Solubility profiles, hydration and desolvation of curcumin complexed with γ-cyclodextrin and hydroxypropyl-γ-cyclodextrin

Article abstract of DOI:10.1016/j.molstruc.2016.12.028

In this study, we investigated curcumin (CUR) solubility profiles and hydration/desolvation effects of this substance formulated with γ-cyclodextrin (γ-CD) and hydroxypropyl-γ-cyclodextrin (HP-γ-CD) excipients. The CUR/HP-γ-CD complex was found to be more stable in solution with the highest apparent stability constant for CUR/HP-γ-CD (K_c?=?1.58*10⁴?M^?1) as the more soluble form in distilled water. The in silico calculations, including molecular docking, Monte Carlo (MC), and molecular dynamics (MD) simulations, indicated that water molecules play an important role in host-guest complexation mediating the CUR binding to cyclodextrins via hydrogen bond formations. The CUR hydration/desolvation effects contributed to the complex formation by elevating the CUR binding affinity to both CDs. The CUR/HP-γ-CD complex after the CUR hydration was determined with a minimal Gibbs free energy of binding (ΔG_bind?=??9.93?kcal*mol^?1) due to the major hydrophobic (vdW) forces. Overall, the results of this study can aid a development of cyclodextrin-based drug delivery vectors, signifying the importance of water molecules during the formulation processes.

Full text of DOI:10.1016/j.molstruc.2016.12.028

Accepted Manuscript
Solubility profiles, hydration and desolvation of curcumin complexed with γ-
cyclodextrin and hydroxypropyl-γ-cyclodextrin
Sergey Shityakov, Ramin Ekhteiari Salmas, Serdar Durdagi, Norbert Roewer, Carola
Förster, Jens Broscheit
PII:
S0022-2860(16)31318-7
10.1016/j.molstruc.2016.12.028
MOLSTR 23226
DOI:
Reference:
To appear in: Journal of Molecular Structure
Received Date: 22 July 2016
Revised Date: 6 December 2016
Accepted Date: 9 December 2016
Please cite this article as: S. Shityakov, R.E. Salmas, S. Durdagi, N. Roewer, C. Förster,
J. Broscheit, Solubility profiles, hydration and desolvation of curcumin complexed with γ-
cyclodextrin and hydroxypropyl-γ-cyclodextrin, Journal of Molecular Structure (2017), doi: 10.1016/
j.molstruc.2016.12.028.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
1
2
Solubility profiles, hydration and desolvation of curcumin complexed with
γ-cyclodextrin and hydroxypropyl-γ-cyclodextrin
3
4
Sergey Shityakov1,*, Ramin Ekhteiari Salmas3, Serdar Durdagi3, Norbert Roewer1,2,
Carola Förster1, and Jens Broscheit1,2
5
1Department of Anesthesia and Critical Care, University of Würzburg, 97080 Würzburg,
6
Germany
7
2Sapiotec Ltd., 97078 Würzburg, Germany
8
3Department of Biophysics, School of Medicine, Bahcesehir University, Istanbul, Turkey
Author to whom correspondence should be addressed; E-Mail: Shityakov_s@ukw.de
Tel: +49-931-2013-0016; Fax: +49-931-2013-0019.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Graphical Abstract:
Abstract: In this study, we investigated curcumin (CUR) solubility profiles and
hydration/desolvation effects of this substance formulated with γ-cyclodextrin (γ-CD) and
hydroxypropyl-γ-cyclodextrin (HP-γ-CD) excipients. The CUR/HP-γ-CD complex was found
to be more stable in solution with the highest apparent stability constant for CUR/HP-γ-CD
(Kc = 1.58*104 M-1) as the more soluble form in distilled water. The in silico calculations,
including molecular docking, Monte Carlo (MC), and molecular dynamics (MD) simulations,
indicated that water molecules play an important role in host-guest complexation mediating the
CUR binding to cyclodextrins via hydrogen bond formations. The CUR hydration/desolvation
effects contributed to the complex formation by elevating the CUR binding affinity to both
CDs. The CUR/HP-γ-CD complex after the CUR hydration was determined with a minimal
Gibbs free energy of binding (ꢀGbind = ‒9.93 kcal*mol-1) due to the major hydrophobic (vdW)
forces. Overall, the results of this study can aid a development of cyclodextrin-based drug
1

ACCEPTED MANUSCRIPT
delivery vectors, signifying the importance of water molecules during the formulation
processes.
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
Keywords: curcumin, cyclodextrins, complex stability, hydrated molecular docking,
hydrogen bonds, Monte Carlo, molecular dynamics simulations, free energy calculations
1. Introduction
Curcumin (CUR) is a natural chemical produced by some plants of the ginger family, and it is
used as a dietary herbal supplement or a food coloring agent (Volate et al., 2005; Baumann et
al., 2009). This substance has also been found to possess antiviral and anticancer properties,
which were intensively characterized over the last few decades (Nagabhushan and Bhide,
1992; Prasad and Tyagi, 2015). The CUR compound is essentially insoluble at physiological
pH and, since it is very unstable, it undergoes a rapid hydrolytic degradation (Tonnesen et al.,
2002). Various attempts have been made to produce water-soluble and more stable CUR by
formulating it with different excipients, including hydrophilic cyclodextrins (CDs) to
minimize its hydrolysis and high decomposition rate (Yadav et al., 2009; Marcolino et al.,
2011).
CDs have a wide range of biomedical applications in fields such as pharmacy, chemistry,
biotechnology, and medicine. They are oligosaccharides produced by bacteria from starch via
its enzymatic degradation and typically exist as pristine hexameric α-, heptameric β- and
octameric γ-CD forms and their derivatives (Zidovetzki and Levitan, 2007). CDs have a
hydrophilic outer surface, and a less polar but more lipophilic (amphiphilic) central cavity
leading to a CD inclusion complexation with different drug-like substances through the
electrostatic, van der Waals (vdW), hydrophobic, charge-transfer, and H-bonding interactions.
The CD-drug complex formation usually contributes to a retardation of the degradation rate
during solvation process (Tonnesen et al., 2002). This complexation process mainly depends
on the interaction of the guest molecule with CDs and the difference in the interactions of
bound water and water with the bulk solvent. Therefore, studies on molecular recognition and
binding require careful consideration of solvent effects.
The solvated drug molecules interact with waters before they bind to CDs with a subsequent
release of water molecules to the bulk waters in aqueous solution. This so-called “desolvation
process” is usually unfavorable during the change in Gibbs free energy of binding (ꢀGbind
)
and described as a “desolvation” penalty (Baldwin, 2010). However, some water molecules
2

