Specific binding capacity of β-cyclodextrin with cis and trans enalapril: Physicochemical characterization and structural studies by molecular modeling

Article abstract of DOI:10.1016/j.bmc.2008.08.032

The main objective of this work was to study an inclusion complex between enalapril (ENA), and β-cyclodextrin (β-CD). From nuclear magnetic resonance (NMR) we determined that the complex showed a 1:1 stoichiometry, with an apparent formation constant (K_C) of 439 and 290 M^-1 for the cis and trans isomers, respectively. The molecular modeling and NMR techniques demonstrated that the aromatic moiety of ENA was inserted into the hydrophobic cavity of β-CD. When studying the chemical stability of ENA complexed to β-CD, a clear stabilizing effect was observed in both the aqueous solution and solid state.

Full text of DOI:10.1016/j.bmc.2008.08.032

Bioorganic & Medicinal Chemistry 16 (2008) 8403–8412
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Specific binding capacity of b-cyclodextrin with cis and trans enalapril:
Physicochemical characterization and structural studies by molecular modeling
*
Ariana Zoppi, Mario A. Quevedo, Marcela R. Longhi
Departamento de Farmacia, Facultad de Ciencias Químicas, Universidad Nacional de Córdoba, Ciudad Universitaria, 5000 Córdoba, Argentina
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 2 July 2008
Revised 12 August 2008
Accepted 14 August 2008
Available online 19 August 2008
The main objective of this work was to study an inclusion complex between enalapril (ENA), and b-cyclo-
dextrin (b-CD). From nuclear magnetic resonance (NMR) we determined that the complex showed a 1:1
stoichiometry, with an apparent formation constant (KC) of 439 and 290 Mꢀ1 for the cis and trans isomers,
respectively. The molecular modeling and NMR techniques demonstrated that the aromatic moiety of
ENA was inserted into the hydrophobic cavity of b-CD. When studying the chemical stability of ENA com-
plexed to b-CD, a clear stabilizing effect was observed in both the aqueous solution and solid state.
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
Enalapril maleate
b-Cyclodextrin
Drug stability
NMR spectroscopy
Molecular modeling
1. Introduction
ular encapsulation with cyclodextrins (CDs) constitutes an alterna-
tive for the development of new pharmaceutical dosage forms.10–12
Enalapril (ENA, Fig. 1a) is the ethyl ester of the long-acting
angiotensin converting enzyme inhibitor enalaprilat (ET, Fig. 1b),
and is indicated for the treatment of essential and renovascular
hypertension. Since the oral absorption of ENA is superior to that
of ET, the former is used in oral dosage form.1 This drug, however,
suffers degradation in the solid state as a consequence of its inter-
action with excipients, a fact that causes an important stability
problem when considering its formulation. Diketopiperazine
(DKP, Fig. 1c) is its main degradation product, formed in solid state
and by photodegradation in solution, while in solutions of pH 5 or
above, ET is the main degradation product formed by hydrolysis.2–8
As stated above, the stability of ENA in the solid state is a crucial
aspect in its formulation procedures, with the intramolecular
nucleophilic attack by the secondary amine on the proline carbox-
ylic group being the key step for the formation of DKP.3 To over-
come this problem, several methods have been attempted, for
example, the formation of a maleate salt to reduce the attacking
ability of the nitrogen atom in its amine group.3 By means of these
strategies, ENA formulation stability has been improved, but this
achievement has been insufficient to assure long term stability of
the formulations that are currently in use, with several reports
indicating that degradation products are still a concern.2,8,9
Considering that CDs frequently form inclusion complexes with
drugs, and that the flexibility of the included molecule is consider-
ably diminished through steric restriction and intermolecular
interactions, then the preparation of inclusion complexes of ENA
with CDs could restrain the possibility of an intramolecular nucle-
ophilic attack, not only in the solid state but also in solution.
Among CDs, b-cyclodextrin (b-CD, Fig. 2) is the most widely used
in pharmaceutical industry, mainly due to its capability of includ-
ing a wide variety of drugs, but also because of the low cost of bulk
production.
Considering the above mentioned aspects, the purpose of this
study was to prepare and characterize an ENA:b-CD inclusion com-
plex, with the aim of enhancing the stability properties of the for-
mulations currently in use. The study of the three-dimensional
structure of this complex was carried out by means of experimen-
tal and theoretical techniques. Furthermore, the effects obtained
regarding the stability of ENA as a consequence of its inclusion into
the b-CD was studied in both the solid state and solution.
2. Results and discussion
Another widely used strategy to increase the stability of drugs is
the formation of complexes with macromolecules, of which molec-
2.1. ENA:b-CD complex structure elucidation
2.1.1. Spectroscopic studies
NMR spectroscopy is at present the most useful tool for the
study of CD complexes, which provides qualitative and quantita-
* Corresponding author. Tel.: +54 351 433 4163; fax: +54 351 433 4127.
E-mail address: mrlcor@fcq.unc.edu.ar (M.R. Longhi).
0968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.08.032

8404
A. Zoppi et al. / Bioorg. Med. Chem. 16 (2008) 8403–8412
O
a
HO
CH3
CH2 CH2 CH
C
NH
CH
C
O
N
O
O
OCH2CH3
O
b
HO
c
H3C
N
CH3
CH
CH2 CH2 CH
N
CH2 CH2 CH
NH
C
O
N
C
C
O
O
OCH2CH3
O
OH
Figure 1. Molecular structure. (a) Enalapril, (b) enalaprilat and (c) diketopeparazine.
