- Email: [email protected]
Effect of antioxidant activity of caffeic acid with cyclodextrins using ground mixture method
Accepted Manuscript Title: Effect of antioxidant activity of caffeic acid with cyclodextrins using ground mixture method Author: Ryota Shiozawa, Yutaka Inoue, Isamu Murata, Ikuo Kanamoto PII: DOI: Reference:
S1818-0876(17)30352-5 https://doi.org/doi:10.1016/j.ajps.2017.08.006 AJPS 462
To appear in:
Asian Journal of Pharmaceutical Sciences
Received date: Revised date: Accepted date:
8-5-2017 8-8-2017 15-8-2017
Please cite this article as: Ryota Shiozawa, Yutaka Inoue, Isamu Murata, Ikuo Kanamoto, Effect of antioxidant activity of caffeic acid with cyclodextrins using ground mixture method, Asian Journal of Pharmaceutical Sciences (2017), https://doi.org/doi:10.1016/j.ajps.2017.08.006. 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.
1
Title page
2
Effect of antioxidant activity of caffeic acid with cyclodextrins
3
using ground mixture method
4 5
Ryota Shiozawaa, Yutaka Inouea*, Isamu Murataa, Ikuo Kanamotoa
6 7
a
Faculty of Pharmacy and Pharmaceutical Sciences, Josai University, 1-1 Keyakidai, Sakado-shi, Saitama, 3500295, Japan
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Corresponding author: Yutaka Inouea*
24
Mailing address: Laboratory of Drug Safety Management, Faculty of Pharmacy and
25
Pharmaceutical Sciences, Josai University, 1-1 Keyakidai, Sakado-shi, Saitama, 350-0295,
26
Japan.
27
Tel.: +81-49-271-7317; Fax: +81-49-271-7317
28
E-mail: [email protected]
29
1 Page 1 of 30
1
Graphical Abstract
2
Caffeic acid/α-cyclodextrin (molar ratio=1/1), the vinylene group of the caffeic acid molecule
3
appears to be included from the wider to the narrower rim of the ring of α-cyclodextrin.
4
Caffeic acid /β-cyclodextrin (molar ratio=1/1), the aromatic ring of the caffeic acid molecule
5
appears to be included from the wider to the narrower rim of the ring of βCD. This suggests
6
that forms of inclusion differed in the caffeic acid /α-cyclodextrin (molar ratio=1/1) and the
7
caffeic acid /β-cyclodextrin (molar ratio=1/1).
8 9 10 11 12 13 Caffeic acid/β-cyclodextrin
Caffeic acid/α-cyclodextrin
Structural images of caffeic acid/cyclodextrin complex from side view.
14
Abstract:
15 16
In the current study, we prepared a ground mixture (GM) of caffeic acid (CA) with
17
α-cyclodextrin (αCD) and with β-cyclodextrin (βCD), and then comparatively assessed the
18
physicochemical properties and antioxidant capacities of these GMs. Phase solubility
19
diagrams indicated that both CA/αCD and CA/βCD formed a complex at a molar ratio of 1/1.
20
In addition, stability constants suggested that CA was more stable inside the cavity of αCD
21
than inside the cavity of βCD. Results of powder X-ray diffraction (PXRD) indicated that the
22
characteristic diffraction peaks of CA and CD disappeared and a halo pattern was produced
23
by the GMs of CA/αCD and CA/βCD (molar ratios=1/1). Dissolution testing revealed that
24
both GMs had a higher rate of dissolution than CA alone did. Based on the 1H-1H NOESY 2 Page 2 of 30
1
NMR spectra for the GM of CA/αCD, the vinylene group of the CA molecule appeared to be
2
included from the wider to the narrower rim of the αCD ring. Based on spectra for the GM of
3
CA/βCD, the aromatic ring of the CA molecule appeared to be included from the wider to the
4
narrower rim of the βCD ring. This suggests that the structures of the CA inclusion
5
complexes differed between those involving αCD rings and those involving βCD rings.
6
Results of a DPPH radical-scavenging activity test indicated that the GM of CA/αCD had a
7
higher antioxidant capacity than that of the GM of CA/βCD. The differences in the
8
antioxidant capacities of the GMs of CA/αCD and CA/βCD are presumably due to
9
differences in stability constants and structures of the inclusion complexes.
10 11
Keywords: Caffeic acid; Cyclodextrins; Inclusion; Antioxidant activity; Solubility; Ground
12
mixture
13
3 Page 3 of 30
1
1. Introduction
2 3
In recent years, cardiovascular diseases have become the second leading cause of death in
4
Japan, after cancer [1]. Therefore, several health foods are being investigated to treat the risk
5
factors of cardiovascular diseases, which include dyslipidemia- and diabetes-induced
6
atherosclerosis. Chlorogenic acid, a coffee polyphenol, is one such health food that has been
7
reported to protect against lifestyle diseases, such as dyslipidemia and diabetes [2-4]. Caffeic
8
acid (CA) is a decomposition product of chlorogenic acid that has the same physiological
9
activity as chlorogenic acid does. Although CA has been reported to protect against
10
dyslipidemia and blood glucose elevation owing to its antioxidant activity, its application is
11
limited by poor water solubility and oral absorption [5, 6]. Cyclodextrin (CD) is a
12
ring-shaped molecule consisting of 6, 7, or 8 glucose units for α, β, or γCD, respectively,
13
which are linked by α-(1→4)-glucosidic bonds. In the safety assessment of the Food and
14
Agriculture Organization of the United Nations/World Health Organization (FAO/WHO)
15
Joint Expert Committee on Food Additives (JECFA), the acceptable daily intake (ADI) was
16
set as ‘‘not specified’’ for α- and γCD and ‘‘up to 5 mg/kg/day’’ for βCD. CD molecules
17
have a hydrophobic cavity in their center that can incorporate various guest molecules to
18
form inclusion complexes [7]. Inclusion within CD is known to stabilize pharmaceuticals [8]
19
and improve their bioavailability by improving their solubility [9]. Various methods, such as
20
coprecipitation [10], kneading [11], and lyophilization [12], have been reported for the
21
preparation of inclusion complexes. Co-grinding, a solvent-free, mechanochemical method of
22
forming inclusion complexes, causes changes in physicochemical properties because of the
23
application of mechanical energy (i.e., by the grinding of the solid material), and has been
24
used for the complexation of piperine, which is poorly water soluble [13,14]. The preparation
25
of (R)-α-lipoic acid and CD inclusion complexes, which is affected by the cavity size of CD,
26
has also been reported [15]. As the steric properties and antioxidant capacity of drugs are 4 Page 4 of 30
1
reported to be affected by the type of CD [16], it is useful to evaluate the physicochemical
2
properties of inclusion complexes prepared using CDs of different cavity sizes. In addition,
3
CDs have been reported to increase antioxidant capacity by their inclusion of polyphenols,
4
such as chlorogenic acid [17]. However, there is no report of increased antioxidant capacity
5
of CA using CDs. Fundamental research to evaluate the antioxidant capacity of inclusion
6
complexes prepared using CDs of different cavity sizes to improve the antioxidant capacity
7
of CA may have clinical applications. We have already reported the preparation of CA and
8
γCD inclusion complexes by co-grinding and demonstrated that the antioxidant capacity is
9
retained owing to improvement in CA inclusion and elution [18]. Since αCD and βCD have a
10
narrower cavity than γCD does, they may form different inclusion complexes and thus affect
11
the antioxidant capacity of drugs differently. Therefore, in this study, we prepared CA/CD
12
inclusion complexes using αCD and βCD, and evaluated their physicochemical properties,
13
dissolution properties, and antioxidant capacities.
14
5 Page 5 of 30
1
2. Materials and methods
2 3
2.1. Materials
4
CA (Fig. 1A) was purchased from Wako Pure Chemical Industries Co. Ltd (Japan). αCD
5
and βCD (Fig. 1B) were a generous gift from CycloChem Co. Ltd. (Tokyo, Japan) and were
6
stored
7
2,2-Diphenyl-1-picrylhydrazyl (DPPH) was purchased from Sigma-Aldrich Co. Ltd. All other
8
chemicals and solvents were of analytical grade and were purchased from Wako Pure
9
Chemical Industries Co. Ltd.
at
40
°C
and
82%
relative
humidity
for
7
days,
prior
to
use.
(Fig. 1)
10 11 12
2.2. Preparation of the physical mixture (PM) and ground mixture (GM)
13
The PM was prepared by mixing CA and CD in different molar ratios (2/1, 1/1, and 1/2)
14
using a vortex mixer for 1 min. The GM was prepared by placing the PM (1 g) in an alumina
15
cell and grinding for 60 min using a vibration rod mill (TI-500ET, CMT Co. Ltd., Japan).
16 17
2.3. Methods
18
2.3.1. Phase solubility studies
19
Phase solubility studies were performed as previously described by Higuchi and Connors
20
[19]. A supersaturated amount of CA (100 mg) was added to an aqueous solution (10 ml) of
21
αCD (0-40 mM) or βCD (0-15 mM), and the mixture was placed in a medium-sized, shaking
22
instrument for 24 h at 200 rpm and 25±0.5 °C. After the suspension reached equilibrium, the
23
solution was immediately filtered through a 0.45 μm membrane filter. The solubility was
24
quantified by high-performance liquid chromatography using a Waters® e2795 separations
25
module (Nippon Waters Co., Ltd., Japan) and Cosmosil 5C18-AR-II packed column (4.6
26
mm×150 mm) at a detection wavelength of 285 nm. The sample injection volume was 30 μl, 6 Page 6 of 30
1
and the column temperature was 40 °C. The mobile phase consisted of distilled
2
water/acetonitrile/methanol/acetic acid (862/113/20/5), and the CA retention time was 10 min.
3
The apparent stability constant (Ks) of the CA/CD inclusion complexes was calculated from
4
the slope of the phase solubility diagram and solubility of CA in the absence of CD (S 0) using
5
equation (1).
6
7
Ks
slope
(1)
S 0 (1 slope )
8 9
2.3.2. Differential scanning calorimetry (DSC)
10
The thermal behavior of samples was recorded using a differential scanning calorimeter
11
(Thermo plus EVO, Rigaku Co., Japan). All samples were weighed (2 mg) and heated at a
12
scanning rate of 5.0 °C/min under nitrogen flow (60 ml/min). Aluminum pans and lids were
13
used for all samples.
14 15
2.3.3. Powder X-ray diffraction (PXRD)
16
PXRD studies were performed using a powder X-ray diffractometer (MiniFlex II, Rigaku
17
Co., Japan) with Cu Kα radiation (30 kV, 15 mA). The diffraction intensity was measured
18
using a NaI scintillation counter. The samples were scanned from 2θ = 3–35° at a scan rate of
19
4°/min. The powder sample was spread on a glass plate to maintain a flat sample plane, and
20
measurements were performed.