ACCEPTED MANUSCRIPT
67
68
69
70
71
72
73
74
75
76
77
78
79
80
are not displaced, and they might mediate the CD-drug complexation via altering binding site
topography followed by the increase in overall binding affinity.
To tackle this issue, different quantitative and qualitative approaches have been established
and applied to assess the energy contribution implied by the presence or displacement of
water molecules, such as the hydrated ligand molecular docking (Forli and Olson, 2012). In
particular, the AutoDock method was recently revised to evaluate the solvation and
desolvation phenomena via including explicit displaceable waters during molecular docking
procedure to improve overall docking precision and scoring function without excessive
computational needs (Forli and Olson, 2012). On the other hand, to the best of our knowledge,
there is no report on the hydration effect of CUR complexed with cyclodextrins so far.
Therefore, in the present study, we have investigated the solubility profiles and the CUR
hydration/desolvation process contribution to the CUR complexation with γ-CD and
hydroxypropyl-γ-CD (HP-γ-CD) using combined experimental and computational techniques.
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
2. Materials and Methods
2.1. Materials
CUR was purchased from Sigma-Aldrich GmbH (Steinheim am Albuch, Germany). The
CUR/γ-CD and CUR/HP-γ-CD complexes as pure forms (98% purity) were prepared as per
the following protocols: 45 g of γ-CD and 5 g of CUR were weighed and suspended in 25 ml
distilled water with a mortar and a pestle. Similarly, 1 g of HP-γ-CD and 44 mg of CUR were
weighed and suspended in 1.5 ml of 96% ethanol and were homogenized as above. After
complete homogenization, the suspension was dried in vacuo under ambient conditions for
three days. Finally, the dried complexes were pulverized in the mortar.
2.2. Quantification of curcumin
The CUR concentrations were determined by UV/Vis spectrophotometry using an Agilent
8453 spectroscopy system (Agilent Technologies, Budapest, Hungary). The samples were
diluted with a 50 vol% ethanol-water mixture to yield an absorbance recordable at the 430 nm
range. All CUR measurements were performed without any interference from CDs presented
in the complex.
2.3. Solubility studies and determination of stability constants
3

ACCEPTED MANUSCRIPT
99
100
101
102
103
104
105
The solubility method was carried out according to the Higuchi & Connors method (Higuchi
and Connors, 1965). Solubilities were measured by adding an excess amount of curcumin to
distilled water containing different amounts of various kinds of cyclodextrins. The suspension
formed was equilibrated under continuous agitation for 24 h at 25 ± 3.0 °C and then filtered
through a 0.45 µm nominal pore size PVDF filter to yield a clear curcumin solution. The
apparent stability constant (Kc) for a CUR-CD complexes was obtained from the slope of the
phase-solubility diagram according to the following equation:
(1)
slope
Kc =
S0 (1− slope)
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
where S0 is the saturation concentration of CUR in the solvent without cyclodextrin. The data
are represented as means ± S.D. of three independent experiments.
2.4. Molecular docking and semi-empirical calculations
The 3D coordinates of the curcumin structure were retrieved from the PubChem database
(Figure 1[A]). Since no 3D structure for HP-
γ-CD was available, the molecule was
constructed from the -CD (PDB ID: 1P2G) crystal structure (Pinotsis et al., 2003) with the
γ
PyMol v.1.2 software (Figure 1 [B]). Prior to molecular docking and semi-empirical
calculations, the all-simulated structures were minimized with the GTKDynamo v.1.8.1
software (Bachega et al., 2013) using the conjugate gradient method with 200 maximum
iterations and a threshold tolerance gradient of 0.1 Å. The host and guest structure
preparations for molecular docking included the Gasteiger partial charge assignment
(Gasteiger et al., 1980) and rotatable bonds definition. Rigid-flexible standard and hydrated
molecular docking were applied to the center of the cyclodextrin structure. AutoDock
v.4.2.5.1 (Goodsell et al., 1996; Forli et al., 2015) integrated into the PyMol AutoDock/Vina
plugin (Seeliger and Groot, 2010) was used in the study. The grid spacing of 0.375 Å, with a
dimension size of 60 Å from x, y, and z, was used to create the grid maps. In order to increase
a conformational sampling of the drug, a number of genetic algorithm dockings (ga_run) were
set to 100. Docking output results were represented by the approximation function as the
estimated Gibbs free energy of binding (ꢀGbind). The calculated octanol/water partition
(ClogP) coefficient for curcumin using the weighted algorithm with electrolyte concentration
of 0.1 mol/dm3 and its aqueous solubility (ClogS) were determined by the Marwin Sketch
v.14.7.14.0 tool. The single-point calculations for all simulated structures using the Austin
4