OH
a mixture of the Z-(cis) and E-(trans) isomers (Fig. 3). Therefore,
when possible, the signals corresponding to each isomer were
identified.
O
O
HO
O
As can be seen in Table 1, in the presence of b-CD, almost all
ENA proton resonances were modified, and exhibited both upfield
and downfield displacements with respect to those of the pure
drug. Among the protons that had the most marked upfield dis-
placement were those corresponding to the aromatic ring
(H170190 and H160180200), indicating that this moiety was located
in a rich electronic density environment that produced a shielding
effect. This evidence suggests that the aromatic ring of ENA may
have been included into the b-CD hydrophobic cavity, near the gly-
OH
HO
OH
O
OH
HO
HO
O
OH
O
O
OH
O
HO
OH
OH
cosidic linkage oxygens that are rich in
p electrons. A lower upfield
O
displacement was observed for H140, which indicates that the alkyl
chain that connects the benzyl ring with the rest of the molecule
was still buried into the b-CD cavity, but was closer to the wide
rim of the molecule torus. On the other hand, a downfield displace-
ment was observed for H90, H110 and H120 protons demonstrating a
deshielding effect, which might be explained by the presence of
hydrogen bond interactions between the carbonyl group of the
ethyl ester moiety in ENA and the hydroxyl groups of b-CD. Based
on the additional downfield displacement observed for H20, H70
and H80, we can postulate that the secondary amine and carbonyl
moieties of the amide and carboxylic group also established hydro-
gen bond interactions with the hydrophilic rim of the macromole-
cule. The displacement of the chemical shifts for H30, H40 and H130
was difficult to calculate due to these signals being multiplets,
while those for H70 and H50 protons could not be determined be-
cause of peak overlapping with b-CD protons.
OH
OH
OH
O
O
OH
OH
OH
O
O
OH
HO
O
Figure 2. Molecular structure of b-CD.
tive insight into their structure, as well as their thermodynamic
and kinetic properties.13
2.1.1.1. 1H NMR studies. In order to study the structure of the
ENA:b-CD complex, the spectra of pure ENA and b-CD were com-
pared to their corresponding signals in the inclusion complex.
cis–trans Isomerization may occur in proline-containing molecules
such as ENA, with this drug being present in an aqueous solution as
The analysis of the signals corresponding to b-CD also sheds
lights on the structure of this complex. b-CDs are toroidal mole-
cules (Fig. 4) with a truncated cone shape, with CH groups carrying
O
b
a
HO
1'
8'
8'
16'
15'
17'
CH3
16'
15'
5'
2'
17'
CH3
4'
14'
9'
13'
6'
2'
13'
CH2 CH2 CH
3'
14'
9'
6'
18'
CH2 CH2 CH
NH
CH
7'
C
N
18'
NH
CH
7'
C
N
3'
10'
4'
C
O
10'
19'
20'
C
O
19'
5'
20'
O
OCH2CH3
11' 12'
O
OCH2CH3
11'12'
1'
HO
O
E-(trans)-enalapril
Z-(cis)-enalapril
Figure 3. NMR signal notation. ENA isomers (a) Z-(cis) and (b) E-(trans).

A. Zoppi et al. / Bioorg. Med. Chem. 16 (2008) 8403–8412
8405
Table 1
as the aromatic ring of ENA. As was expected, only small changes in
the chemical shifts were observed for the H1, H2 and H4 protons
located outside the cavity.
Chemical shifts for the protons of ENA in the free state and in the 1:1 complex
Protons of ENA
ENA free (ppm)
ENA complex (ppm)
D
d (ppm)
In order to assess the stoichiometry of the complex, Job’s meth-
od was applied.14 The continuous variation plots for the protons of
ENA and b-CD, whose chemical shift displacements were the most
marked, are shown in Figure 5. It can be observed that the maxi-
mums for both plots were found at r = 0.5, indicating a 1:1 stoichi-
ometry for the ENA:b-CD complex.
Figure 6 shows Scott’s plots for the ENA:b-CD complex, which
were obtained based on the H-9E/Z and H-8E/Z protons of ENA. Kc
values were then estimated from these plots and are presented
in Table 3. The greater Kc value calculated for the complex with
the Z-(cis) ENA isomer, compared to that of the E-(trans) isomer,
may be due to the fact that the former is more hydrophobic than
the latter one.15
H80E
1.2546
1.2716
1.2935
2.7498
3.2051
3.3939
4.1760
4.2566
7.3234
7.4021
1.2691
1.2926
1.3167
2.7388
3.2722
3.4303
4.2014
4.2905
7.2864
7.3764
0.0145
0.021
0.0232
H80Z
H120
H140
ꢀ0.011
0.0671
0.0364
0.0254
0.0339
ꢀ0.037
ꢀ0.0257
H90Z
H90E
H110
H20
H170190
H160180200
OH
H
H2
6
4
H1
H4
HH33
H5
H
O
5
2.1.1.2. ROESY. To study the spatial interaction between host
and guest, and the corresponding three-dimensional geometry of
the complex, a 2D ROESY experiment was carried out (Fig. 7).