21 22
2.3.4. Fourier transform infrared (FTIR) spectroscopy
23
The FTIR spectra of samples were recorded using an FTIR spectrometer (FT/IR-410,
24
JASCO Inc., Japan) by the potassium bromide (KBr) disk method. The samples were scanned
25
32 times over a range of 650-4,000 cm-1 with a resolution of 4 cm-1. The tablets were 7 Page 7 of 30
1
prepared by adding KBr to the sample (sample/KBr = 1/10 (w/w)) and manually compressing
2
the mixture. Background correction was performed using a blank KBr tablet.
3 4
2.3.5. Measurement of 1H-1H nuclear Overhauser effect spectroscopy (NOESY) NMR spectra
5
The NOESY spectra were recorded on a 700 MHz Varian NMR spectrometer (Agilent
6
Technologies) under the following conditions: solvent: D2O; resonant frequency: 699.6 MHz;
7
pulse width: 90°; relaxation delay: 0.500 s; scan time: 0.500 s; temperature: 25 °C; 256
8
increments.
9 10
2.3.6. Dissolution profile
11
Dissolution testing of samples was performed using a dissolution apparatus (NTR-593,
12
Toyama Sangyo Co. Ltd., Japan) with 900 ml of distilled water at 37±0.5 °C and 50 rpm,
13
according to the paddle method (Japanese Pharmacopoeia 17 th Edition). CA (100 mg) was
14
weighed accurately and loaded into the paddle apparatus. Dissolved samples (10 ml) were
15
collected at 5, 10, 15, 30, 60, 90, and 120 min through 0.45 μm membrane filters, with
16
replacement of an equal volume of fresh dissolution medium at the same temperature, to
17
maintain a constant volume of dissolution medium. The concentration of CA in the samples
18
was determined by the same method as described for the phase solubility studies (section
19
2.3.1).
20 21
2.3.7. DPPH radical scavenging activity
22
The DPPH radical scavenging activity of the samples was measured using a microplate
23
reader (SpectraMax 190, Molecular Devices, Japan). Briefly, each sample and DPPH
24
methanolic solution (100 μM) were mixed in a microplate at a ratio of 1/1 (v/v) and incubated
25
at 25 °C for 5 min. The absorbance of DPPH in samples (As) was measured at 517 nm. The
26
absorbance of a mixture of methanol/water (1/1 (v/v)) (Br), which has a radical removal rate 8 Page 8 of 30
1
of 100%, and a mixture of DPPH/water (1/1 (v/v)) (A0), which has a radical removal rate of
2
0%, was also measured. The radical scavenging rate was calculated using equation (2) [17].
3 4
Radical scavenging rate = [1-(As-Br)/(A0-Br)] ×100
(2)
5 6 7
3. Results and discussion
8 9
3.1. Phase solubility studies
10 11
Phase solubility studies were conducted to assess the molar ratio of the CA/CD inclusion
12
complexes and their stability constants in an aqueous solution. Results of the solubility
13
testing indicated that the solubility of CA increased linearly for αCD and βCD, producing an
14
AL type of phase solubility diagram according to the classification system of Higuchi et al.
15
(Fig. 2). Typically, an AL diagram indicates that a complex is formed at a molar ratio of 1/1
16
[20], so both CA/αCD and CA/βCD formed a complex at a molar ratio of 1/1 in solution. The
17
stability constant (Ks) was determined next. Ks was 1547.5 M-1 for αCD and 371.4 M-1 for
18
βCD (Table 1). A study by Iacovino et al. reported that, when a complex has a stability
19
constant exceeding 1000 M-1, the guest molecule and CD are unlikely to dissociate [21]. CA
20
included in αCD had a stability constant of 1547.5 M-1, so CA was more stable inside the
21
cavity of αCD than it was inside the cavity of βCD.
22
(Fig2) (Table 1)
23 24
3.2. Differential scanning calorimetry (DSC)
25 26
Phase solubility diagrams suggested that both CA/αCD and CA/βCD formed a complex at
27
a molar ratio of 1/1 in an aqueous solution. When CA/αCD and CA/βCD are ground together, 9 Page 9 of 30
1
they presumably still form an inclusion complex at a molar ratio of 1/1. It has been reported
2
that when a guest molecule does not melt or its peak shifts as a result of inclusion complex
3
formation caused by co-grinding, changes in its thermal behavior become evident [22]. Thus,
4
DSC was performed in the current study to examine the thermal behavior of the GMs. Results
5
of DSC revealed an endothermic peak due to melting at 223ºC for intact CA (Fig. 3A). The
6
PM of CA/αCD (molar ratio=1/1) had an endothermic peak due to the melting of CA at
7
210ºC, and the PM of CA/βCD (molar ratio=1/1) had an endothermic peak due to the melting
8
of CA at 216ºC, so CA crystals were presumably present (Fig. 3B and F). The GM of
9
CA/αCD (molar ratio=2/1) had an endothermic peak due to CA at 207ºC (Fig. 3C). The
10
endothermic peak due to CA disappeared for the GM of CA/αCD at molar ratios of 1/1 and
11
1/2 (Fig. 3D and E). The GM of CA/βCD (molar ratio=2/1) had an endothermic peak due to
12
CA at 197ºC (Fig. 3G). The endothermic peak due to CA disappeared for the GM of CA/αCD
13
at molar ratios of 1/1 and 1/2 (Fig. 3H and I). In our previous study, the disappearance of the
14
CA melting peak was confirmed by DSC measurement for the GM of the CA/γCD complex
15
[18]. Therefore, the GMs of CA/αCD and CA/βCD are also considered inclusion complexes.
16
The endothermic peak due to CA disappeared because the CA and αCD molecules and the
17
CA and βCD molecules formed inclusion complexes. The GM of CA/αCD (molar ratio=2/1)
18
and the GM of CA/βCD (molar ratio=2/1) showed an endothermic peak due to CA. The
19
reason may be that there was excess CA that did not complex with the CDs. This suggests
20
that both CA/αCD and CA/βCD were present in the GMs at a molar ratio of CA/CD of 1/1. (Fig. 3)
21 22 23
3.3. Powder X-ray diffraction (PXRD)
24 25
The results of DSC suggested that CA could form an inclusion complex with either αCD or
26
βCD at a stoichiometric ratio of 1/1. Thus, PXRD was performed to examine the crystalline 10 Page 10 of 30
1
state of the CA/CDs in the GMs. Intact CA and ground CA had a characteristic diffraction
2
peak at 2θ=26.9º, αCD had a characteristic diffraction peak at 2θ=11.8º, and βCD had a
3
characteristic diffraction peak at 2θ=12.3º (Fig. 4A, C and I). The PM of CA/αCD (molar
4
ratio=1/1) and PM of CA/βCD (molar ratio=1/1) had diffraction peaks due to CA at 2θ=26.9º,
5
αCD at 2θ=12.1º, and βCD at 2θ=12.3º (Fig. 4E and K). The GM of CA/αCD (molar
6
ratio=2/1) had a diffraction peak due to CA at 2θ=26.9º, but the peak disappeared and a halo
7
pattern was produced with the GM of CA/αCD at molar ratios of both 1/1 and 1/2 (Fig. 4F, G
8
and H). The GM of CA/βCD (molar ratio=2/1) had a diffraction peak due to CA at 2θ=26.9º,
9
but the peak disappeared and a halo pattern was produced by the GM of CA/βCD at molar
10
ratios of both 1/1 and 1/2 (Fig. 4L, M and N). A halo pattern has been confirmed in the GM
11
of the CA/γCD complex [18]. The GM of CA/αCD and CA/βCD also have similar patterns,
12
which suggests that they form inclusion complexes. Co-grinding disrupts the structures of
13
crystals, and these amorphous solids can produce a halo pattern [23]. A previous study has
14
reported that supplying mechanical energy by grinding and friction can result in the
15
formation of amorphous inclusion complexes [14]. The GMs of CA/αCD and CA/βCD (both
16
molar ratios of 2/1) had an endothermic peak due to CA. The reason may be that excess CA
17
exists outside the CD complexes. This suggests that both CA/αCD and CA/βCD in a GM
18
(molar ratio=1/1) form an inclusion complex.
19
(Fig. 4)
20 21
3.4. Fourier transform infrared (FTIR) spectroscopy
22 23
FTIR spectroscopy was performed in order to examine the molecular state of the complexes
24
in a solid state. The intact CA had a peak due to the OH groups of its aromatic ring at 3433
25
cm-1. A peak due to the carbonyl group (C=O) was produced at 1644 cm-1, and a peak due to
26
the aromatic ring (C=C) was produced at 1622 cm-1 (Fig. 5A). In the PM of CA/αCD (molar 11 Page 11 of 30
1
ratio=1/1) and the PM of CA/βCD (molar ratio=1/1), the OH groups from the aromatic ring
2
in the CA molecule produced a peak at 3433 cm-1, the carbonyl group(C=O) produced a peak
3
at 1644 cm-1, and the aromatic ring (C=C) produced a peak at 1622 cm-1 (Fig. 5D and E). In
4
the GM of CA/αCD (molar ratio=1/1) and the GM of CA/βCD (molar ratio=1/1), the O-H
5
stretching band attributed to the hydroxyl groups of the aromatic ring in the CA molecules
6
were still observed at the same position as with the CA crystals. However, the peaks assigned
7
to aromatic ring (C-C) and the carbonyl group of CA were significantly shifted to 1622 and
8
1685 cm-1, respectively. The results suggest a molecular interaction between the CA and CD
9
molecules (Fig. 5F and G). Shifts in the CA aromatic rings and carbonyl group peaks have
10
already been confirmed in the GM of CA/γCD complexes [18]. The GMs of CA/αCD and
11
CA/βCD have similar peak shifts, which suggest that they also form inclusion complexes. CA
12
is known to form a dimeric structure as a result of intermolecular hydrogen bonding between
13
the carbonyl group (C=O) of one carboxylic acid and the hydroxyl group (-OH) of another
14
[24]. Higashi et al. reported that complex formation by molecules with this dimeric structure
15
causes hydrogen bonds between molecules to dissociate, resulting in an upward shift in the
16
peak due to carbonyl groups (C=O) [25]. Thus, molecular interaction is presumably occurring
17
between CA and αCD and between CA and βCD in the GMs (both molar ratios of 1/1) in the
18
solid state. Since the peak corresponding to the hydroxyl groups (-OH) of the aromatic ring of
19
CA at 3433 cm-1 did not change, the CD cavity and OH groups of CA presumably do not
20
interact in the GM of CA/αCD or in the GM of CA/βCD (both molar ratios of 1/1). (Fig. 5)
21 22 23
3.5. Dissolution profile
24 25
The results of the DSC, PXRD, and FTIR spectra suggest that CA and αCD as well as CA
26
and βCD in the GMs (both at molar ratios of 1/1) form an inclusion complex in the solid state. 12 Page 12 of 30
1
Thus, dissolution testing was performed next to ascertain changes in the dissolution behavior
2
of CA as a result of the formation of the inclusion complexes. Test samples were intact CA,
3
ground CA, PM of CA/αCD, PM of CA/βCD, GM of CA/αCD, and GM of CA/βCD (all
4
mixtures at molar ratios of 1/1). The dissolution rate of CA after 5 min was 24% for intact CA
5
and 23% for ground CA, so both had a slow rate of dissolution (Fig. 6). The dissolution rate
6
of CA after 5 min was 37% for the PM of CA/αCD and 59% for the PM of CA/βCD, so the
7
PMs showed a higher rate of dissolution than that of the intact CA. The dissolution rate of CA
8
after 5 min was 100% for both the GM of CA/αCD and the GM of CA/βCD, so the GMs
9
demonstrated improved dissolution in comparison to that of intact CA and the PMs. It has
10
been reported that disrupting the molecular arrangement of a drug by rendering it amorphous
11
and inclusion complex formation were possible mechanisms of enhanced dissolution [26, 27].