ACCEPTED MANUSCRIPT
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
Model 1 (AM1) method (Dewar et al., 1985; Rocha et al., 2006) were performed using
GTKDynamo v.1.8.1 software (Bachega et al., 2013).
2.5. Atomistic molecular dynamics (MD) simulations
The docking poses of CUR produced by AutoDock v.4.2.5.1 (Goodsell et al., 1996; Forli et
al., 2015) were used as an initial structure suitable for MD simulations. All classical MD
simulations were performed using a GPU-accelerated version of Desmond 2015.4 code
(Bowers et al., 2006). The OPLS 2005 force field (Banks et al., 2005) was used to calculate the
interactions between atoms. The long-range electrostatic interaction calculations were set up
by using the particle-mesh Ewald (PME) method (Essmann et al., 1995). The short-range van
der Waals (vdW) and Coulomb interactions were defined by a cut-off radius of 9.0 Å. A Nose-
Hoover thermostat (Hoover et al., 1985) and Martyna-Tobias-Klein method (Martyna et al.,
1994) were used to maintain the system at the body temperature of 310 K and the pressure of
1.01325 bars. A time step of 2.0 fs was used during the MD simulations. The systems were
minimized using 2000 iterations, and a convergence threshold of 1 kcal*mol-1*Å-1. To perform
the structural relaxation, all the systems were equilibrated using default algorithms
implemented in the Desmond software. Finally, 100 ns atomistic MD trajectories were
produced for each of the analyzed systems.
2.6. Monte Carlo simulations
All Monte Carlo (MC) simulations implemented in ROSETTA code (Meiler et al., 2006) were
employed in order to generate different systematic conformers for the standard and hydrated
variants of CUR bound to γ-CD and HP-γ-CD. The trajectory frames as obtained by the MC
protocol were subjected to conformational clustering to be used further for the Molecular
Mechanics with Generalized Born and Surface Area Solvation (MM-GBSA) energy analysis
(Gohlke et al., 2004).
2.7. MM-GBSA calculations
The MM-GBSA method (Gohlke et al., 2004) was applied to estimate the free energies of
binding (ꢀGMM-GBSA) for the CUR-CD complexes. The Prime module in the Schrödinger
package (Schrödinger Release 2014-2: Prime, version 3.6) was used to calculate the ꢀGMM-
GBSA values. The VSGB 2.0 solvation model (Li et al., 2011) was used to simulate the implicit
solvation for the analyzed systems. The ꢀGMM-GBSA values were determined by subtracting the
5

ACCEPTED MANUSCRIPT
162
163
total individual free energies of cyclodextrin (ꢀGCD) and CUR (ꢀGCUR) from the complex
free energy (ꢀGcomplex) as depicted according to the following equation:
(2)
∆GMM −GBSA = ∆Gcomplex − (∆GCD + ∆GCUR
)
164
or re-written as a sum of the energetic terms:
∆GMM −GBSA = ∆GCoul + ∆Gcov + ∆GH −bond + ∆Glipo + ∆Gsolv + ∆GvdW
(3)
165
where ∆GCoul ,∆Gcov ,GH −bond ,∆Glipo,∆Gsolv and ∆GvdW are Columb, covalent, hydrogen-bond,
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
lipophilic, GB electrostatic solvation and vdW energies, respectively.
3. Results and Discussion
Molecular modelling techniques, such as molecular docking, have been proven useful in
predicting the binding modes and interaction profiles of drug-like molecules complexed with
different target structures including CDs (Ansari et al., 2012; Kar et al., 2013; Ahsan et al.,
2015; Shityakov et al., 2016). In the standard and hydrated molecular docking experiments,
the enol form of CUR, that is more energetically stable in the solid phase (Manolova et al.,
2014) because of intramolecular hydrogen bonding, was chosen as the ligand. While
γ-CD
and HP- -CD were selected to be the excipient macromolecules. Taking into account the
γ
flexibility of CUR and CD, different conformational variants for the complexes are possible.
By adding the ligand molecule to the CDs, 100 docking conformation instances for each
CUR/CD pair were generated.
The conformation of a guest compound can depend strongly on the presence of solvent
molecules participating in the binding and mediation of the interactions between different
ligand substitutes and the receptor (Forli and Olson, 2012). For this reason, studies of
molecular complexation involving cyclodextrins require careful consideration of the solvent
effects. Nonetheless, the formation of the CUR/CD complex takes place primarily in the
aqueous phase, and is followed by the drying step under ambient conditions. Even then, the
complex still contains approximately 6.9% residual water, established by the Karl Fischer
titration method.
In an aqueous environment, the
γ
-CD cavity with the largest volume (427 Å3) in comparison
to other CDs (174 Å3 for -CD and 262 Å3 for
α
β
-CD) is occupied with water molecules; and
most or all of these water molecules are excluded from the cavity upon binding with a guest
compound. This mildly lipophilic cavity, which contains about 9 weakly-held and easily
displaced water molecules with higher enthalpy than the bulk waters (Tabushi et al., 1978;
6

ACCEPTED MANUSCRIPT
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
Tanhuanpää et al., 2001). Also, because of the H-bond bivalency of this binding site, a water
molecule can “invert” a receptor's hydrogen bond acceptor region into a donor (Forli and
Olson, 2012). The cavity is further stabilized by intramolecular H-bonds between the adjacent
primary and secondary hydroxyl groups in γ-CD and hydroxypropyl groups in HP-γ-CD.
Water molecules were attached to CUR before docking by hydrating all polar groups capable
of hydrogen bond interactions. This is important information in de novo or early stages of
drug design, to optimize the ligands to best fit the binding site. As already mentioned in the
literature, γ-CD, and its derivative can be topologically described as torus-shaped molecules
with the larger and the smaller entries, where the hydroxyl groups are exposed to the solvent
(Munro et al., 2004). Previous studies (Szejtly, 1998; Uekama et al., 1998; Ma et al., 2000;
Zhao et al., 2002) have verified that the ligand molecule is usually inserted into CDs via the
larger outer rim opening (0.85 nm) instead of the smaller (0.75 nm) one. In fact, our docking
simulations demonstrate a similar scenario where CUR was inserted into the binding cavity of
the amphiphilic CDs from the larger interface in both the standard and hydrated variants. The
water molecules remained in the vicinity of the outer rim, and some of them were deeply
submerged into the cyclodextrin binding cavity, interacting with the enolic hydroxyl group of
CUR via H-bonds (Figure 2 and 3).
Based on the above observation of the CUR docking poses, only the complex structure with
predicted the lowest ꢀGbind value in the binding region was considered as the top-docking
result. The results are listed in Table 1 for standard and hydrated molecular dockings. CUR
was found to bind strongly to CDs exceeding the binder/non-binder energy threshold (non-
binders < ‒6.0 kcal*mol-1 < binders), where this threshold for various drug-like molecules in
other AutoDock experiments has been determined (Shityakov et al., 2012; Shityakov et al.,
2014). The CUR hydration improves its binding affinity to both cyclodextrins, this was
confirmed by a minimal predicted Gibbs free energy of binding (ꢀGbind) values. The CUR
hydration/desolvation effects contributed to the complex formation by elevating the CUR
binding affinity to both CDs. The CUR/HP-γ-CD complex after the CUR hydration was
determined with a minimal Gibbs free energy of binding (ꢀGbind = ‒9.93 kcal*mol-1) and Kd
of 0.05 µM. The thermodynamic equilibrium constants (Kd) for all of the docking poses were
calculated from the ꢀGbind values as follows: Kd = exp([ꢀGbind*1000]/[R*T]), where R (gas
constant) is 1.98 cal*(mol*K)-1 and T (room temperature) is 298.15 Kelvin (Shityakov et al.,
2012).
7