ROESY is a two-dimensional technique based on the Nuclear Over-
hauser Effect (NOEs), in which cross-peaks may be observed be-
tween protons if the corresponding internuclear distance is
smaller than 3–4 Å.16 An expansion of the ROESY spectrum ob-
tained is shown in Figure 7b, where an interaction between the
aromatic protons of ENA (H160–H200) and H3, H5 and H6 protons
--O
H
HO
2
1
3
H
OH
H
O--
CH26-OH
7
Figure 4. Molecular structure, NMR signal notation and schematic representation
of the b-CD.
Table 2
Chemical shifts for the protons of b-cyclodextrin in the free state and in the 1:1
complex
Protons of b-CD
b-CD free (ppm)
b-CD complex (ppm)
Dd (ppm)
0.0012
H1
H2
H3
H4
H5
H6
5.1000
3.6793
3.9951
3.6133
3.8837
3.9078
5.0902
3.6696
3.9480
3.6086
3.7641
3.8698
ꢀ0.0098
ꢀ0.0097
ꢀ0.0471
ꢀ0.0047
ꢀ0.1196
ꢀ0.0380
H-8'
H-8'
H-9'
H-9'
E
Z
E
Z
r2= 0.993
r2= 0.988
r2= 0.991
r2= 0.996
0.0010
0.0008
0.0006
0.0004
0.0002
0.0000
the H1, H2 and H4 protons located on the exterior surface of the
torus, which is of a hydrophilic nature. The interior of the torus
is lined with two rings of CH groups (bearing H3 and H5) and also
with glycosidic oxygens (O4), thus offering an environment of
much lower polarity than that of water, and therefore can be con-
sidered to be of hydrophobic nature. Table 2 shows the displace-
ments of the chemical shifts for b-CD protons in the presence or
absence of ENA. A marked upfield displacement for H3, H5 and
H6 is observed, which could be due to the inclusion into the b-
0.005
0.006
0.007
0.008
[β-CD]
0.009
0.010
Figure 6. Scott’s plots for the protons H-90Z (d), H-9E0 (j), H-80Z (.) and H-8E0 (N) of
ENA with increasing concentrations of b-CD.
CD hydrophobic cavity of groups that are rich in
p electrons, such
0.36
0.34
0.32
0.30
0.28
0.26
0.24
0.22
0.20
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0.0
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
r
r
Figure 5. Continuous variation plot for (a) H90E (j), H9Z0 (d), H8Z0 (N) and H8E0 (.) protons of ENA in the presence of different relative concentrations of b-CD, (b) H3 (j), H5
(d) and H6 (N) protons of b-CD in the presence of different relative concentrations of ENA.

8406
A. Zoppi et al. / Bioorg. Med. Chem. 16 (2008) 8403–8412
Table 3
2.1.2. Computational studies
Complex formation constant for the inclusion complex of Z-(cis) and E-(trans) ENA
with b-CD
In order to obtain the initial conformation of ENA, a conforma-
tional search was performed over all the rotatable bonds in the
molecule. In this way, we found that the Z-(cis) conformation
was energetically favored at pH 7.0 respect to the E-(trans) confor-
mation, and thus the former was used to initiate the docking
studies.
On analyzing the carbon and oxygen affinity grids to be used in
the docking procedures (Fig. 8a and b), we were able to confirm
that carbon atoms are preferably accommodated in the interior
cavity of the b-CD torus (hydrophobic cavity), while oxygen atoms
are preferably located near the hydroxyl functional groups in the
outer rim of the molecule.
Proton of ENA
Complex formation constant (Mꢀ1
)
H-9E
H-9Z
H-80E
H-80Z
E-trans
Z-cis
285
455
295
423
290
7
439 23
of b-CD was found, indicating that the aromatic ring of the drug is
deeply internalized into the cyclodextrin cavity.
Figure 7c, shows the intermolecular cross-peaks corresponding
to the alkyl chain connecting the aromatic ring of ENA, in which an
interaction was found between H140 with H3 and H5, suggesting
that this methylene group is also deeply inserted into the b-CD
cavity. On the other hand, H130 only exhibited correlations with
the H3 proton, indicating that it is located closer to the wide rim
of the molecule torus, and is thus not so deeply inserted into the
hydrophobic cavity.
2.1.2.1. Docking studies. From the results obtained by the
docking procedures, we found three clusters of conformations
for the ENA:b-CD complex (Fig. 9a–c). As can be seen, clusters
of conformations were found corresponding to complexes in
which the aromatic ring (a), the proline (b) and the ethyl ester
moiety (c) are included into the b-CD hydrophobic cavity. By
analyzing the docked energy values, it was concluded that the
Figure 7. (a) 2D ROESY spectrum of ENA:b-CD complex. Expansion from the 2D ROESY spectrum of ENA:b-CD complex. (b) In F1 are shown protons H3, H5 and H6 of b-CD,
and in F2 protons H160,180,200 and H170,190 of ENA (c) in F1 are shown protons H3, H5 and H6 of b-CD, and in F2 protons H130 and H140 of ENA.

A. Zoppi et al. / Bioorg. Med. Chem. 16 (2008) 8403–8412
8407
Figure 8. (a) Carbon affinity map located inside the b-CD cavity (hydrophobic) and (b) oxygen affinity map located near the outer rim of the torus (hydrophilic).