12
The results of the PXRD revealed that CA was amorphous, and they indicated inclusion
13
complex formation (Fig.2). The FTIR spectra revealed the interaction between the aromatic
14
ring of CA (which is a hydrophobic group) and the CDs. Thus, multiple factors, such as CA
15
in an amorphous form and inclusion complex formation, help improve the elution of CA in
16
CA/CDs. The improved dissolution in the PM of CA/αCD and in the PM of CA/βCD (both
17
molar ratios of 1/1) was presumably due to the formation of inclusion complexes in an
18
aqueous solution. In addition, improvement of dissolution was confirmed in the GM of the
19
CA/γCD complex [18]. Similarly, the GM of CA/αCD and the GM of CA/βCD have
20
improved solubility. It is likely that the formation of inclusion complexes contributes to the
21
dissolution property.
22
(Fig. 6)
23 24
3.6. Measurement of 1H-1H nuclear Overhauser effect spectroscopy (NOESY) NMR spectra
25 26
The results of the dissolution tests revealed that the GMs had an improved rate of 13 Page 13 of 30
1
dissolution compared to that of intact CA. The molecular interaction between CA and αCD
2
and between CA and βCD (both molar ratios of 1/1) in the GMs in solution may affect their
3
solubility. 1H-1H NOESY NMR spectroscopy was performed to examine these molecular
4
interactions in detail. 1H-1H NOESY NMR spectroscopy can reveal spatial interactions
5
between a guest molecule and the CD cavity, so it is used to predict the positioning of the
6
guest molecule within the inclusion complex [28]. In the GM of CA/αCD (molar ratio=1/1),
7
the H-3 proton (3.66 ppm) in the CD cavity, the H-a proton (6.08, 6.06 ppm) and H-b proton
8
(7.38, 7.40 ppm) of the vinylene group of CA, and the H-c proton (6.98, 6.99 ppm) and H-e
9
proton (6.96 ppm) of the aromatic ring produced cross peaks (Fig. 7A). The H-a and H-b
10
protons of the vinylene group of CA and the H-3 proton in the cavity of αCD in particular
11
produced intense cross peaks. Typically, the H-3 proton is found at the wider rim of the ring
12
of CD and the H-6 proton is found at the narrower rim of the ring of CD [28]. The appearance
13
of cross peaks indicates that the distance between protons is less than 4Å, and a more intense
14
peak indicates a smaller distance [29]. Thus, the vinylene group of the CA molecule appears
15
to be oriented from the wider to the narrower rim of the αCD ring (Fig. 8A). In the GM of
16
CA/βCD (molar ratio=1/1), cross peaks were produced by H-3 (3.67 ppm), H-5 (3.53, 3.54
17
ppm), and H-6 (3.60 ppm) in the cavity, by H-a (6.02, 6.05 ppm) and H-b (7.31, 7.33 ppm) in
18
the vinylene group of CA, and by H-c (6.80, 6.81 ppm), H-d (6.68, 6.69 ppm), and H-e (6.88
19
ppm) in the aromatic ring (Fig. 7B). In the GM of CA/βCD, intense cross peaks were
20
produced by H-b in the vinylene group of CA, by H-e in the aromatic ring, and by H-5 in the
21
cavity of βCD. No cross peaks were produced by H-c or H-d in the benzene ring or by H-3 in
22
the cavity of βCD. Thus, the aromatic ring of the CA molecule appears to be oriented from
23
the wider to the narrower rim of the ring of βCD (Fig. 8B). The results of the FTIR revealing
24
an interaction between the aromatic ring and vinylene group of the CA and the CDs has been
25
confirmed. This result suggests that the inclusion complex of the GM CA/CDs also involves
26
interactions of the aromatic ring and vinylene group of CA similar to that observed in 14 Page 14 of 30
1
aqueous solution. In the GM of CA/αCD (molar ratio=1/1), the aromatic ring was partially
2
included, while the aromatic ring was completely included in the GM of CA/βCD (molar
3
ratio=1/1). In the CA/γCD complex, there was a strong cross peak between the proton derived
4
from the aromatic ring of the CA molecule and the narrow edge of γCD [18]. Thus, the
5
aromatic ring of the CA molecule appears to be oriented from the wider to the narrower rim
6
of the ring of γCD. This result suggests that the GM of CA/αCD has a different inclusion
7
form compared to that of CA/βCD and CA/γCD. (Fig. 7), (Fig. 8)
8 9 10
3.7. DPPH radical scavenging test
11 12
It has been reported that inclusion complex formation increases antioxidant capacity [17].
13
The presence of a guest molecule in a CD molecule results in increased electron density,
14
resulting in a greater likelihood that the guest molecule will release protons as radicals to
15
quench the DPPH radical. Thus, a DPPH radical-scavenging activity test was conducted to
16
assess antioxidant capacity in the GM of CA/αCD (molar ratio=1/1) and the GM of CA/βCD
17
(molar ratio=1/1). The results of this test indicate that the 50% inhibitory concentration (IC50)
18
of DPPH radical scavenging was 2.57 µg/ml for intact CA, 1.42 µg/ml for the GM of
19
CA/αCD, and 1.77 µg/ml for the GM of CA/βCD (Fig. 9). The IC50 was significantly lower
20
for the GMs of CA/αCD and CA/βCD (both molar ratios of 1/1) compared to that of intact
21
CA, and increased antioxidant capacity was noted. The GM of CA/αCD (molar ratio=1/1)
22
tended to have higher antioxidant capacity. This might be explained by the stability constant
23
in aqueous solution affecting the antioxidant capacity; the results of solubility testing
24
indicated that CA/αCD had a higher stability constant than CA/βCD did [30]. 1H-1H NOESY
25
NMR spectra revealed that the form of the inclusion complex in the GM of CA/αCD (molar
26
ratio=1/1) was different from that of the GM of CA/βCD (molar ratio=1/1) (Fig. 7). 15 Page 15 of 30
1
Additionally, βCD has a more hydrophobic cavity than αCD does [31]. Thus, the form of the
2
CA inclusion in the GM of CA/βCD involves capture of the aromatic ring. In the GM of
3
CA/αCD, the inclusion primarily involves the vinylene group and not the aromatic ring. The
4
antioxidant capacity of CA is due to the hydroxyl groups attached to the aromatic ring,
5
suggesting that the differences in the orientation of the CA within the CD affected the
6
antioxidant capacity [32]. The GM of the CA/γCD complex did not improve antioxidant
7
capacity [18]. The GM of the CA/γCD complex has a lower stability constant than the GM of
8
the CA/αCD complex does, and the CA of the latter is complexed in a manner that captures
9
the aromatic rings. We believe that these two reasons explain the improved antioxidant
10
capacity of the GM of CA/αCD. (Fig. 9)
11 12 13
4. Conclusion
14
In the current study, phase solubility diagrams indicated that both CA/αCD and CA/βCD
15
form a complex at a molar ratio of 1/1. In addition, the results revealed that CA was
16
accommodated more stably inside the cavity of αCD than inside the cavity of βCD.
17
The results of the DSC, PXRD, FTIR, and dissolution tests revealed that both CA/αCD and
18
CA/βCD could form a 1/1 inclusion complex with an enhanced dissolution rate through the
19
co-grinding method. H- H NOESY NMR spectra revealed that the orientation of the guest
20
molecule inclusion differed depending on the form of CD. The results of a test of DPPH
21
radical-scavenging activity revealed that differences in stability constants and forms of
22
inclusion affected the antioxidant capacity of CA. These findings indicate that the complexes
23
of CA/αCD and CA/βCD have improved solubility and improved antioxidant capacity,
24
suggesting that CA may warrant increased use in functional foods.
1
1
25
16 Page 16 of 30
1 2
Acknowledgements The authors are grateful to Cyclo Chem Co., Ltd. for the provision of αCD and βCD.
3 4 5 6
Conflict of interests This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
7 8
References
9 10
[1] Ministry of Health, Labour and Welfare: General Welfare and Labour (2015). http://www.mhlw.go.jp/english/wp/wp-hw9/dl/01e.pdf
11 12
[2] Ishikawa A, Yamashita H, Hiemori M, et al. Characterization of Inhibitors of postprandial
13
hyperglycemia from the leaves of nerium indicum. J Nutr Sci Vitaminol
14
2007;53:166-173.
15
[3]
Hemmerle H, Burger HJ, Below P, et al. Chlorogenic Acid and Synthetic Chlorogenic
16
Acid Derivatives: Novel Inhibitors of Hepatic Glucose-6-phosphate Translocase. J Med
17
Chem 1997;40(2):137-145.
18
[4]
rescues diet-induced insulin resistance and obesity in mice. Nutr Metab 2015;22:12-19.
19 20
[5]
Jung UJ, Lee MK, Park YB, et al. Antihyperglycemic and antioxidant properties of caffeic acid in db/db mice. J Pharmacol Exp Ther 2006;318(2):476-483.
21 22
Ghadieh HE, Smiley ZN, Kopfman MW, et al. Chlorogenic acid/chromium supplement
[6]
Natella F, Nardini M, Belelli F, et al. Coffee drinking induces incorporation of phenolic
23
acids into LDL and increases the resistance of LDL to ex vivo oxidation in humans. Am
24
J Clin Nutr 2007;86(3):604-609.