ACCEPTED MANUSCRIPT
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
The increase in CUR binding affinity to CDs after hydration might be explained by the higher
number of intermolecular H-bonds formed between the host and the guest: 4 for CUR/γ-CD
and 2 for CUR/HP-γ-CD without ligand hydration (Figure 2 [A] and 3 [A]); and 8 for CUR/γ-
CD and 4 for CUR/HP-γ-CD with ligand hydration (Figure 2 [B] and 3 [B]). Thus, the CUR
hydrated docking to CDs illustrates the capabilities of the new hydration method to predict the
position of weakly bound water molecules; and we ranked them in accordance with
experimental findings, improving the docking accuracy.
The minimal predicted ꢀGbind and Kd values indicated that very lipophilic CUR (logP = 3.28)
compound (Pawar et al., 2012) with very low calculated water solubility (0.01 – 0.06 mg*ml-
1) appeared to be the strongest binder to HP-γ-CD in both variants of molecular docking. As a
result of the higher hydrophobicity potential (ClogP = ‒5.30) defined for this modified
cyclodextrin than for its parental pristine form (ClogP = ‒14.17).
Following this, the top-docked poses of CUR were analyzed for whether the pose was still the
likely bound pose and there are 99 other possible host-guest configurations. Since, there are
several commonly occurring docked poses formed during the clustering with the root-mean-
square-deviation (RMSD) cut-off value of 2.0 Å, the most popular pose was present multiple
times for the standard and hydrated docking simulations. Additionally, when the guest
interacts with water before it binds to the host, then those water molecules must be displaced
to be further released into the bulk of the waters molecules in the solution (Furuki et al.,
1993). The change in ꢀGbind in this process is often unfavorable and considered to be the
“desolvation” penalty (Furuki et al., 1993). The determination of water to be displaced or
bridged is the balance between its energetic contribution to ligand-receptor binding and the
ability to stabilize a ligand pose through its displacement (Forli and Olson, 2012). Following
this rule, during the hydrated complexation of CUR with the γ-CD and HP-γ-CD molecules
most of the water molecules were displaced because of high water free energy (ꢀGwat = ‒0.2
kcal*mol-1) and only few were strongly (ꢀGwat = ‒0.62 kcal*mol-1) or weakly (ꢀGwat = ‒0.36
kcal*mol-1) attached via H-bonds to the inclusion complex (Table 2). Clearly, the CUR
inclusion in the CD cavity cannot lead to the displacement of all water molecules presented in
it. The more closely a guest can be fitted in the CD binding site, the greater will be the
number of water molecules released into the bulk. Thus, the combination of these two effects,
such as CUR inclusion and water displacement, impacts the equilibrium formation constants
for CD formulated drugs in solution.
8

ACCEPTED MANUSCRIPT
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
Quantitation of CUR was also determined from the solubility isotherms of the CUR/γ-CD and
CUR/HP-γ-CD complexes in a distilled water. The solubility diagrams for all complexes were
non-linear Cauchy–Lorentz distribution for CUR/γ-CD and showed second-order polynomial
(quadratic) curve fitting for CUR/HP-γ-CD (Figure 4 [A, B]). The CUR concentration in
CUR/HP-γ-CD was about 60-fold greater than that of γ-CD complex due to improved
aqueous solubility. The CUR/HP-γ-CD solution was supersaturated when the concentration of
HP-γ-CD was higher than 11.6 mM. The CUR/γ-CD precipitate formation was started at 1.9
mM of γ-CD disturbing the CUR release probably by binding to its free fraction. The plateau
section of the curve suggested that the precipitation of the complex was finally attained. An
estimate of Kc constant was calculated using the analytical detection limit as highest possible
S0 value (S0 = 2.37 µM) showing the increase in apparent stability for CUR/HP-γ-CD (Kc =
1.58*104 M-1) in contrast with the CUR/γ-CD complex (Kc = 1.02*103 M-1). Observations
from previous complexation studies on curcuminoids also suggested that the bulky moieties
of the two phenyl groups of CUR fit better to the bigger γ-CD that β-CD cavity, where the Kc
values of CUR in HP-β-CD and HP-γ-CD was reported to be more than 5.0*104 M-1 and
16*104 M-1, respectively (Tonnesen et al., 2002).
To compare our experimental and theoretical results, we calculated the reference binding ratio
(BRref) to determine the complexation strength, which is based on the relative content of CUR
in the γ-CD (
) or in HP-γ-CD (
−CD ) complex, molecular weight of the complex
CHP−
γ
Cγ
−CD
275
constituents and degree of substitution (n) for HP-γ-CD using the following equation:
CHP− −CD (MWCUR + MWHP− + n MWres )
(4)
γ
γ −CD
BRref
=
Cγ −CD (MWHP−
+ MWHP−
)
−CD
γ
−CD
γ
276
277
278
279
280
281
282
where MWCUR, MWCD, and MWres are the molecular weights of curcumin,
γ
-CD, HP-
γ
-CD and
hydroxypropyl residues In reality, the HP-γ-CD structure is randomly substituted having
.
hydroxypropyl substituents at all positions of primary (C6 position) and secondary (C2 and
C3 position) interface. To simplify the molecular docking procedure, all hydroxyl groups (n =
24) of the
CD model. Alternatively, the theoretical binding ratio (BR) was simply defined as the
ratio between the binding affinities of the CUR/ -CD and CUR/HP- -CD complexes as:
γ-CD molecule were substituted for the hydroxypropyl groups in the case of HP-γ-
ꢀGbind
γ
γ
(5)
∆GCUR / HP−
∆GCUR /
γ
−CD
BR =
γ
−CD
283
284
As shown in Figures 3 and 4 , the CUR position is roughly detected to be docked in the center
of the CD structures using 1:1 complex stoichiometry, where the binding cavity is located.
9