Figure 9. Three-dimensional conformation of the complexes between the Z-cis isomer of ENA and b-CD (clusters a, b and c) obtained by molecular docking.
complex in which the aromatic ring of ENA was included into
the hydrophobic cavity of b-CD (a) was slightly more stable than
the other two conformations. Considering the standard error
associated to the quantitation of the docked energy reported
for autodock3 (around 2.00 kcal/mol), one would expect that
these conformations are equally probable. Therefore, in order
to gain insight into the most stable complex, these three confor-
mations obtained by molecular docking were refined, to include
the effect of the presence of explicit solvent molecules, temper-
ature and b-CD flexibility. Thus, the three clusters obtained were
subjected to posterior refinement by means of molecular dynam-
ics methods.
2.1.2.2. Molecular dynamics studies. The quality of the molec-
ular dynamics trajectories corresponding to the clusters (a), (b) and
(c) were evaluated by analyzing the temperature, and kinetic,

8408
A. Zoppi et al. / Bioorg. Med. Chem. 16 (2008) 8403–8412
Figure 10. Plots of temperature, kinetic energy, potential energy and total energy for the molecular dynamics simulations of the ENA:b-CD complex (Cluster (a) (red), (b)
(black) and (c) (yellow)).
potential and total energy profiles for 13 ns of simulation
(Fig. 10a–d). As can be seen, no abnormal events occurred during
the heating, equilibration on production runs, indicating that the
conditions under which the trajectories were obtained were
adequate. For the next step, the rmsd versus time profiles
(Fig. 11a–c) were constructed and analyzed in order to evaluate
if any significant conformational change occurred in the complex
during the molecular dynamics run. From the plots in Figure 11,
it can be seen that cluster (a) shows a slight change in its structure
at around 5 ns, with the benzyl ring now being accommodated
deeply into the b-CD hydrophobic cavity, with this conformation
becoming stable after 6 ns. When cluster (b) was analyzed, we
found that a considerable conformational fluctuation occurred up
to 7 ns of simulation, in which the proline moiety moved toward
the wide rim of the torus, thereby optimizing the establishment
of new hydrogen bond interactions. Cluster (c) exhibited a consid-
erable increase in its rmsd between 3 and 4 ns, which was due to
the exclusion of the ethyl ester moiety of ENA from the b-CD
hydrophobic cavity.
ity during the simulation, and secondly, it was found to be less
favorable on analysis of the total binding energy (ꢀ10.98 kcal/
mol).
Based on the energetic component analysis, we conclude that
the complex conformation in which the aromatic ring of ENA is in-
cluded into the b-CD hydrophobic cavity is the favored conforma-
tion. Although the electrostatic component for cluster (a) was
considerably lower than that of cluster (b) (ꢀ28.92 and
ꢀ95.34 kcal/mol, respectively), the Van der Waals component
was higher (ꢀ25.38 and ꢀ15.13 kcal/mol, respectively). A main
force that drove the complex formation was the solvation compo-
nent, which was positive for the three evaluated conformations.
However, considering that the most hydrophobic part of the ENA
molecule was the aromatic ring, do to its inclusion into the hydro-
phobic cavity and removal from the bulk of the water molecules,
then as expected, the energy required to solvate the ligand during
the complex formation was discovered to be considerably lower in
cluster (a) compared to (b) (36.33 kcal/mol and 95.04 kcal/mol,
respectively). Related to this, the solvation energy dependence
for the formation of cyclodextrin inclusion complexes has been
previously described.13 The formation of complex (a) is in agree-
ment with spectroscopic data, which indicated that the aromatic
ring is included into the b-CD hydrophobic cavity. This was con-
firmed by observing the upfield displacement in H160–H200 pro-
tons of the ENA molecule, as well as using data provided by the
2D ROESY experiments.
2.1.2.3. Energetic component analysis. Following on from the
aspects discussed with respect to the rmsd versus time plots, ener-
getic component analyses (MM-PBSA) were performed over the
time interval corresponding to 7–13 ns of simulation (Table 4).
When the total free energy of binding for the three clusters was
compared, we found that cluster (a) was clearly favored, with an
energy of ꢀ18.63 kcal/mol. This was followed by cluster (b) with
an energy of ꢀ16.00 kcal/mol. The factibility of the formation of
the cluster (c) is discarded based on two observations. First, it
was been demonstrated that ENA was excluded from the b-CD cav-
2.1.2.4. Effect of the cis/trans isomerism in the complex
formation. As can be seen in Table 3, the formation constant
for the complex corresponding to the Z-cis conformation of ENA

A. Zoppi et al. / Bioorg. Med. Chem. 16 (2008) 8403–8412
8409
mer, with the benzyl ring of the ENA molecule being inserted dee-
ply into the b-CD cavity, and the proline moiety being positioned
near the hydrophilic ring of the torus. After 13 ns of molecular
dynamics simulations and also using the MM-PBSA analysis, we
found a free energy of binding of ꢀ15.81 kcal/mol, which com-
pared to the value obtained for the Z-cis conformation
(ꢀ18.63 kcal/mol), thereby supporting the higher formation con-
stant found for the latter. When analyzing the corresponding ener-
getic components, we found electrostatic, Van der Waals and
polar/non-polar contribution of ꢀ37.02, ꢀ27.49 and 48.70 kcal/
mol, respectively. Based on these data, we can conclude that the
force driving the ENA:b-CD complex formation (polar and non-po-
lar solvation energy) was also due to the higher affinity of the Z-cis
conformation of ENA when complexed to b-CD, since the energy
necessary to solvate the E-trans isomer was higher than that of
its counterpart, the Z-cis one (48.70 and 36.33 kcal/mol,
respectively).