25 26
[7]
Brewster ME, Loftsson T. Cyclodextrins as pharmaceutical solubilizers. Adv Drug Deliv Rev 2007;59(7):645-666. 17 Page 17 of 30
1
[8]
Inoue Y, Iohara D, Sekiya N, et al. Ternary inclusion complex formation and
2
stabilization of limaprost, a prostaglandin E1 derivative, in the presence of α- and
3
β-cyclodextrins in the solid state. Int J Pharm 2016;509(1-2):338-347.
4
[9]
Arima H, Yunomae K, Miyake K, et al. Comparative studies of the enhancing effects of
5
cyclodextrins on the solubility and oral bioavailability of tacrolimus in rats. J Pharm Sci
6
2001;90(6):690-701.
7
[10] Anselmi C, Centini M, Ricci M, et al. Analytical characterization of a ferulic
8
acid/gamma-cyclodextrin inclusion complex. J Pharm Biomed Anal
9
2005;40(4):875-881.
10
[11] Yáñez C, Cañete-Rosales P, Castillo JP, et al. Cyclodextrin inclusion complex to
11
improve physicochemical properties of herbicide bentazon: exploring better
12
formulations. PLoS One 2012;7(8):e41072.
13 14 15
[12] Wang J, Cao Y, Sun B, et al. Characterisation of inclusion complex of trans-ferulic acid and hydroxypropyl-β-cyclodextrin. Food Chemistry 2011;124:1069-1075. [13] Hu Y, Gniado K, Erxleben A, et al. Mechanochemical Reaction of Sulfathiazole with
16
Carboxylic Acids: Formation of a Cocrystal, a Salt, and Coamorphous Solids Cryst.
17
Growth Des 2014;14(2):803-813.
18
[14] Ezawa T, Inoue Y, Tunvichien S, et al. Changes in the Physicochemical Properties of
19
Piperine/β-Cyclodextrin due to the Formation of Inclusion Complexes. Int J Med Chem
20
2016. doi:10.1155/2016/8723139
21
[15] Ikeda H, Ikuta N, Nakata D, et al. NMR Studies of Inclusion Complexes Formed by
22
(R)-α-Lipoic Acid with α-, β-, and γ-Cyclodextrins. Bulletin of the Chemical Society of
23
Japan 2015;88(8):1123-1127.
24
[16] Pápay Z.E, Sebestyén Z, Ludányi K, et al. Comparative evaluation of the effect of
25
cyclodextrins and pH on aqueous solubility of apigenin. Journal of Pharmaceutical and
26
Biomedical Analysis 2016;117:210-216. 18 Page 18 of 30
1
[17] Chao J, Wang H, Zhao W, et al. Investigation of the inclusion behavior of chlorogenic
2
acid with hydroxypropyl-β cyclodextrin. Int J Biol Macromol 2012;50:277-282.
3
[18] Inoue Y, Suzuki K, Ezawa T, et al. Examination of the physicochemical properties of
4
caffeic acid complexed with γ-cyclodextrin. J Incl Phenom Macrocycl Chem
5
2015;83:289-298.
6 7 8
[19] Higuchi T, Connors KA. Phase-solubility techniques. Adv Anal Chem Instrum 1961;4:117-212. [20] Göktürk S, Çalışkan E, Talman RY, et al. A study on solubilization of poorly soluble
9
drugs by cyclodextrins and micelles:complexation and binding characteristics of
10
sulfamethoxazole and trimethoprim. Scientific World Journal 2012;2012:718791.
11
[21] Iacovino R, Rapuano F, Caso JV, et al. β-Cyclodextrin Inclusion Complex to Improve
12
Physicochemical Properties of Pipemidic Acid: Characterization and Bioactivity
13
Evaluation. Int J Mol Sci 2013;14(7):13022-13041.
14
[22] Suzuki R, Inoue Y, Tsunoda Y, et al. Effect of γ-cyclodextrin derivative complexation
15
on the physicochemical properties and antimicrobial activity of hinokitiol. J Incl
16
Phenom Macrocycl Chem 2015;83:177-186.
17
[23] Iwata M, Fukami T, Kawashima D, et al. Effectiveness of mechanochemical treatment
18
with cyclodextrins on increasing solubility of glimepiride. Pharmazie
19
2009;64(6):390-394.
20
[24] Xing Y, Peng HY, Zhang MX, et al. Caffeic acid product from the highly
21
copper-tolerant plant Elsholtzia splendens post-phytoremediation: its extraction,
22
purification,and identification. J Zhejiang Univ Sci B 2012;13:487-493.
23 24 25 26
[25] Higashi K, Tozuka Y, Moribe K, et al. Salicylic Acid/γ-Cyclodextrin 2:1 and 4:1 Complex Formation by Sealed-Heating Method. J Pharm 2010;10:4192-4200. [26] Trapani G, Latrofa A, Franco M, et al. Complexation of zolpidem with 2-hydroxypropyl-beta-, methyl-beta-, and 2-hydroxypropyl-gamma-cyclodextrin: effect 19 Page 19 of 30
1
on aqueous solubility, dissolution rate, and ataxic activity in rat. J Pharm Sci
2
2000;89(11):1443-1451.
3
[27] Corciova A, Ciobanu C, Poiata A, et al. Antibacterial and antioxidant properties of
4
hesperidin:β-cyclodextrin complexes obtained by different techniques. J Incl Phenom
5
Macrocycl Chem 2015;81:71-84.
6
[28] Inoue Y, Watanabe S, Suzuki R, et al. Evaluation of actarit/γ-cyclodextrin complex
7
prepared by different methods. J Incl Phenom Macrocycl Chem 2015;81:161-168.
8
[29] Kirschner DL, Green TK. Nonaqueous synthesis of a selectively modified, highly
9
anionic sulfopropyl ether derivative of cyclomaltoheptaose (β-cyclodextrin) in the
10
presence of 18-crown-6. Carbohydrate Research 2005;340:1773-1779.
11
[30] Aytac Z, Uyar T. Antioxidant activity and photostability of α-tocopherol/β-cyclodextrin
12
inclusion complex encapsulated electrospun polycaprolactone nanofibers. European
13
Polymer Journal 2016;79:140-149.
14 15 16
[31] Naidoo KJ, Chen JY, Jansson JLM, et al. Molecular Properties Related to the Anomalous Solubility of β-Cyclodextrin. J Phys Chem B 2004;108(14):4236-4238 [32] Kfoury M, Landy D, Auezova L, et al. Effect of cyclodextrin complexation on
17
phenylpropanoids’ solubility and antioxidant activity. Beilstein J Org Chem
18
2014;10:2322-2331.
19 20
Figure and Table legends
21 22
Figures:
23 24 25
A)
OH
26 20
O
OH OH
Page 20 of 30
1 2 3 4 5 6 7 8 9 10 11 12
Fig. 1. Chemical structure.
13
(A) Caffeic acid (CA), (B-1) αCyclodextrin (αCD), (B-2) βCyclodextrin (βCD)
14
21 Page 21 of 30
1 2 3 4 5 6
40
7 8 9
30
10
12 13 14
CA (mM)
11
20
15
αCD
16 17
10
βCD
18 19 20 21 22 23 24
0 0
10
20 CDs (mM)
30
40
Fig. 2. Phase solubility diagram of CA/CDs. Results are expressed as means ±S.D. (n = 3).
25
22 Page 22 of 30
1 2
(A)
3 4
223
5 6
(B) 210
7 8
(C
207
9 10 11 12
(D) Endothermic (E)
13 14
(F)
15
216
16 17
(G) 197
18 19
(H
20 21
(I)
22 23 24 25
100
200
150
250
Temperature(ºC)
26
Fig. 3. DSC curves of CA/CD systems.
27
(A) CA intact, (B) PM (CA/αCD=1/1), (C) GM (CA/αCD=2/1), (D) GM (CA/αCD=1/1), (E) GM (CA/αCD=1/2), (F) PM (CA/βCD=1/1), (G) GM (CA/βCD=2/1), (H) GM (CA/βCD=1/1), (I) GM (CA/βCD=1/2)
28 29 30
23 Page 23 of 30
1 2 3 4 5 6 (A)
7 8 9
(B)
10 11 12
(I)
(C)
13
16 17 18
Intensity (cps)
15
Intensity (cps)
14 (D)
(E)
19
(J)
(K)
(F)
(L)
(G)
(M)
(H)
(N)
20 21 22 23 24 25 26 27 28 29 30 31 32
5
10
15
20
25
30
5
2θ degree
10
15
20
25
30
2θ degree
Fig. 4. PXRD patterns of CA/CDs. (A) CA intact, (B) CA ground, (C) αCD, (D) αCD ground, (E) PM (CA/αCD=1/1), (F) GM (CA/αCD=2/1), (G) GM (CA/αCD=1/1), (H) GM (CA/αCD=1/2), (I) βCD, (J) βCD ground, (K) PM (CA/βCD=1/1), (L) GM (CA/βCD=2/1), (M) GM (CA/βCD=1/1), (N) GM (CA/βCD=1/2)
33
24 Page 24 of 30
1 2 3 4
(A)
5
22
3433
6
44
7
(B)
8 9
(C)
10 11 12 13
15 16 17 18
Transmission (%)
14
(D) 22 46
(E)
3433
22 46
19
3433
20
(F)
21
85
13 44
22 3433
23
(G)
24 25
13
85
26
44
3433
27 28 29 30 31 32
3700
3000
2600 1850 Wavenumber(cm-1 )
1600 1500
1300
Fig. 5. FTIR spectra of CA/CD systems. (A) CA intact, (B) αCD, (C) βCD, (D) PM (CA/αCD=1/1), (E) PM (CA/βCD=1/1), (F) GM (CA/αCD=1/1), (G) GM (CA/βCD=1/1)
33
25 Page 25 of 30
1 2
Percentage dissolved (%)
100
80
60
40
CA intact CA ground GM(CA/αCD=1/1)
20
GM(CA/βCD=1/1) PM(CA/αCD=1/1) PM(CA/βCD=1/1)
0 0
30
60
90
120
Time (min) 3 4
Fig. 6. Dissolution profiles of CA/CD systems. Results are expressed as means ±S.D. (n = 3).
5
26 Page 26 of 30
1 2 3 c
4 O
5 6 7
d
a OH
b
e
4
OH OH
6
OH O
5 2
HO
3
e
O
OH
b
1
c d
Glucopyranose
Caffeic acid
a
(A)
8
4 2
9 10 11 12
5 6
13
3
14 15 16 17
e
18
b
cd
a
19 20
(B) 4 2
21 22 23
5
24 25
6
26
3
27 28 29 1
30 31
1
Fig. 7. H- H NOESY NMR spectra of CA/CD systems. (A) GM (CA/αCD=1/1), (B) GM (CA/βCD=1/1)
32 33
27 Page 27 of 30
1 2 3 4 5 6
(A)
(B)
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Fig. 8. Proposed structural images of CA/CDs complex. (A) side view of GM (CA/αCD), (B) side view of GM (CA/βCD)
23
28 Page 28 of 30
1 2 3
＊
4
＊
5
3.0
6
2.57
7 8
2.5 2.05
9
2.03
10
2.0
12 13 14
IC50 (μg/ml)
11
1.77 1.42
1.5
15 16
1.0
17 18
0.5
19 20 21
0.0
22 23 24 25 26
Fig. 9. IC50 of DPPH radical scavenging test of CA/CD systems.