ACCEPTED MANUSCRIPT
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
This stoichiometry was confirmed via a good correlation between the experimental data and
the 1:1 binding isotherm for the CUR/CD complexes (Hegge et al., 2009). However, the
formation of inclusion complexes with 1:2 (guest:host) stoichiometry is possible for
CUR/HP-γ-CD as it was observed in the CUR absorption spectra sing UV-Vis spectroscopy
(Hegge et al., 2009). The theoretical molecular docking results were in good agreement with
the experimentally derived data judging by the theoretical binding ratio for the standard (BR =
1.15) and hydrated (BR = 1.29) molecular docking in comparison with the reference (BRref =
1.37).
Furthermore, the three monomers (CUR, γ-CD and HP-γ-CD) and four inclusion complexes,
including those with the hydrated ligand were minimized using the conjugate gradient
approach in order to perform the single-point AM1 calculation on the assessment of their
relative chemical stability. The AM1 energies of 7 species and the energy difference (ꢀE)
between the inclusion complexes and their constituent monomers are reported in Table 3. The
relative stabilities of CUR/CDs inclusion complexes were measured by evaluating their ꢀE,
which can be addressed as the stabilizing energy of complexation. Hence, the low ꢀE values,
the more stable the inclusion complex. The energies of all the complexes were lower than the
energies of their two constituents except for CUR/HP-γ-CD, indicating that the association of
CUR and CDs had formed mainly quite stable complexes. Among the four inclusion
complexes, the more hydrated CUR/γ-CD complex was found with the lowest energy term
(ꢀE = ‒191.55 kcal*mol-1) due to the numerous stabilizing CD intramolecular H-bonds and
less excessive water displacement effect.
Additionally, in order to study the effect of the CDs on the CUR conformational change that
can be adapted into the cavity, as well as their influence on the binding property, the CUR/CD
complexes were subjected to the MC simulations implemented in the ROSETTA code (Meiler
et al., 2006), and 100 ns MD simulations with the MM-GBSA approach in evaluating binding
affinity. In classical MD simulations, the conformational changes of compounds and proteins
are connected in time. These simulations mimic the dynamical behavior of the studied
systems, from which time-dependent values of conformational and thermodynamic properties
can be estimated. Series of atomic coordinates (trajectories) are output in the result by using
Newton’s equations of motion. However, in MC simulations, each conformation can be
determined only on its predecessor. The MC protocol predicts conformations randomly and
employs energetic criteria to detect whether or not to confirm the new conformation.
10

ACCEPTED MANUSCRIPT
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
To approximate the MD simulations in a more or less realistic way, all the hydroxyl groups
were substituted with hydroxypropyl residues in the case of modified cyclodextrin models.
Since the biophysical and biological behaviors of a compound often depend mainly upon the
conformations that it can adopt, the time-dependent configurations (translations and rotations)
of the CUR into the γ-CD and HP-γ-CD were studied and clearly monitored (Supplementary
material 1 and 2) based on the MD trajectory frames, in which the CD heavy atoms were
aligned with respect to the starting structure. These analyses are feasible to illustrate the level
of the conformational changes and movements of the CUR into the different environments
from the starting point. The efficiency of this fashion can be enhanced by clustering the
different configurations of the CUR that group together the similar conformations, from
which the representative conformers can be selected. The relatively straightforward clustering
method is the RMSD parameter between pairs of conformations (sometimes referred to as
distance analysis). The CUR spatial arrangements were set establishing non-bonded forces
with the CDs atoms and the water molecules into the cavity. The strong binding of CUR into
the HP-γ-CD provides a high level of shape complementary with the hydrophobic pocket of
the HP-γ-CD.
The relative structural stability of the CUR in the HP-γ-CD cavity is dependent upon internal
and non-covalent forces between the modified CD and CUR atoms. The CUR molecule in the
HP-γ-CD complex incurs a higher entropic penalty compared to CUR in the γ-CD complex.
The free energy difference is considered to be a highly important factor in the thermodynamic
concept and it would appear to be necessary for estimating the average ꢀGbind values between
the CUR and CDs based on the MC and MD trajectory frames. The energetic analysis reveals
that the highest CD binding affinity to CUR belongs to HP-γ-CD after ligand hydration with
the ꢀGMM-GBSA value of −86.38 kcal*mol-1 and the ꢀGbind value of −71.4 kcal*mol-1. This was
due to strong hydrophobic (vdW) forces causing a decrease in the CUR hydration followed by
excessive desolvation effects increasing water displacement (Figure 5 and 6 [A-D]). On the
other hand, the hydration/desolvation effect of CUR/γ-CD characterized by less water
displacement and high ligand hydration was more pronounced. However, the hydrophobic
(vdW) forces in CUR/γ-CD were less prominent.
Furthermore, the relative free energies of binding CUR to the hydrated and dehydrated CDs
by the MD simulations were estimated to be −69.67 kcal*mol-1 and −46.44 kcal*mol-1,
respectively. The same systems (CUR bound to hydrated/dehydrated γ-CD and HP-γ-CD)
were examined by the MC simulations, providing free energy differences of −51.90 kcal*mol-
11