2.2. Complex formation in solid state
2.2.1. IR studies
The reported lyophilization method resulted in a simple and
convenient way to obtain the complex between ENA and b-CD in
the solid state. The formation of this complex may be studied by
IR spectroscopy based on the ligand, since it possesses carbonyl
functional groups that present bands between 1680 and
1700 cmꢀ1, which can be significantly modified as a consequence
of complexation.13
Due to the fact that ENA showed three important bands at
1753, 1728 and 1647 cmꢀ1 (Fig. 12 a), which were attributed
to carbonyl stretching of the carboxylic acid, ester and tertiary
amide, respectively, then the modification of these signals may
be used to identify the establishment of intermolecular interac-
tions involving these groups, thus confirming the formation of
an inclusion complex. The infrared spectra of different samples
are presented in Figure 12. As shown in the spectra for the phys-
ical mixture, it consists of an overlay of the pure compound
spectra whereas the spectrum of the complex shows differences
from that of pure ENA. For the IR spectrum corresponding to the
ENA:b-CD complex, it is observed that the band at 1753 cmꢀ1 is
not present, but two new bands, due to the formation of the car-
boxylate anion, become visible at 1550 and 1400 cmꢀ1. Also, the
bands at 1728 and 1647 cmꢀ1 became broader and were shifted
to lower frequencies. This spectral shift may have originated
from the establishment of stable hydrogen bond interactions be-
Figure 11. Plots of the root-mean-square deviation (RMSD) versus time for the
three clusters of the ENA:b-CD complex subjected to molecular dynamics
simulations.
Table 4
Energetic component analysis obtained by MM-PBSA for the three clusters of the
ENA:b-CD complex analyzed by molecular dynamics simulations
Energetic component
Cluster (a)
(kcal/mol)
Cluster (b)
(kcal/mol)
Cluster (c)
(kcal/mol)
Electrostatic
Van der Waals
Gas state total
Polar and non-polar contrib.
ꢀ28.92
ꢀ25.38
ꢀ54.96
36.33
ꢀ95.34
ꢀ15.13
ꢀ111.04
95.04
ꢀ101.36
ꢀ6.99
ꢀ109.04
98.06
Free energy of binding
ꢀ18.63
ꢀ16.00
ꢀ10.98
is higher than that of the E-trans isomer. In order to study the
structural basis of this observation, the E-trans isomer was docked
to the b-CD molecule, and then the same molecular dynamic and
energetic component analysis was performed as described above.
The docking pose thus obtained was similar to that of the Z-cis iso-
Figure 12. IR spectra of (a) ENA, (b) b-CD, (c) ENA:b-CD physical mixture and (d)
ENA:b-CD freeze-dried complex.

8410
A. Zoppi et al. / Bioorg. Med. Chem. 16 (2008) 8403–8412
the characteristic events observed for the pure compounds were
found.
For the results obtained for the complex prepared by lyophiliza-
tion, the DSC curve shows that the endothermic peak of pure ENA
at 146 °C is not present, indicating that the melting event did not
take place, which might be due to changes in the crystalline form
of the solid, or to an inclusion complexation. Also, the TGA curve
shows that the decomposition of ENA starts at above 235 °C, a tem-
perature that is considerably higher than that for pure ENA
(153 °C). This fact indicates that the formation of the inclusion
complex considerably enhances the stability of ENA in solid state.
In addition, these observed curve patterns are in agreement with IR
spectroscopic data with respect to the formation of the inclusion
complex.
2.3. Stability studies
After performing the corresponding stability assay, we found
that the main degradation product of ENA in pH 7.0 buffer solution
was ET, with a retention time of 0.8 min. This observation was con-
sistent with bibliographic data.1
The corresponding degradation rates for free ENA (ko) and for
the ENA:b-CD complex at different b-CD concentrations (kobs) were
calculated (Fig. 15). In both cases, a pseudo first-order kinetic deg-
radation behavior was found, demonstrating that the presence of
b-CD did not influence the reaction order or the degradation
mechanism.
Figure 13. DSC curves of (a) ENA, (b) b-CD, (c) ENA:b-CD physical mixture, (d)
ENA:b-CD freeze-dried complex.
tween ENA and b-CD. Furthermore, a reduction of the intensities
of the vibrational bands of ENA was observed for the complex, a
fact that is consistent with the vibrational constrictions imposed
on the guest molecule in the cyclodextrin cavity. Based on this
evidence, we can confirm the existence of intermolecular inter-
actions between ENA and b-CD, and consequently the formation
of the corresponding inclusion complex.
Considering that we previously determined that the complex
has a 1:1 stoichiometry, we then applied the values obtained for
ko and kobs in order to calculate the corresponding formation con-
stant of the complex (Kc) and the degradation rate of the drug
within it (kc), at 70 °C as described by Loftsson and Brewster10 In
this way, free ENA exhibited a shelf life time (t90) of 6.7 h
(ko = 1.56 ꢁ 10ꢀ2 hꢀ1), while the drug within the complex showed
a t90 of 14 h (kc = 0.74 ꢁ 10ꢀ2 hꢀ1). This allows us to conclude that
the drug complexed with b-CD is 2.1 times more stable than its
free counterpart.