27
Results are expressed as means ± S.D. (n = 3) ＊ :P<0.05 vs CA (Tukey’s test).
28
29 Page 29 of 30
1
Table 1 Apparent stability constant (Ks) of CA inclusion complexes. CA complex
Ks ± S.D. (M-1)
αCD
1547.5±30.9
βCD
371.4±6.22
γCD
391.2±25.8
2 3
30 Page 30 of 30
S1818-0876(17)30352-5 https://doi.org/doi:10.1016/j.ajps.2017.08.006 AJPS 462
To appear in:
Asian Journal of Pharmaceutical Sciences
Received date: Revised date: Accepted date:
8-5-2017 8-8-2017 15-8-2017
Please cite this article as: Ryota Shiozawa, Yutaka Inoue, Isamu Murata, Ikuo Kanamoto, Effect of antioxidant activity of caffeic acid with cyclodextrins using ground mixture method, Asian Journal of Pharmaceutical Sciences (2017), https://doi.org/doi:10.1016/j.ajps.2017.08.006. 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.
1
Title page
2
Effect of antioxidant activity of caffeic acid with cyclodextrins
3
using ground mixture method
4 5
Ryota Shiozawaa, Yutaka Inouea*, Isamu Murataa, Ikuo Kanamotoa
6 7
a
Faculty of Pharmacy and Pharmaceutical Sciences, Josai University, 1-1 Keyakidai, Sakado-shi, Saitama, 3500295, Japan
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Corresponding author: Yutaka Inouea*
24
Mailing address: Laboratory of Drug Safety Management, Faculty of Pharmacy and
25
Pharmaceutical Sciences, Josai University, 1-1 Keyakidai, Sakado-shi, Saitama, 350-0295,
26
Japan.
27
Tel.: +81-49-271-7317; Fax: +81-49-271-7317
28
E-mail: [email protected]
29
1 Page 1 of 30
1
Graphical Abstract
2
Caffeic acid/α-cyclodextrin (molar ratio=1/1), the vinylene group of the caffeic acid molecule
3
appears to be included from the wider to the narrower rim of the ring of α-cyclodextrin.
4
Caffeic acid /β-cyclodextrin (molar ratio=1/1), the aromatic ring of the caffeic acid molecule
5
appears to be included from the wider to the narrower rim of the ring of βCD. This suggests
6
that forms of inclusion differed in the caffeic acid /α-cyclodextrin (molar ratio=1/1) and the
7
caffeic acid /β-cyclodextrin (molar ratio=1/1).
8 9 10 11 12 13 Caffeic acid/β-cyclodextrin
Caffeic acid/α-cyclodextrin
Structural images of caffeic acid/cyclodextrin complex from side view.
14
Abstract:
15 16
In the current study, we prepared a ground mixture (GM) of caffeic acid (CA) with
17
α-cyclodextrin (αCD) and with β-cyclodextrin (βCD), and then comparatively assessed the
18
physicochemical properties and antioxidant capacities of these GMs. Phase solubility
19
diagrams indicated that both CA/αCD and CA/βCD formed a complex at a molar ratio of 1/1.
20
In addition, stability constants suggested that CA was more stable inside the cavity of αCD
21
than inside the cavity of βCD. Results of powder X-ray diffraction (PXRD) indicated that the
22
characteristic diffraction peaks of CA and CD disappeared and a halo pattern was produced
23
by the GMs of CA/αCD and CA/βCD (molar ratios=1/1). Dissolution testing revealed that
24
both GMs had a higher rate of dissolution than CA alone did. Based on the 1H-1H NOESY 2 Page 2 of 30
1
NMR spectra for the GM of CA/αCD, the vinylene group of the CA molecule appeared to be
2
included from the wider to the narrower rim of the αCD ring. Based on spectra for the GM of
3
CA/βCD, the aromatic ring of the CA molecule appeared to be included from the wider to the
4
narrower rim of the βCD ring. This suggests that the structures of the CA inclusion
5
complexes differed between those involving αCD rings and those involving βCD rings.
6
Results of a DPPH radical-scavenging activity test indicated that the GM of CA/αCD had a
7
higher antioxidant capacity than that of the GM of CA/βCD. The differences in the
8
antioxidant capacities of the GMs of CA/αCD and CA/βCD are presumably due to
9
differences in stability constants and structures of the inclusion complexes.
10 11
Keywords: Caffeic acid; Cyclodextrins; Inclusion; Antioxidant activity; Solubility; Ground
12
mixture
13
3 Page 3 of 30
1
1. Introduction
2 3
In recent years, cardiovascular diseases have become the second leading cause of death in
4
Japan, after cancer [1]. Therefore, several health foods are being investigated to treat the risk
5
factors of cardiovascular diseases, which include dyslipidemia- and diabetes-induced
6
atherosclerosis. Chlorogenic acid, a coffee polyphenol, is one such health food that has been
7
reported to protect against lifestyle diseases, such as dyslipidemia and diabetes [2-4]. Caffeic
8
acid (CA) is a decomposition product of chlorogenic acid that has the same physiological
9
activity as chlorogenic acid does. Although CA has been reported to protect against
10
dyslipidemia and blood glucose elevation owing to its antioxidant activity, its application is
11
limited by poor water solubility and oral absorption [5, 6]. Cyclodextrin (CD) is a
12
ring-shaped molecule consisting of 6, 7, or 8 glucose units for α, β, or γCD, respectively,
13
which are linked by α-(1→4)-glucosidic bonds. In the safety assessment of the Food and
14
Agriculture Organization of the United Nations/World Health Organization (FAO/WHO)
15
Joint Expert Committee on Food Additives (JECFA), the acceptable daily intake (ADI) was
16
set as ‘‘not specified’’ for α- and γCD and ‘‘up to 5 mg/kg/day’’ for βCD. CD molecules
17
have a hydrophobic cavity in their center that can incorporate various guest molecules to
18
form inclusion complexes [7]. Inclusion within CD is known to stabilize pharmaceuticals [8]
19
and improve their bioavailability by improving their solubility [9]. Various methods, such as
20
coprecipitation [10], kneading [11], and lyophilization [12], have been reported for the
21
preparation of inclusion complexes. Co-grinding, a solvent-free, mechanochemical method of
22
forming inclusion complexes, causes changes in physicochemical properties because of the
23
application of mechanical energy (i.e., by the grinding of the solid material), and has been
24
used for the complexation of piperine, which is poorly water soluble [13,14]. The preparation
25
of (R)-α-lipoic acid and CD inclusion complexes, which is affected by the cavity size of CD,
26
has also been reported [15]. As the steric properties and antioxidant capacity of drugs are 4 Page 4 of 30
1
reported to be affected by the type of CD [16], it is useful to evaluate the physicochemical
2
properties of inclusion complexes prepared using CDs of different cavity sizes. In addition,
3
CDs have been reported to increase antioxidant capacity by their inclusion of polyphenols,
4
such as chlorogenic acid [17]. However, there is no report of increased antioxidant capacity
5
of CA using CDs. Fundamental research to evaluate the antioxidant capacity of inclusion
6
complexes prepared using CDs of different cavity sizes to improve the antioxidant capacity
7
of CA may have clinical applications. We have already reported the preparation of CA and
8
γCD inclusion complexes by co-grinding and demonstrated that the antioxidant capacity is
9
retained owing to improvement in CA inclusion and elution [18]. Since αCD and βCD have a
10
narrower cavity than γCD does, they may form different inclusion complexes and thus affect
11
the antioxidant capacity of drugs differently. Therefore, in this study, we prepared CA/CD
12
inclusion complexes using αCD and βCD, and evaluated their physicochemical properties,
13
dissolution properties, and antioxidant capacities.
14
5 Page 5 of 30
1
2. Materials and methods
2 3
2.1. Materials
4
CA (Fig. 1A) was purchased from Wako Pure Chemical Industries Co. Ltd (Japan). αCD
5
and βCD (Fig. 1B) were a generous gift from CycloChem Co. Ltd. (Tokyo, Japan) and were
6
stored
7
2,2-Diphenyl-1-picrylhydrazyl (DPPH) was purchased from Sigma-Aldrich Co. Ltd. All other
8
chemicals and solvents were of analytical grade and were purchased from Wako Pure
9
Chemical Industries Co. Ltd.
at
40
°C
and
82%
relative
humidity
for
7
days,
prior
to
use.
(Fig. 1)
10 11 12
2.2. Preparation of the physical mixture (PM) and ground mixture (GM)
13
The PM was prepared by mixing CA and CD in different molar ratios (2/1, 1/1, and 1/2)
14
using a vortex mixer for 1 min. The GM was prepared by placing the PM (1 g) in an alumina
15
cell and grinding for 60 min using a vibration rod mill (TI-500ET, CMT Co. Ltd., Japan).
16 17
2.3. Methods
18
2.3.1. Phase solubility studies
19
Phase solubility studies were performed as previously described by Higuchi and Connors
20
[19]. A supersaturated amount of CA (100 mg) was added to an aqueous solution (10 ml) of
21
αCD (0-40 mM) or βCD (0-15 mM), and the mixture was placed in a medium-sized, shaking
22
instrument for 24 h at 200 rpm and 25±0.5 °C. After the suspension reached equilibrium, the
23
solution was immediately filtered through a 0.45 μm membrane filter. The solubility was
24
quantified by high-performance liquid chromatography using a Waters® e2795 separations
25
module (Nippon Waters Co., Ltd., Japan) and Cosmosil 5C18-AR-II packed column (4.6
26
mm×150 mm) at a detection wavelength of 285 nm. The sample injection volume was 30 μl, 6 Page 6 of 30
1
and the column temperature was 40 °C. The mobile phase consisted of distilled
2
water/acetonitrile/methanol/acetic acid (862/113/20/5), and the CA retention time was 10 min.
3
The apparent stability constant (Ks) of the CA/CD inclusion complexes was calculated from
4
the slope of the phase solubility diagram and solubility of CA in the absence of CD (S 0) using
5
equation (1).