ACCEPTED MANUSCRIPT
1
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
and −49.41 kcal*mol-1, respectively. Both simulation methods were able to accurately
determine the effects of the combinatorial explicit and implicit solvent models of the cavity
that were being used in the energy calculations. The stability of the CUR was a consequence
of the hydrophobic effect in the pocket of the CDs, from which the contribution of the
entropic term could be considered. The vdW forces between the CUR, γ-CD and HP-γ-CD
systems were calculated to be −27.59 kcal*mol-1 and −40.57 kcal*mol-1, respectively. The
vdW interactions of CUR/ γ-CD and HP-γ-CD in the presence of explicit water molecules
showed the average energies to be −24.37 kcal*mol-1 and −37.03 kcal*mol-1. The observed
differences in the non-covalent energies for the explicitly hydrated and dehydrated pockets of
the CDs were expected due to the water-mediated Coulombic forces. These results obtained
through the MD simulations, obviously were in a satisfactory qualitative agreement with those
calculated through the MC method. However, the relative free energy value was highly
sensitive to the configurations of the CD complexes used for the above calculations. This was
because, the conformations of the molecules were obtained by different simulation methods
(MC and MD) in each case. This was considered to be a major drawback of the force-field
methods. In principle, free energy is a very difficult value to estimate for flexible systems,
such as the CDs used in the current study, which share many closely separated minima.
Related thermodynamic terms, such as entropy and partial molar free energy, were also
difficult to estimate, due to the poor sampling produced by these simulations.
4. Conclusions
Molecular docking, MC, and MD simulations were performed in this study to investigate the
CUR hydration/desolvation effects for its complexation with γ-CD and HP-γ-CD. The
calculations indicate that water molecules play an important role in host-guest complexation,
mediating the binding of ligands with their targets via hydrogen bond formations. The CUR
hydration increases its binding affinity to both CDs, which is confirmed by the ꢀGbind and
ꢀGMM-GBSA values. Although the CUR/CD complex affinities were significantly improved
after the CUR hydration, as newly formed hydrogen bonds are responsible for the stability of
the inclusion complexes, hydrophobic (vdW) forces in CUR/HP-γ-CD are dominant over
hydration/desolvation effects. In addition, the CUR/HP-γ-CD complex was found to be more
stable in aqueous solution with the highest apparent stability constant for CUR/HP-γ-CD (Kc
= 1.58*104 M-1) as the more soluble form in distilled water. Overall, the results of this study
12

ACCEPTED MANUSCRIPT
can aid drug delivery vector development, underscoring the importance of hydration effect on
the formulation of different drug-like molecules.
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
Acknowledgments
Special thanks are extended to Dr. Biswapriya B. Misra from the Department of Biology,
Genetics Institute, University of Florida for his assistance in the paper's writing. The authors
are also grateful to the BMBF (Bundesministerium für Bildung und Forschung) for supporting
this work through a grant (BMBF13N11801) to Jens Broscheit.
Conflicts of Interest
The authors declare no conflict of interest.
Reference
1. Volate, S. R., D. M. Davenport, et al. (2005). "Modulation of aberrant crypt foci and
apoptosis by dietary herbal supplements (quercetin, curcumin, silymarin, ginseng and
rutin)." Carcinogenesis 26(8): 1450-1456.
2. Baumann, L., H. Woolery-Lloyd, et al. (2009). ""Natural" ingredients in cosmetic
dermatology." J Drugs Dermatol 8(6 Suppl): s5-9.
3. Nagabhushan, M. and S. V. Bhide (1992). "Curcumin as an inhibitor of cancer." J Am
Coll Nutr 11(2): 192-198.
4. Prasad, S. and A. K. Tyagi (2015). "Curcumin and its analogues: a potential natural
compound against HIV infection and AIDS." Food Funct 6(11): 3412-3419.
5. Tonnesen, H. H., M. Masson, et al. (2002). "Studies of curcumin and curcuminoids.
XXVII. Cyclodextrin complexation: solubility, chemical and photochemical stability."
Int J Pharm 244(1-2): 127-135.
6. Yadav, V. R., S. Suresh, et al. (2009). "Effect of cyclodextrin complexation of curcumin
on its solubility and antiangiogenic and anti-inflammatory activity in rat colitis model."
AAPS PharmSciTech 10(3): 752-762.
7. Marcolino, V. A., G. M. Zanin, et al. (2011). "Interaction of curcumin and bixin with
beta-cyclodextrin: complexation methods, stability, and applications in food." J Agric
Food Chem 59(7): 3348-3357.
13

ACCEPTED MANUSCRIPT
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
8. Zidovetzki, R. and I. Levitan (2007). "Use of cyclodextrins to manipulate plasma
membrane cholesterol content: evidence, misconceptions and control strategies."
Biochim Biophys Acta 1768(6): 1311-1324.
9. Baldwin, R. L. (2010). "Desolvation penalty for burying hydrogen-bonded peptide
groups in protein folding." J Phys Chem B 114(49): 16223-16227.
10. Forli, S. and A. J. Olson (2012). "A force field with discrete displaceable waters and
desolvation entropy for hydrated ligand docking." J Med Chem 55(2): 623-638.
11. Higuchi T, Connors KA. Phase solubility techniques. In: Reilly CN, editor. Advances in
Analytical Chemistry Instrumentation. Vol. 4. New York, NY: Interscience; 1965. pp.
117–212.
12. Pinotsis, N., D. D. Leonidas, et al. (2003). "The binding of beta- and gamma-
cyclodextrins to glycogen phosphorylase b: kinetic and crystallographic studies."
Protein Sci 12(9): 1914-1924.
13. Bachega, J. F., L. F. Timmers, et al. (2013). "GTKDynamo: a PyMOL plug-in for
QC/MM hybrid potential simulations." J Comput Chem 34(25): 2190-2196.
14. Gasteiger, J. and M. Marsili (1980). "Iterative Partial Equalization of Orbital
Electronegativity - a Rapid Access to Atomic Charges." Tetrahedron 36(22): 3219-
3228.
15. Goodsell, D. S., G. M. Morris, et al. (1996). "Automated docking of flexible ligands:
applications of AutoDock." J Mol Recognit 9(1): 1-5.
16. Forli, S. (2015). "Charting a Path to Success in Virtual Screening." Molecules 20(10):
18732-18758.
17. Seeliger, D. and B. L. de Groot (2010). "Ligand docking and binding site analysis with
PyMOL and Autodock/Vina." J Comput Aided Mol Des 24(5): 417-422.
18. Dewar, M. J. S., E. G. Zoebisch, et al. (1985). "The Development and Use of Quantum-
Mechanical Molecular-Models .76. Am1 - a New General-Purpose Quantum-
Mechanical Molecular-Model." J Am Chem Soc 107(13): 3902-3909.
19. Rocha, G. B., R. O. Freire, et al. (2006). "RM1: A reparameterization of AM1 for H, C,
N, O, P, S, F, Cl, Br, and I." Journal of Computational Chemistry 27(10): 1101-1111.
20. Bowers, K. J., R. O. Dror, et al. (2006). "The midpoint method for parallelization of
particle simulations." Journal of Chemical Physics 124(18).
21. Banks, J. L., H. S. Beard, et al. (2005). "Integrated modeling program, applied chemical
theory (IMPACT)." J Comput Chem 26(16): 1752-1780.
14