2.2.2. DSC and TGA studies
The DSC and TGA curves of pure ENA, pure b-CD, their physical
mixture, and the corresponding inclusion complex are shown in
Figures 13 and 14, respectively. The DSC curve of pure ENA shows
a sharp endothermic peak at 146 °C, corresponding to its melting,
thus demonstrating its crystalline anhydrous state. From 153 °C,
another endothermic shoulder is observed, with a maximum at
163 °C, while the TGA curve showed a loss of the mass fraction
of 29% during this temperature interval. As was previously pro-
posed by Lin et al.,17 the observed features for free enalapril male-
ate may be indicating a complex reaction that includes the
formation of diketopiperazine.
3. Conclusion
From the results present, we can conclude that the production
of an inclusion complex between enalapril maleate and b-CD was
achieved by applying a lyophilization method, which might be of
use at an industrial scale.
Taking into account the stability issues of ENA, we found that
the strategy applied in this work resulted in an enhanced stability
of the drug, not only in solid state as observed in the thermal anal-
ysis studies, but also in solution as described in the accelerated
degradation experiments. These observations suggest that the for-
mulation of ENA with b-CD may overcome the problems currently
described for solid dosage forms of this drug, and also provides an
interesting alternative for the preparation of orally administered
solutions of ENA.
When analyzing the DSC and TG curves of b-CD, it can be seen
that between 50 and 150 °C an endothermic event takes place,
with a weight loss of 11% of the mass fraction, corresponding to
its water content. Decomposition of b-CD occurs over 300 °C. Also
expected when analyzing the physical mixture of ENA and b-CD,
Additionally, we were able to successfully combine molecular
modeling studies with experimental spectroscopic assays, in order
to elucidate the molecular bases and thereby explains the complex
formation and the specific intermolecular interactions taking place.
The methodology applied might then be extensible to the prepara-
tion of b-CD inclusion complexes with other molecules that exhibit
similar stability issues, thus allowing a preliminary computational
study of the potential complex.
Finally, we can postulate that the preparation of an oral dosage
containing the ENA:b-CD complex could be very useful for the
treatment of essential and renovascular hypertension, although
further physical and chemical studies are needed to assess the bio-
availability of ENA from the complex.
Figure 14. TG curves of (a) b-CD (— —), (b) ENA (----), (c) ENA:b-CD physical
mixture (——) and (d) ENA:b-CD freeze-dried complex (— – —).

A. Zoppi et al. / Bioorg. Med. Chem. 16 (2008) 8403–8412
8411
0.016
0.015
0.014
0.013
0.012
0.011
0.010
-0.2
-0.4
-0.6
-0.8
-1.0
-1.2
-1.4
-1.6
10
20
30
40
50
60
70
80
90 100
0.000
0.002
0.004
0.006
0.008
0.010
0.012
β-CD Concentration (M x 103)
Time (h)
Figure 15. (a) Plot of pseudo first-order kinetic degradation of ENA, (j) free and in the presence of b-CD 11 mM (d) and (b) plot of kobs versus b-CD concentration for the
degradation of ENA.
against [b-CD]t, Kc could be calculated from the intercept obtained
by a least squares linear analysis. In this experiment, the concentra-
tion of ENA was kept constant at 2 mM, while b-CD concentration
was varied from 5 to 10 mM for the D2O of pD 7.0.
4. Materials and methods
4.1. Materials
2D ROESY: the geometry of the inclusion complex was studied
by two-dimensional Rotating frame Overhauser experiments (2D
ROESY), with spinlock for mixing phase sensitive, using 180x
180 ꢀ x pulses for polarization transfer. The spectra were mea-
sured with a relaxation delay of 2 s, p15 pulse for ROESY spinlock
Enalapril maleate, D2O 99.9 atom% D, and NaOD 40% used in
spectroscopic studies were purchased from SigmaÒ, while b-CD
(MW = 1135) was kindly supplied by Roquette (France). All other
materials and solvents were of analytical reagent grade. A Milli-Q
Water Purification System (MilliporeÒ, USA) generated the water
used in these studies. Also a H2KPO4/HK2PO4 pH 7.0 buffer was
used (pH-meter ORION SA520, USA), with its ionic strength being
adjusted to 0.5 by addition of KCl.
*
of 20 ms and l4, spinlock loop, (p15/p25 2) = 400.
Before Fourier transformation, the matrix was zero filled to
4096 (F2) by 2048 (F1) and Gaussian apodization functions were
applied in both dimensions. Working concentrations of both ENA
and b-CD were 5 mM in D2O, at pD 7.0.
4.2. NMR studies
All experiments were performed on a BrukerÒ Avance II High
Resolution Spectrometer, equipped with a Broad Band Inverse
probe (BBI) and a Variable Temperature Unit (VTU). The spectra
were measured at 298 K, adjusting the solutions to pD 7.0 by the
addition of NaOD/D2O.
4.3. Molecular modeling procedures
The initial molecular structure of ENA was built using the
graphical interface Gabedit, after which a systematic conforma-
tional search was performed in order to obtain the minimum en-
ergy conformation. In this way, a semiempirical method (AM1)
was applied during the conformational search, with the final min-
imum energy conformation being optimized using an ab-initio
(HF-6-31G*) method. Both semiempirical and ab-initio methods
are implemented in the Gaussian 98 software.19 The initial struc-
ture of b-CD was obtained from the Cambridge Structural Database
(PDB code BCDEXD10).