6
7
Ks
slope
(1)
S 0 (1 slope )
8 9
2.3.2. Differential scanning calorimetry (DSC)
10
The thermal behavior of samples was recorded using a differential scanning calorimeter
11
(Thermo plus EVO, Rigaku Co., Japan). All samples were weighed (2 mg) and heated at a
12
scanning rate of 5.0 °C/min under nitrogen flow (60 ml/min). Aluminum pans and lids were
13
used for all samples.
14 15
2.3.3. Powder X-ray diffraction (PXRD)
16
PXRD studies were performed using a powder X-ray diffractometer (MiniFlex II, Rigaku
17
Co., Japan) with Cu Kα radiation (30 kV, 15 mA). The diffraction intensity was measured
18
using a NaI scintillation counter. The samples were scanned from 2θ = 3–35° at a scan rate of
19
4°/min. The powder sample was spread on a glass plate to maintain a flat sample plane, and
20
measurements were performed.
21 22
2.3.4. Fourier transform infrared (FTIR) spectroscopy
23
The FTIR spectra of samples were recorded using an FTIR spectrometer (FT/IR-410,
24
JASCO Inc., Japan) by the potassium bromide (KBr) disk method. The samples were scanned
25
32 times over a range of 650-4,000 cm-1 with a resolution of 4 cm-1. The tablets were 7 Page 7 of 30
1
prepared by adding KBr to the sample (sample/KBr = 1/10 (w/w)) and manually compressing
2
the mixture. Background correction was performed using a blank KBr tablet.
3 4
2.3.5. Measurement of 1H-1H nuclear Overhauser effect spectroscopy (NOESY) NMR spectra
5
The NOESY spectra were recorded on a 700 MHz Varian NMR spectrometer (Agilent
6
Technologies) under the following conditions: solvent: D2O; resonant frequency: 699.6 MHz;
7
pulse width: 90°; relaxation delay: 0.500 s; scan time: 0.500 s; temperature: 25 °C; 256
8
increments.
9 10
2.3.6. Dissolution profile
11
Dissolution testing of samples was performed using a dissolution apparatus (NTR-593,
12
Toyama Sangyo Co. Ltd., Japan) with 900 ml of distilled water at 37±0.5 °C and 50 rpm,
13
according to the paddle method (Japanese Pharmacopoeia 17 th Edition). CA (100 mg) was
14
weighed accurately and loaded into the paddle apparatus. Dissolved samples (10 ml) were
15
collected at 5, 10, 15, 30, 60, 90, and 120 min through 0.45 μm membrane filters, with
16
replacement of an equal volume of fresh dissolution medium at the same temperature, to
17
maintain a constant volume of dissolution medium. The concentration of CA in the samples
18
was determined by the same method as described for the phase solubility studies (section
19
2.3.1).
20 21
2.3.7. DPPH radical scavenging activity
22
The DPPH radical scavenging activity of the samples was measured using a microplate
23
reader (SpectraMax 190, Molecular Devices, Japan). Briefly, each sample and DPPH
24
methanolic solution (100 μM) were mixed in a microplate at a ratio of 1/1 (v/v) and incubated
25
at 25 °C for 5 min. The absorbance of DPPH in samples (As) was measured at 517 nm. The
26
absorbance of a mixture of methanol/water (1/1 (v/v)) (Br), which has a radical removal rate 8 Page 8 of 30
1
of 100%, and a mixture of DPPH/water (1/1 (v/v)) (A0), which has a radical removal rate of
2
0%, was also measured. The radical scavenging rate was calculated using equation (2) [17].
3 4
Radical scavenging rate = [1-(As-Br)/(A0-Br)] ×100
(2)
5 6 7
3. Results and discussion
8 9
3.1. Phase solubility studies
10 11
Phase solubility studies were conducted to assess the molar ratio of the CA/CD inclusion
12
complexes and their stability constants in an aqueous solution. Results of the solubility
13
testing indicated that the solubility of CA increased linearly for αCD and βCD, producing an
14
AL type of phase solubility diagram according to the classification system of Higuchi et al.
15
(Fig. 2). Typically, an AL diagram indicates that a complex is formed at a molar ratio of 1/1
16
[20], so both CA/αCD and CA/βCD formed a complex at a molar ratio of 1/1 in solution. The
17
stability constant (Ks) was determined next. Ks was 1547.5 M-1 for αCD and 371.4 M-1 for
18
βCD (Table 1). A study by Iacovino et al. reported that, when a complex has a stability
19
constant exceeding 1000 M-1, the guest molecule and CD are unlikely to dissociate [21]. CA
20
included in αCD had a stability constant of 1547.5 M-1, so CA was more stable inside the
21
cavity of αCD than it was inside the cavity of βCD.
22
(Fig2) (Table 1)
23 24
3.2. Differential scanning calorimetry (DSC)
25 26
Phase solubility diagrams suggested that both CA/αCD and CA/βCD formed a complex at
27
a molar ratio of 1/1 in an aqueous solution. When CA/αCD and CA/βCD are ground together, 9 Page 9 of 30
1
they presumably still form an inclusion complex at a molar ratio of 1/1. It has been reported
2
that when a guest molecule does not melt or its peak shifts as a result of inclusion complex
3
formation caused by co-grinding, changes in its thermal behavior become evident [22]. Thus,
4
DSC was performed in the current study to examine the thermal behavior of the GMs. Results
5
of DSC revealed an endothermic peak due to melting at 223ºC for intact CA (Fig. 3A). The
6
PM of CA/αCD (molar ratio=1/1) had an endothermic peak due to the melting of CA at
7
210ºC, and the PM of CA/βCD (molar ratio=1/1) had an endothermic peak due to the melting
8
of CA at 216ºC, so CA crystals were presumably present (Fig. 3B and F). The GM of
9
CA/αCD (molar ratio=2/1) had an endothermic peak due to CA at 207ºC (Fig. 3C). The
10
endothermic peak due to CA disappeared for the GM of CA/αCD at molar ratios of 1/1 and
11
1/2 (Fig. 3D and E). The GM of CA/βCD (molar ratio=2/1) had an endothermic peak due to
12
CA at 197ºC (Fig. 3G). The endothermic peak due to CA disappeared for the GM of CA/αCD
13
at molar ratios of 1/1 and 1/2 (Fig. 3H and I). In our previous study, the disappearance of the
14
CA melting peak was confirmed by DSC measurement for the GM of the CA/γCD complex
15
[18]. Therefore, the GMs of CA/αCD and CA/βCD are also considered inclusion complexes.
16
The endothermic peak due to CA disappeared because the CA and αCD molecules and the
17
CA and βCD molecules formed inclusion complexes. The GM of CA/αCD (molar ratio=2/1)
18
and the GM of CA/βCD (molar ratio=2/1) showed an endothermic peak due to CA. The
19
reason may be that there was excess CA that did not complex with the CDs. This suggests
20
that both CA/αCD and CA/βCD were present in the GMs at a molar ratio of CA/CD of 1/1. (Fig. 3)
21 22 23
3.3. Powder X-ray diffraction (PXRD)
24 25
The results of DSC suggested that CA could form an inclusion complex with either αCD or
26
βCD at a stoichiometric ratio of 1/1. Thus, PXRD was performed to examine the crystalline 10 Page 10 of 30
1
state of the CA/CDs in the GMs. Intact CA and ground CA had a characteristic diffraction
2
peak at 2θ=26.9º, αCD had a characteristic diffraction peak at 2θ=11.8º, and βCD had a
3
characteristic diffraction peak at 2θ=12.3º (Fig. 4A, C and I). The PM of CA/αCD (molar
4
ratio=1/1) and PM of CA/βCD (molar ratio=1/1) had diffraction peaks due to CA at 2θ=26.9º,
5
αCD at 2θ=12.1º, and βCD at 2θ=12.3º (Fig. 4E and K). The GM of CA/αCD (molar
6
ratio=2/1) had a diffraction peak due to CA at 2θ=26.9º, but the peak disappeared and a halo
7
pattern was produced with the GM of CA/αCD at molar ratios of both 1/1 and 1/2 (Fig. 4F, G
8
and H). The GM of CA/βCD (molar ratio=2/1) had a diffraction peak due to CA at 2θ=26.9º,
9
but the peak disappeared and a halo pattern was produced by the GM of CA/βCD at molar
10
ratios of both 1/1 and 1/2 (Fig. 4L, M and N). A halo pattern has been confirmed in the GM
11
of the CA/γCD complex [18]. The GM of CA/αCD and CA/βCD also have similar patterns,
12
which suggests that they form inclusion complexes. Co-grinding disrupts the structures of
13
crystals, and these amorphous solids can produce a halo pattern [23]. A previous study has
14
reported that supplying mechanical energy by grinding and friction can result in the
15
formation of amorphous inclusion complexes [14]. The GMs of CA/αCD and CA/βCD (both
16
molar ratios of 2/1) had an endothermic peak due to CA. The reason may be that excess CA
17
exists outside the CD complexes. This suggests that both CA/αCD and CA/βCD in a GM
18
(molar ratio=1/1) form an inclusion complex.