ACCEPTED MANUSCRIPT
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
22. Essmann, U., L. Perera, et al. (1995). "A Smooth Particle Mesh Ewald Method." Journal
of Chemical Physics 103(19): 8577-8593.
23. Hoover, W. G. (1985). "Canonical Dynamics - Equilibrium Phase-Space Distributions."
Physical Review A 31(3): 1695-1697.
24. Martyna, G. J., D. J. Tobias, et al. (1994). "Constant-Pressure Molecular-Dynamics
Algorithms." Journal of Chemical Physics 101(5): 4177-4189.
25. Meiler, J. and D. Baker (2006). "ROSETTALIGAND: protein-small molecule docking
with full side-chain flexibility." Proteins 65(3): 538-548.
26. Gohlke, H. and D. A. Case (2004). "Converging free energy estimates: MM-PB(GB)SA
studies on the protein-protein complex Ras-Raf." J Comput Chem 25(2): 238-250.
27. Li, J., R. Abel, et al. (2011). "The VSGB 2.0 model: a next generation energy model for
high resolution protein structure modeling." Proteins 79(10): 2794-2812.
28. Ansari, M. Y.; Dikhit, M. R.; Sahoo, G. C.; Das, P., Comparative modeling of HGPRT
enzyme of L. donovani and binding affinities of different analogs of GMP.
International journal of biological macromolecules 2012, 50 (3), 637-49.
29. Kar, R. K.; Ansari, M. Y.; Suryadevara, P.; Sahoo, B. R.; Sahoo, G. C.; Dikhit, M. R.;
Das, P., Computational elucidation of structural basis for ligand binding with
Leishmania donovani adenosine kinase. BioMed research international 2013, 2013,
609289.
30. Ahsan, M. J.; Choudhary, K.; Jadav, S. S.; Yasmin, S.; Ansari, M. Y.; Sreenivasulu, R.,
Synthesis, antiproliferative activity, and molecular docking studies of curcumin
analogues bearing pyrazole ring. Med Chem Res 2015, 24 (12), 4166-4180.
31. Shityakov, S., R.E. Salmas, et al, (2016) “Characterization, In Vivo Evaluation and
Molecular Modeling of Different Propofol-Cyclodextrin Complexes to Assess Their
Drug Delivery Potential at The Blood-Brain Barrier Level” J Chem Inf Model 56(10):
1914-1922
32. Manolova, Y., V. Deneva, et al. (2014). "The effect of the water on the curcumin
tautomerism: a quantitative approach." Spectrochim Acta A Mol Biomol Spectrosc 132:
815-820.
33. Tabushi, I., Y. I. Kiyosuke, et al. (1978). "Approach to Aspects of Driving Force of
Inclusion by Alpha-Cyclodextrin." J Am Chem Soc 100(3): 916-919.
34. Tanhuanpaa, K., K. H. Cheng, et al. (2001). "Characteristics of pyrene
phospholipid/gamma-cyclodextrin complex." Biophysical Journal 81(3): 1501-1510.
15

ACCEPTED MANUSCRIPT
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
35. Munro, I. C., P. M. Newberne, et al. (2004). "Safety assessment of gamma-
cyclodextrin." Regul Toxicol Pharmacol 39 Suppl 1: S3-13.
36. Szejtli, J. (1998). "Introduction and General Overview of Cyclodextrin Chemistry."
Chemical Reviews 98(5): 1743-1754.
37. Uekama, K., F. Hirayama, et al. (1998). "Cyclodextrin Drug Carrier Systems."
Chemical Reviews 98(5): 2045-2076.
38. Ma, X. Y., Z. X. Liao, et al. (2000). "Study of the podophyllotoxin/beta-cyclodextrin
inclusion complex." Journal of Inclusion Phenomena and Macrocyclic Chemistry 36(3):
335-342.
39. Zhao, D. Y., K. J. Liao, et al. (2002). "Study of the supramolecular inclusion of beta-
cyclodextrin with andrographolide." Journal of Inclusion Phenomena and Macrocyclic
Chemistry 43(3-4): 259-264.
40. Shityakov, S., J. Broscheit, et al. (2012). "alpha-Cyclodextrin dimer complexes of
dopamine and levodopa derivatives to assess drug delivery to the central nervous
system: ADME and molecular docking studies." Int J Nanomedicine 7: 3211-3219.
41. Shityakov, S. and C. Forster (2014). "In silico structure-based screening of versatile P-
glycoprotein inhibitors using polynomial empirical scoring functions." Adv Appl
Bioinform Chem 7: 1-9.
42. Pawar, Y. B., B. Munjal, et al. (2012). "Bioavailability of a lipidic formulation of
curcumin in healthy human volunteers." Pharmaceutics 4(4): 517-530.
43. Furuki, T., F. Hosokawa, et al. (1993). "Microscopic Medium Effects on a Chemical-
Reaction - a Theoretical-Study of Decarboxylation Catalyzed by Cyclodextrins as an
Enzyme Model." J Am Chem Soc 115(7): 2903-2911.
44. Hegge, A. B., M. Masson, et al. (2009). "Investigation of curcumin-cyclodextrin
inclusion complexation in aqueous solutions containing various alcoholic co-solvents
and alginates using an UV-VIS titration method. Studies of curcumin and
curcuminoides, XXXV." Pharmazie 64(6): 382-389.
507
508
509
510
16

ACCEPTED MANUSCRIPT
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
B
R
O
R
R
O
O
O
O
O
O
O
R
R
O
R
O
O
O
R
O
O
O
R
R
R
O
O
A
O
O
O
O
R
HO
OH
R
O
R
R
O
O
O
R
R
O
O
R
O
R
CH3
O
OH
CH3
R
O
O
O
R
O
O
O
O
R
O
O
O
R
O
O
R
O
O
R
R = H or CH2CH(OH)CH3
Figure 1. 2D chemical structure of CUR in enol form (A) and γ-CD (R = H) or HP-γ-CD (R =
CH2CH(OH)CH3) molecules (B)
17