4.2.1. 1H NMR
Spectra were obtained at 400.16 MHz, with the chemical shift of
the residual solvent at 4.8 ppm used as an internal reference. In-
duced changes in the 1H chemical shifts for ENA and b-CD (
Dd)
originated due to their complexation were calculated using the fol-
lowing equation:
For the molecular docking procedure, restrained electrostatic
potential (RESP) fitted charges were assigned to ENA, while for b-
CD, charges corresponding to the Glycamm04 force field imple-
mented in the Amber9 software package were assigned.20 Auto-
dock v3.05 was used to perform the dockings runs,21 by applying
a Lamarckian Genetic Algorithm (LGA) to generate the population
evaluated, using the following parameters: a population of 150
individuals, a maximum numbers of 250,000,000 evaluations and
5000 generations as an end criterion. An elitism value of 1 was se-
lected, with a probability of mutation and crossing over of 0.02 and
0.8, respectively. Thirty docking runs were performed, and the
three lowest energy cluster conformations were selected for pos-
terior molecular dynamics assays.
The Amber9 software package was used for the molecular
dynamics studies.20 Atomic charges for ENA and b-CD were identi-
cal to those used for the molecular docking procedures described
above. The complexes obtained by molecular docking were sol-
vated with a pre-equilibrated TIP3P octahedric water box, by per-
forming an initial minimization of the solvent followed by a
minimization of the whole system. These minimized systems were
D
d ¼ dcomplex ꢀ dfree
ð1Þ
The stoichiometry of the complex was studied by applying the con-
tinuous variation method,14 using the NMR spectra obtained for a
series of ENA and b-CD mixtures in which the sum of their concen-
trations was kept constant at 10 mM, but varying their molecular
fraction between 0 and 1. After registering the corresponding
changes in the most sensitive NMR signals, the stoichiometry of
the inclusion complex was calculated.
The apparent formation constant (Kc) of the inclusion complex
in solution was determined by applying Scott’s equation (Eq.
(2))18
:
½b-CDꢂt ½b-CDꢂt
1
cDdc
¼
þ
ð2Þ
D
dobs
D
dc
K
where [b-CD]t is the total b-CD concentration;
ence between the chemical shift of free ENA and the corresponding
chemical shift of ENA in the presence of b-CD at each concentration;
D
D
dobs is the differ-
dc is the difference between the chemical shift of a pure sample of
the complex and its free components. By plotting [b-CD]t / dobs
D

8412
A. Zoppi et al. / Bioorg. Med. Chem. 16 (2008) 8403–8412
heated to 300 K for 20 picoseconds (ps), followed by a 100 ps equil-
ibration run. Production runs were performed for 13 nanoseconds
(ns), using a timestep of 2 femtoseconds (fs) and under constant
pressure and temperature, with the SHAKE algorithm being ap-
plied to constrain bonds involving hydrogens. The trajectories ob-
tained were monitored by plotting temperature, total energy,
potential energy and root-mean-square deviation (rmsd) of the
corresponding structures.
The hydrogen bond and energetic component analyses were
performed using the Ptraj and Molecular Mechanics Poisson-Bolz-
mann Surface Area (MM-PBSA) modules of Amber9, respectively.20
MM-PBSA analyses were performed over a 4 ns segment, with indi-
vidual snapshots sampled every 10 frames of the trajectory. The
corresponding structures were visualized using the VMD soft-
ware.22 Molecular dynamics simulations were performed at the
Fimm Cluster resource (Parallab, Bergen Center for Computational
Science, University of Bergen, Norway).
150 mm, 5 lm, Agilent Technologies, USA) reversed-phase column.
The mobile phase was an acetonitrile:pH 2.2 potassium phosphate
buffer (10 mM) 32:68 mixture, at a flow rate of 2.0 ml/min. Assays
were performed at 60 °C, by injecting 50 ll of solution in each chro-
matographic run. Under these conditions, the retention times for ET,
ENA and DKP were 0.8 min, 1.5 min and 5.0 min, respectively.
The observed first-order rate constant for degradation of free
ENA (k0) and ENA complexed with b-CD (kobs) were determined
by applying a linear regression analysis for the plots of the natural
logarithm of the remaining ENA concentration versus time. The
complex formation constant (Kc) and the degradation of the drug
within the complex (kc) were calculated from Lineweaver–Burk
plots.10
Acknowledgments
The authors thank the Secretaría de Ciencia y Técnica de la Uni-
versidad Nacional de Córdoba (SECyT), and the Consejo Nacional de
Investigaciones Científicas y Tecnológicas de la Nación (CONICET)
for financial support. We also thank the Ferromet S.A. (agent of Ro-
quette in Argentina) for their donation of b-cyclodextrin. We
would also like to specially thank Dr. Petter Bjorstad (Research
Director at Bergen Center for Computational Science (BCCS)) for
kindly providing the access to BCCS computing resources, and Dr.
Paul Hobson, native speaker, for revision of the manuscript.
4.4. Preparation of the ENA:b-CD complex in solid state
A solution of ENA and b-CD (1:1 molar ratio) was prepared in
distilled water, and the pH 7.0 was adjusted by addition of NaOH
1 N. The resulting solution was placed in an ultrasonic bath for
1 h, after which it was incubated at 25.0 ( 0.1) °C in a thermostatic
water bath (Circulators HAAKE F3-K, Germany) for 48 h. Then, the
solutions were frozen at ꢀ40 °C before being freeze-drying started
(Freeze Dri 4.5 Labconco Corp., Kansas City, MI).
References and notes
1. Analytical Profiles of Drug Substances; Ip, D. P., Brenner, G. S., Florey, K., Eds.;
Academic Press: New York, 1987; Vol. 16.