19
(Fig. 4)
20 21
3.4. Fourier transform infrared (FTIR) spectroscopy
22 23
FTIR spectroscopy was performed in order to examine the molecular state of the complexes
24
in a solid state. The intact CA had a peak due to the OH groups of its aromatic ring at 3433
25
cm-1. A peak due to the carbonyl group (C=O) was produced at 1644 cm-1, and a peak due to
26
the aromatic ring (C=C) was produced at 1622 cm-1 (Fig. 5A). In the PM of CA/αCD (molar 11 Page 11 of 30
1
ratio=1/1) and the PM of CA/βCD (molar ratio=1/1), the OH groups from the aromatic ring
2
in the CA molecule produced a peak at 3433 cm-1, the carbonyl group(C=O) produced a peak
3
at 1644 cm-1, and the aromatic ring (C=C) produced a peak at 1622 cm-1 (Fig. 5D and E). In
4
the GM of CA/αCD (molar ratio=1/1) and the GM of CA/βCD (molar ratio=1/1), the O-H
5
stretching band attributed to the hydroxyl groups of the aromatic ring in the CA molecules
6
were still observed at the same position as with the CA crystals. However, the peaks assigned
7
to aromatic ring (C-C) and the carbonyl group of CA were significantly shifted to 1622 and
8
1685 cm-1, respectively. The results suggest a molecular interaction between the CA and CD
9
molecules (Fig. 5F and G). Shifts in the CA aromatic rings and carbonyl group peaks have
10
already been confirmed in the GM of CA/γCD complexes [18]. The GMs of CA/αCD and
11
CA/βCD have similar peak shifts, which suggest that they also form inclusion complexes. CA
12
is known to form a dimeric structure as a result of intermolecular hydrogen bonding between
13
the carbonyl group (C=O) of one carboxylic acid and the hydroxyl group (-OH) of another
14
[24]. Higashi et al. reported that complex formation by molecules with this dimeric structure
15
causes hydrogen bonds between molecules to dissociate, resulting in an upward shift in the
16
peak due to carbonyl groups (C=O) [25]. Thus, molecular interaction is presumably occurring
17
between CA and αCD and between CA and βCD in the GMs (both molar ratios of 1/1) in the
18
solid state. Since the peak corresponding to the hydroxyl groups (-OH) of the aromatic ring of
19
CA at 3433 cm-1 did not change, the CD cavity and OH groups of CA presumably do not
20
interact in the GM of CA/αCD or in the GM of CA/βCD (both molar ratios of 1/1). (Fig. 5)
21 22 23
3.5. Dissolution profile
24 25
The results of the DSC, PXRD, and FTIR spectra suggest that CA and αCD as well as CA
26
and βCD in the GMs (both at molar ratios of 1/1) form an inclusion complex in the solid state. 12 Page 12 of 30
1
Thus, dissolution testing was performed next to ascertain changes in the dissolution behavior
2
of CA as a result of the formation of the inclusion complexes. Test samples were intact CA,
3
ground CA, PM of CA/αCD, PM of CA/βCD, GM of CA/αCD, and GM of CA/βCD (all
4
mixtures at molar ratios of 1/1). The dissolution rate of CA after 5 min was 24% for intact CA
5
and 23% for ground CA, so both had a slow rate of dissolution (Fig. 6). The dissolution rate
6
of CA after 5 min was 37% for the PM of CA/αCD and 59% for the PM of CA/βCD, so the
7
PMs showed a higher rate of dissolution than that of the intact CA. The dissolution rate of CA
8
after 5 min was 100% for both the GM of CA/αCD and the GM of CA/βCD, so the GMs
9
demonstrated improved dissolution in comparison to that of intact CA and the PMs. It has
10
been reported that disrupting the molecular arrangement of a drug by rendering it amorphous
11
and inclusion complex formation were possible mechanisms of enhanced dissolution [26, 27].
12
The results of the PXRD revealed that CA was amorphous, and they indicated inclusion
13
complex formation (Fig.2). The FTIR spectra revealed the interaction between the aromatic
14
ring of CA (which is a hydrophobic group) and the CDs. Thus, multiple factors, such as CA
15
in an amorphous form and inclusion complex formation, help improve the elution of CA in
16
CA/CDs. The improved dissolution in the PM of CA/αCD and in the PM of CA/βCD (both
17
molar ratios of 1/1) was presumably due to the formation of inclusion complexes in an
18
aqueous solution. In addition, improvement of dissolution was confirmed in the GM of the
19
CA/γCD complex [18]. Similarly, the GM of CA/αCD and the GM of CA/βCD have
20
improved solubility. It is likely that the formation of inclusion complexes contributes to the
21
dissolution property.
22
(Fig. 6)
23 24
3.6. Measurement of 1H-1H nuclear Overhauser effect spectroscopy (NOESY) NMR spectra
25 26
The results of the dissolution tests revealed that the GMs had an improved rate of 13 Page 13 of 30
1
dissolution compared to that of intact CA. The molecular interaction between CA and αCD
2
and between CA and βCD (both molar ratios of 1/1) in the GMs in solution may affect their
3
solubility. 1H-1H NOESY NMR spectroscopy was performed to examine these molecular
4
interactions in detail. 1H-1H NOESY NMR spectroscopy can reveal spatial interactions
5
between a guest molecule and the CD cavity, so it is used to predict the positioning of the
6
guest molecule within the inclusion complex [28]. In the GM of CA/αCD (molar ratio=1/1),
7
the H-3 proton (3.66 ppm) in the CD cavity, the H-a proton (6.08, 6.06 ppm) and H-b proton
8
(7.38, 7.40 ppm) of the vinylene group of CA, and the H-c proton (6.98, 6.99 ppm) and H-e
9
proton (6.96 ppm) of the aromatic ring produced cross peaks (Fig. 7A). The H-a and H-b
10
protons of the vinylene group of CA and the H-3 proton in the cavity of αCD in particular
11
produced intense cross peaks. Typically, the H-3 proton is found at the wider rim of the ring
12
of CD and the H-6 proton is found at the narrower rim of the ring of CD [28]. The appearance
13
of cross peaks indicates that the distance between protons is less than 4Å, and a more intense
14
peak indicates a smaller distance [29]. Thus, the vinylene group of the CA molecule appears
15
to be oriented from the wider to the narrower rim of the αCD ring (Fig. 8A). In the GM of
16
CA/βCD (molar ratio=1/1), cross peaks were produced by H-3 (3.67 ppm), H-5 (3.53, 3.54
17
ppm), and H-6 (3.60 ppm) in the cavity, by H-a (6.02, 6.05 ppm) and H-b (7.31, 7.33 ppm) in
18
the vinylene group of CA, and by H-c (6.80, 6.81 ppm), H-d (6.68, 6.69 ppm), and H-e (6.88
19
ppm) in the aromatic ring (Fig. 7B). In the GM of CA/βCD, intense cross peaks were
20
produced by H-b in the vinylene group of CA, by H-e in the aromatic ring, and by H-5 in the
21
cavity of βCD. No cross peaks were produced by H-c or H-d in the benzene ring or by H-3 in
22
the cavity of βCD. Thus, the aromatic ring of the CA molecule appears to be oriented from
23
the wider to the narrower rim of the ring of βCD (Fig. 8B). The results of the FTIR revealing
24
an interaction between the aromatic ring and vinylene group of the CA and the CDs has been
25
confirmed. This result suggests that the inclusion complex of the GM CA/CDs also involves
26
interactions of the aromatic ring and vinylene group of CA similar to that observed in 14 Page 14 of 30
1
aqueous solution. In the GM of CA/αCD (molar ratio=1/1), the aromatic ring was partially
2
included, while the aromatic ring was completely included in the GM of CA/βCD (molar
3
ratio=1/1). In the CA/γCD complex, there was a strong cross peak between the proton derived
4
from the aromatic ring of the CA molecule and the narrow edge of γCD [18]. Thus, the
5
aromatic ring of the CA molecule appears to be oriented from the wider to the narrower rim
6
of the ring of γCD. This result suggests that the GM of CA/αCD has a different inclusion
7
form compared to that of CA/βCD and CA/γCD. (Fig. 7), (Fig. 8)
8 9 10
3.7. DPPH radical scavenging test
11 12
It has been reported that inclusion complex formation increases antioxidant capacity [17].
13
The presence of a guest molecule in a CD molecule results in increased electron density,
14
resulting in a greater likelihood that the guest molecule will release protons as radicals to
15
quench the DPPH radical. Thus, a DPPH radical-scavenging activity test was conducted to
16
assess antioxidant capacity in the GM of CA/αCD (molar ratio=1/1) and the GM of CA/βCD
17
(molar ratio=1/1). The results of this test indicate that the 50% inhibitory concentration (IC50)
18
of DPPH radical scavenging was 2.57 µg/ml for intact CA, 1.42 µg/ml for the GM of
19
CA/αCD, and 1.77 µg/ml for the GM of CA/βCD (Fig. 9). The IC50 was significantly lower
20
for the GMs of CA/αCD and CA/βCD (both molar ratios of 1/1) compared to that of intact
21
CA, and increased antioxidant capacity was noted. The GM of CA/αCD (molar ratio=1/1)
22
tended to have higher antioxidant capacity. This might be explained by the stability constant
23
in aqueous solution affecting the antioxidant capacity; the results of solubility testing
24
indicated that CA/αCD had a higher stability constant than CA/βCD did [30]. 1H-1H NOESY
25
NMR spectra revealed that the form of the inclusion complex in the GM of CA/αCD (molar
26
ratio=1/1) was different from that of the GM of CA/βCD (molar ratio=1/1) (Fig. 7). 15 Page 15 of 30
1
Additionally, βCD has a more hydrophobic cavity than αCD does [31]. Thus, the form of the
2
CA inclusion in the GM of CA/βCD involves capture of the aromatic ring. In the GM of
3
CA/αCD, the inclusion primarily involves the vinylene group and not the aromatic ring. The
4
antioxidant capacity of CA is due to the hydroxyl groups attached to the aromatic ring,
5
suggesting that the differences in the orientation of the CA within the CD affected the
6
antioxidant capacity [32]. The GM of the CA/γCD complex did not improve antioxidant
7
capacity [18]. The GM of the CA/γCD complex has a lower stability constant than the GM of
8
the CA/αCD complex does, and the CA of the latter is complexed in a manner that captures
9
the aromatic rings. We believe that these two reasons explain the improved antioxidant
10
capacity of the GM of CA/αCD. (Fig. 9)
11 12 13
4. Conclusion
14
In the current study, phase solubility diagrams indicated that both CA/αCD and CA/βCD
15
form a complex at a molar ratio of 1/1. In addition, the results revealed that CA was
16
accommodated more stably inside the cavity of αCD than inside the cavity of βCD.
17
The results of the DSC, PXRD, FTIR, and dissolution tests revealed that both CA/αCD and
18
CA/βCD could form a 1/1 inclusion complex with an enhanced dissolution rate through the
19
co-grinding method. H- H NOESY NMR spectra revealed that the orientation of the guest
20
molecule inclusion differed depending on the form of CD. The results of a test of DPPH
21
radical-scavenging activity revealed that differences in stability constants and forms of
22
inclusion affected the antioxidant capacity of CA. These findings indicate that the complexes
23
of CA/αCD and CA/βCD have improved solubility and improved antioxidant capacity,
24
suggesting that CA may warrant increased use in functional foods.
1
1
25
16 Page 16 of 30
1 2
Acknowledgements The authors are grateful to Cyclo Chem Co., Ltd. for the provision of αCD and βCD.
3 4 5 6
Conflict of interests This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
7 8
References
9 10
[1] Ministry of Health, Labour and Welfare: General Welfare and Labour (2015). http://www.mhlw.go.jp/english/wp/wp-hw9/dl/01e.pdf
11 12
[2] Ishikawa A, Yamashita H, Hiemori M, et al. Characterization of Inhibitors of postprandial
13
hyperglycemia from the leaves of nerium indicum. J Nutr Sci Vitaminol
14
2007;53:166-173.
15
[3]
Hemmerle H, Burger HJ, Below P, et al. Chlorogenic Acid and Synthetic Chlorogenic
16
Acid Derivatives: Novel Inhibitors of Hepatic Glucose-6-phosphate Translocase. J Med
17
Chem 1997;40(2):137-145.
18
[4]
rescues diet-induced insulin resistance and obesity in mice. Nutr Metab 2015;22:12-19.