ACCEPTED MANUSCRIPT
544
B
545A
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
Figure 2: Standard (A) and hydrated (B) AutoDock molecular docking profiles of CUR (enol
form) bound to the γ-CD molecule. Cyclodextrin and its ligand are visualized with lines and
sticks, respectively. Water molecules are represented in spheres, and the hydrogen bonds are
depicted as yellow dashed lines. Water molecule located in the γ-CD binding cavity is pointed
by the bold arrow. Molecules are colored according to their atom types. The majority of
hydrogens atoms are omitted for clarity.
18

ACCEPTED MANUSCRIPT
577
578
B
A
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
Figure 3: Standard (A) and hydrated (B) AutoDock molecular docking profiles of CUR (enol
form) bound to the HP-γ-CD molecule. Cyclodextrin and its ligand are visualized with lines
and sticks, respectively. Water molecules are represented in spheres, and the hydrogen bonds
are depicted as yellow dashed lines. Water molecule located in the γ-CD binding cavity is
pointed by the bold arrow. Molecules are colored according to their atom types. The majority
of hydrogens atoms are omitted for clarity.
19

ACCEPTED MANUSCRIPT
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
A
B
0.010
0.6
Cauchy-Lorentz
Cauchy-Lorentz
Cubic spline
0.008
0.006
0.004
0.002
0.000
Cubic spline
0.4
0.2
0.0
0
4
8
12
16
0
4
8
12
16
γ-CD (mM)
HP-γ-CD (mM)
Figure 4. Solubility isotherms of the CUR/γ-CD (A) and CUR/HP-γ-CD (B) complexes
determined in H2O, pH 7.0. The S0 value (the same for both complexes) states for the CUR
saturation concentration in the solvent in absence of cyclodextrin. The solid lines show fitting
to the isotherms. The data represent means ± S.D. of three independent experiments.
Figure 5: The ꢀGMM-GBSA values for CUR/γ-CD and CUR/HP-γ-CD with or without CUR
hydration (denoted by *) calculated by the MC approach.
20

ACCEPTED MANUSCRIPT
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
Mean= −36.61 ± 4.18
Max= −25.61
Mean= −61.0 ± 5.69
Max= −48.35
Min= −46.44
Max= −27.68
Min= −76.66
Max= −51.11
Mean= −47.21 ± 9.51
Mean= −71.40 ± 9.87
Min= −69.67
Min= −94.25
Figure 6: The ꢀGbind values (in kcal*mol-1) calculated from 100 ns MD simulations for
CUR/γ-CD and CUR/γ-CD* (A, B) together with CUR/HP-γ-CD and CUR/HP-γ-CD* (C, D).
CUR hydration is denoted by an asterisk sign. The data represent means ± S.D.
21

ACCEPTED MANUSCRIPT
675
676
Table 1: Summary of the molecular docking profiles for CUR bound to the γ-CD and HP-γ-
CD molecules.
***
Complex
Nrun Ncl
ꢀGbind
(kcal*mol-1) (µM)
Kd
LE**
RMSDLC Nats
(nm)
Ntors
CUR/γ-CD
CUR/γ-CD*
100 66
100 72
100 51
‒6.83
‒7.69†
‒7.92
‒9.93†
9.45
2.2
‒0.23
‒0.17
‒0.26
‒0.22
78.02
78.18
77.64
77.25
30
45
30
45
10
10
10
10
CUR/HP-γ-CD
CUR/HP-γ-CD* 100 65
1.49
0.05
∆Gbind
**
†
677
678
*-hydrated ligand;
-
LE =
;
***-waters are described by monatomic pseudoatoms; -
Nats
∆Gbind = ∆Glig + ∆Gwat ; Nruns-total number of runs; Ncl-number of distinct clusters formed in
679
680
681
682
683
684
the clustering; RMSDLC-the root-mean-square-deviation difference between the lowest energy
conformation (LC) in the largest cluster and the reference ligand conformation; Nats-number
of heavy atoms; Ntors-number of torsions
Table 2: Summary of displaced (DISP), weakly (WK) and strongly (STR) bound water
molecules during hydrated complexation of CUR with the γ-CD and HP-γ-CD molecules.
Complex
DISP WK STR
CUR/γ-CD
CUR/HP-γ-CD
12
13
2
1
1
1
685
686
687
688
Table 3: The AM1 energies (in kcal*mol-1) of the 7 species, the energy differences (ꢀE)
between the inclusion complexes and the molecules (CUR, γ-CD and HP-γ-CD), which form
the complexes (CUR/ γ-CD and CUR/HP-γ-CD).
Species
AM1 energy ꢀE**
CUR
‒151.19
-
γ-CD
‒1891.39
‒3057.42
‒2045.59
‒2234.13
‒3199.91
-
HP-γ-CD
CUR/γ-CD
CUR/γ-CD*
CUR/HP-γ-CD
-
‒3.01
‒191.55
8.7
CUR/HP-γ-CD* ‒3324.52
‒115.91
689
*-hydrated ligand; **- ꢀE = EComplex ‒ ECD ‒ ECUR
22

ACCEPTED MANUSCRIPT
•
•
•
•
•
Curcumin solubility profiles and hydration/desolvation effects of this substance
formulated with γ-cyclodextrin and hydroxypropyl-γ-cyclodextrin were investigated.
Curcumin/hydroxypropyl-γ-cyclodextrin complex was found to be more stable in solution
than curcumin/hydroxypropyl-γ-cyclodextrin.
Water molecules play an important role in host-guest complexation mediating the
curcumin binding to cyclodextrins via hydrogen bond formations.
Curcumin hydration/desolvation effects contributed to the complex formation by
elevating the curcumin binding affinity to cyclodextrins.
Curcumin/hydroxypropyl-γ-cyclodextrin complex was determined with a minimal free
energy of binding due to hydrophobic forces.