4.5. Fourier transform infrared spectroscopy (FT-IR)
2. Al-Omari, M. M.; Abdelah, M. K.; Badwan, A. A.; Jaber, A. M. Y. J. Pharm. Biomed.
Anal. 2001, 25, 893.
3. Wang, S. L.; Lin, S. Y.; Chen, T. F.; Cheng, W. T. Pharm. Res. 2004, 21, 2127.
4. Cotton, M. L.; Wu, D. W.; Vadas, E. B. Int. J. Pharm. 1987, 40, 129.
5. Stanisz, B. J. Pharm. Biomed. Anal. 2003, 31, 375.
6. Oliva, M. A.; Sombra, L. L.; Olsina, R. A.; Masi, A. N. J. Fluoresc. 2005, 15, 723.
7. Bhardwaj, S. P.; Singh, S. J. Pharm. Biomed. Anal. 2008, 46, 113.
8. Simoncic, Z.; Zupancic, P.; Roskar, R.; Gartner, A.; Kogej, K.; Kmetec, V. Int. J.
Pharm. 2007, 342, 145.
The FT-IR spectra of ENA and its complex with b-CD were mea-
sured as potassium bromide disks on a Nicolet 5 SXC FT-IR Spec-
trometer. The FT-IR spectrum of ENA:b-CD was compared with
those of the corresponding 1:1 molar ratio physical mixture and
also with pure ENA and b-CD. All spectra were obtained and pro-
cessed using EZ OMNIC E.S.P v.5.1 software.
9. Pilatti, C.; Ercolano, I.; Torre, M.; Del, C.; Chiale, C.; Spinetto, M. Drug Dev. Ind.
Pharm. 1999, 25, 807.
10. Loftsson, T.; Brewster, M. E. J. Pharm. Sci. 1996, 85, 1017.
11. Vianna, R. F. L.; Bently, M. V. L. B.; Riebeiro, G. Int. J. Pharm. 1998, 167, 205.
12. Li, J.; Guo, Y.; Zografi, G. J. Pharm. Sci. 2002, 91, 229.
4.6. Differential scanning calorimetry (DSC) and
thermogravimetric analysis (TGA)
The DSC curves of the different samples were recorded on a DSC
TA 2920 and the TGA curves on a TG TA 2920, both by applying a
heating rate of 10 °C minꢀ1. The thermal behavior was studied over
a temperature range of 25–400 °C, by heating 1–3 mg of samples in
aluminum-crimped pans under nitrogen gas flow. Data were ob-
tained and processed using the TA Instruments Universal Analysis
2000 software.
13. Fromming, K. H.; Szejtli, J. Cyclodextrins in Pharmacy; Kluwer Academic
Publishers: Dordrecht, 1994.
14. Job, P. Ann. Chim. 1928, 9, 113.
15. Shoji, A.; Yanagida, A.; Shindo, H.; Ito, Y.; Shibusawa, Y. J. Chromatogr. A 2007,
1157, 101.
16. Scheider, H. J.; Hacket, F.; Rudiger, V.; Ikeda, H. Chem. Rev. 1998, 98, 1755.
17. Lin, S. Y.; Wang, S. L.; Chen, T. F.; Hu, T. C. Eur. J. Pharm. Biopharm. 2002, 54, 249.
18. Ganza-Gonzalez, A.; Vila-Jato, J. L.; Anguiano-Igea, S.; Otero-Espinar, F. J.;
Blanco-Mendez, J. Int. J. Pharm. 1994, 106, 179.
19. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Stratmann, R. E.;
Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin, K. N.; Strain, M.
C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.;
Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.;
Cui, Q.; Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;.
Foresman, J. B; Cioslowski, J.; Ortiz, J. V.; Baboul, A. G.; Stefanov, B. B.; Liu,
G.; Liashenko, A. P.; Piskorz, I.; Komaromi, R.; Gomperts, R. L.; Fox Martin, D. J.;
Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.;
Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Andres,
J. L.; Gonzalez, C.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Software:
GAUSSIAN 98; Gaussian, Inc.: Pittsburgh, PA, USA, 1998. www.gaussian.com.
20. Case, D. A.; Cheatham, T. E., III; Darden, T.; Gohlke, H.; Luo, R.; Merz, K. M.;
Onufriev, A., Jr.; Simmerling, C.; Wang, B.; Woods, R. J. Comput. Chem. 2005, 26,
1668.
4.7. Stability studies
The chemical stability of ENA complexed to b-CD was deter-
mined at 70 °C in pH 7.0 phosphate buffer solutions containing
b-CD concentrations of between 0 and 11 mM. Solutions were pre-
pared by dissolving 20 mg of enalapril maleate in 100 ml of the
aqueous buffer containing the desired b-CD concentration, which
were then placed in a thermostated water bath (Circulators HAAKE
F3-K, Germany). Samples were taken at various time intervals.
The quantitation of ENA was carried out by applying the HPLC
method reported by Pilatti et al.,9 with minor modifications. The
HPLC equipment consisted of an Agilent S1100 system with UV
21. Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart, W. E.; Belew, R. K.;
Olson, A. J. J. Comput. Chem. 1998, 19, 1639.
22. Humphrey, W.; Dalke, A.; Schulten, K. J. Mol. Graph. 1996, 14, 33.
detection at 215 nm, using
a
Zorbax Eclipse XDB-C8 (4.6 ꢁ