). 13C
NMR (125 MHz, DMSO-d6): dC 178.9, 178.6, 170.3
(carbonyls), 136.0, 127.1, 123.6, 120.9, 118.4, 118.1,
111.4, 109.5 (8 aromatic carbons), 52.3, 38.8, 38.6,
23.4, 23.3, 22.3, 21.2, 21.0 (8 aliphatic carbon). ESIMS
(CH3OH) m/z 339.1 (MꢀH+, 100%) 679.2 (2MꢀH+,
100). Anal. Calcd for C19H20N2O4: C, 67.05; H, 5.92;
N, 8.23. Found: C, 66.91; H, 5.92; N, 8.23.
1
148–149 ꢁC. H NMR (500 MHz, DMSO-d6): dH 4.62
(1H, q, J 7.5 Hz), 2.94 (2H, m), 1.73–1.61 (4H, m),
1.38 (3H, d, J 7.5 Hz), 1.29–1.26 (4H, m). 13C NMR
(125 MHz, DMSO-d6): dC 178.9, 178.7, 170.9 (carbon-
yls), 46.9 (chiral carbon), 39.0, 23.1, 21.3, 21.3, 21.1,
14.1 (aliphatic carbon). ESIMS (CH3OH): m/z 224.1
(MꢀH+, 100%).
Acknowledgment
We thank the Louisiana Board of Regents for their
financial support (LEQSF (2001-04)-RD-B-12) for this
work.
References
4.2.8. Preparation of (R)-2-(1,3-dioxo-1H-isoindol-
2(3H,3aH,4H,5H,6H,7H,7aH)-yl)-3-(1H-indol-3-yl)pro-
pionic acid (2aR). The 2aR stereoisomer was prepared
in a 92% isolated yield by following the procedure for
the preparation of 1aS. The spectroscopic characteristics
for this compound were identical to those of 2aS.
1. Cramer, F.; Dietsche, W. Chem. Ber. 1959, 92, 378–385.
2. For instance see: (a) Seeman, J. I.; Secor, H. V. Anal.
Chem. 1988, 60, 2120–2127; (b) Jin, H. L.; Stalcup, A.;
Amstrong, D. W. Chirality 1989, 1, 137–141; (c) Kano,
K.; Hasegawa, H. J. Am. Chem. Soc. 2001, 123, 10616–
10627; (d) Rekharsky, M.; Yamamura, H.; Kawai, M.;
Inoue, Y. J. Org. Chem. 2003, 68, 5228–5235; (e) Liu, Y.;
Yang, Y.-W.; Zhang, H.-Y.; Hu, B.-W.; Ding, F.; Li,
C.-J. Chem. Biodiversity 2004, 1, 481–488.
3. For instance see: (a) Thuaud, N.; Sebille, B.; Deratani, A.;
Popping, B.; Pellet, C. Chromatographia 1993, 36, 373–
380; (b) Akiyama, T.; Hishiya, T.; Asanuma, H.; Komi-
yama, M. J. Inl. Phenom. Macrocycl. Chem. 2001, 41, 149–
4.2.9. Preparation of (R,S)-2-(1,3-dioxo-1H-isoindol-
2(3H,3aH,4H,5H,6H,7H,7aH)-yl)-3-(1H-indol-3-yl)pro-
pionic acid (2aRS). Racemic 2b was prepared in a 92%
isolated yield from racemic phenylalanine by following
the procedure for the preparation of 1aS. The product

2866
B. S. Jursic, P. K. Patel / Carbohydrate Research 341 (2006) 2858–2866
153; (c) Lai, X.-H.; Bai, Z.-W.; Ng, S.-C.; Ching, C.-B.
8. Schneider, H.-J.; Hacket, F.; Rudiger, V.; Ikeda, H. Chem.
¨
Chirality 2004, 16, 592–597, and references cited therein.
Rev. 1998, 98, 1755–1786.
4. For instance see: (a) Sundareson, V.; Abrol, R. Chirality
2005, 17, S30–S39; (b) Meyer, V. R.; Rais, M. Chirality
1989, 1, 167–169, and references cited therein.
5. (a) Jursic, B. S.; Patel, P. K. Tetrahedron 2005, 61, 919–
926; (b) Jursic, B. S.; Patel, P. K. Carbohydr. Res. 2005,
304, 1413–1418.
9. (a) Jursic, B. S.; Goldberg, S. I. J. Org. Chem. 1992, 57,
7172–7174; (b) Jursic, B. S.; Goldberg, S. I. J. Org. Chem.
1992, 57, 7370–7372.
10. Kobetic, R.; Jursic, B. S.; Bonnette, S.; Salamone, S. J.
Tetrahedron Lett. 2001, 42, 6077–6082, and references
cited therein.
6. For general information about CDs see: (a) Comprehensive
Supramolecular Chemistry; Szejtli, J., Osa, T., Eds.;
Pergamon: Oxford, UK, 1996; Vol. 3 (CD); (b) Szejtli, J.
Chem. Rev. 1998, 98, 1743–1754.
7. For instance see: Waters, M. L. Biopolymers 2004, 76,
435–445, and references cited therein.
11. Amato, M. E.; Lombardo, G. M.; Pappalardo, G. C.;
Scarlata, G. J. Mol. Struct. 1995, 350, 71–82, and
references cited therein.
12. Connors, K. A. Binding Constants; Wiley Interscience:
New York, 1987.