19 20
[5]
Jung UJ, Lee MK, Park YB, et al. Antihyperglycemic and antioxidant properties of caffeic acid in db/db mice. J Pharmacol Exp Ther 2006;318(2):476-483.
21 22
Ghadieh HE, Smiley ZN, Kopfman MW, et al. Chlorogenic acid/chromium supplement
[6]
Natella F, Nardini M, Belelli F, et al. Coffee drinking induces incorporation of phenolic
23
acids into LDL and increases the resistance of LDL to ex vivo oxidation in humans. Am
24
J Clin Nutr 2007;86(3):604-609.
25 26
[7]
Brewster ME, Loftsson T. Cyclodextrins as pharmaceutical solubilizers. Adv Drug Deliv Rev 2007;59(7):645-666. 17 Page 17 of 30
1
[8]
Inoue Y, Iohara D, Sekiya N, et al. Ternary inclusion complex formation and
2
stabilization of limaprost, a prostaglandin E1 derivative, in the presence of α- and
3
β-cyclodextrins in the solid state. Int J Pharm 2016;509(1-2):338-347.
4
[9]
Arima H, Yunomae K, Miyake K, et al. Comparative studies of the enhancing effects of
5
cyclodextrins on the solubility and oral bioavailability of tacrolimus in rats. J Pharm Sci
6
2001;90(6):690-701.
7
[10] Anselmi C, Centini M, Ricci M, et al. Analytical characterization of a ferulic
8
acid/gamma-cyclodextrin inclusion complex. J Pharm Biomed Anal
9
2005;40(4):875-881.
10
[11] Yáñez C, Cañete-Rosales P, Castillo JP, et al. Cyclodextrin inclusion complex to
11
improve physicochemical properties of herbicide bentazon: exploring better
12
formulations. PLoS One 2012;7(8):e41072.
13 14 15
[12] Wang J, Cao Y, Sun B, et al. Characterisation of inclusion complex of trans-ferulic acid and hydroxypropyl-β-cyclodextrin. Food Chemistry 2011;124:1069-1075. [13] Hu Y, Gniado K, Erxleben A, et al. Mechanochemical Reaction of Sulfathiazole with
16
Carboxylic Acids: Formation of a Cocrystal, a Salt, and Coamorphous Solids Cryst.
17
Growth Des 2014;14(2):803-813.
18
[14] Ezawa T, Inoue Y, Tunvichien S, et al. Changes in the Physicochemical Properties of
19
Piperine/β-Cyclodextrin due to the Formation of Inclusion Complexes. Int J Med Chem
20
2016. doi:10.1155/2016/8723139
21
[15] Ikeda H, Ikuta N, Nakata D, et al. NMR Studies of Inclusion Complexes Formed by
22
(R)-α-Lipoic Acid with α-, β-, and γ-Cyclodextrins. Bulletin of the Chemical Society of
23
Japan 2015;88(8):1123-1127.
24
[16] Pápay Z.E, Sebestyén Z, Ludányi K, et al. Comparative evaluation of the effect of
25
cyclodextrins and pH on aqueous solubility of apigenin. Journal of Pharmaceutical and
26
Biomedical Analysis 2016;117:210-216. 18 Page 18 of 30
1
[17] Chao J, Wang H, Zhao W, et al. Investigation of the inclusion behavior of chlorogenic
2
acid with hydroxypropyl-β cyclodextrin. Int J Biol Macromol 2012;50:277-282.
3
[18] Inoue Y, Suzuki K, Ezawa T, et al. Examination of the physicochemical properties of
4
caffeic acid complexed with γ-cyclodextrin. J Incl Phenom Macrocycl Chem
5
2015;83:289-298.
6 7 8
[19] Higuchi T, Connors KA. Phase-solubility techniques. Adv Anal Chem Instrum 1961;4:117-212. [20] Göktürk S, Çalışkan E, Talman RY, et al. A study on solubilization of poorly soluble
9
drugs by cyclodextrins and micelles:complexation and binding characteristics of
10
sulfamethoxazole and trimethoprim. Scientific World Journal 2012;2012:718791.
11
[21] Iacovino R, Rapuano F, Caso JV, et al. β-Cyclodextrin Inclusion Complex to Improve
12
Physicochemical Properties of Pipemidic Acid: Characterization and Bioactivity
13
Evaluation. Int J Mol Sci 2013;14(7):13022-13041.
14
[22] Suzuki R, Inoue Y, Tsunoda Y, et al. Effect of γ-cyclodextrin derivative complexation
15
on the physicochemical properties and antimicrobial activity of hinokitiol. J Incl
16
Phenom Macrocycl Chem 2015;83:177-186.
17
[23] Iwata M, Fukami T, Kawashima D, et al. Effectiveness of mechanochemical treatment
18
with cyclodextrins on increasing solubility of glimepiride. Pharmazie
19
2009;64(6):390-394.
20
[24] Xing Y, Peng HY, Zhang MX, et al. Caffeic acid product from the highly
21
copper-tolerant plant Elsholtzia splendens post-phytoremediation: its extraction,
22
purification,and identification. J Zhejiang Univ Sci B 2012;13:487-493.
23 24 25 26
[25] Higashi K, Tozuka Y, Moribe K, et al. Salicylic Acid/γ-Cyclodextrin 2:1 and 4:1 Complex Formation by Sealed-Heating Method. J Pharm 2010;10:4192-4200. [26] Trapani G, Latrofa A, Franco M, et al. Complexation of zolpidem with 2-hydroxypropyl-beta-, methyl-beta-, and 2-hydroxypropyl-gamma-cyclodextrin: effect 19 Page 19 of 30
1
on aqueous solubility, dissolution rate, and ataxic activity in rat. J Pharm Sci
2
2000;89(11):1443-1451.
3
[27] Corciova A, Ciobanu C, Poiata A, et al. Antibacterial and antioxidant properties of
4
hesperidin:β-cyclodextrin complexes obtained by different techniques. J Incl Phenom
5
Macrocycl Chem 2015;81:71-84.
6
[28] Inoue Y, Watanabe S, Suzuki R, et al. Evaluation of actarit/γ-cyclodextrin complex
7
prepared by different methods. J Incl Phenom Macrocycl Chem 2015;81:161-168.
8
[29] Kirschner DL, Green TK. Nonaqueous synthesis of a selectively modified, highly
9
anionic sulfopropyl ether derivative of cyclomaltoheptaose (β-cyclodextrin) in the
10
presence of 18-crown-6. Carbohydrate Research 2005;340:1773-1779.
11
[30] Aytac Z, Uyar T. Antioxidant activity and photostability of α-tocopherol/β-cyclodextrin
12
inclusion complex encapsulated electrospun polycaprolactone nanofibers. European
13
Polymer Journal 2016;79:140-149.
14 15 16
[31] Naidoo KJ, Chen JY, Jansson JLM, et al. Molecular Properties Related to the Anomalous Solubility of β-Cyclodextrin. J Phys Chem B 2004;108(14):4236-4238 [32] Kfoury M, Landy D, Auezova L, et al. Effect of cyclodextrin complexation on
17
phenylpropanoids’ solubility and antioxidant activity. Beilstein J Org Chem
18
2014;10:2322-2331.
19 20
Figure and Table legends
21 22
Figures:
23 24 25
A)
OH
26 20
O
OH OH
Page 20 of 30
1 2 3 4 5 6 7 8 9 10 11 12
Fig. 1. Chemical structure.
13
(A) Caffeic acid (CA), (B-1) αCyclodextrin (αCD), (B-2) βCyclodextrin (βCD)
14
21 Page 21 of 30
1 2 3 4 5 6
40
7 8 9
30
10
12 13 14
CA (mM)
11
20
15
αCD
16 17
10
βCD
18 19 20 21 22 23 24
0 0
10
20 CDs (mM)
30
40
Fig. 2. Phase solubility diagram of CA/CDs. Results are expressed as means ±S.D. (n = 3).
25
22 Page 22 of 30
1 2
(A)
3 4
223
5 6
(B) 210
7 8
(C
207
9 10 11 12
(D) Endothermic (E)
13 14
(F)
15
216
16 17
(G) 197
18 19
(H
20 21
(I)
22 23 24 25
100
200
150
250
Temperature(ºC)
26
Fig. 3. DSC curves of CA/CD systems.
27
(A) CA intact, (B) PM (CA/αCD=1/1), (C) GM (CA/αCD=2/1), (D) GM (CA/αCD=1/1), (E) GM (CA/αCD=1/2), (F) PM (CA/βCD=1/1), (G) GM (CA/βCD=2/1), (H) GM (CA/βCD=1/1), (I) GM (CA/βCD=1/2)
28 29 30
23 Page 23 of 30
1 2 3 4 5 6 (A)
7 8 9
(B)
10 11 12
(I)
(C)
13
16 17 18
Intensity (cps)
15
Intensity (cps)
14 (D)
(E)
19
(J)
(K)
(F)
(L)
(G)
(M)
(H)
(N)
20 21 22 23 24 25 26 27 28 29 30 31 32
5
10
15
20
25
30
5
2θ degree
10
15
20
25
30
2θ degree
Fig. 4. PXRD patterns of CA/CDs. (A) CA intact, (B) CA ground, (C) αCD, (D) αCD ground, (E) PM (CA/αCD=1/1), (F) GM (CA/αCD=2/1), (G) GM (CA/αCD=1/1), (H) GM (CA/αCD=1/2), (I) βCD, (J) βCD ground, (K) PM (CA/βCD=1/1), (L) GM (CA/βCD=2/1), (M) GM (CA/βCD=1/1), (N) GM (CA/βCD=1/2)
33
24 Page 24 of 30
1 2 3 4
(A)
5
22
3433
6
44
7
(B)
8 9
(C)
10 11 12 13
15 16 17 18
Transmission (%)
14
(D) 22 46
(E)
3433
22 46
19
3433
20
(F)
21
85
13 44
22 3433
23
(G)
24 25
13
85
26
44
3433
27 28 29 30 31 32
3700
3000
2600 1850 Wavenumber(cm-1 )
1600 1500
1300
Fig. 5. FTIR spectra of CA/CD systems. (A) CA intact, (B) αCD, (C) βCD, (D) PM (CA/αCD=1/1), (E) PM (CA/βCD=1/1), (F) GM (CA/αCD=1/1), (G) GM (CA/βCD=1/1)
33
25 Page 25 of 30
1 2
Percentage dissolved (%)
100
80
60
40
CA intact CA ground GM(CA/αCD=1/1)
20
GM(CA/βCD=1/1) PM(CA/αCD=1/1) PM(CA/βCD=1/1)
0 0
30
60
90
120
Time (min) 3 4
Fig. 6. Dissolution profiles of CA/CD systems. Results are expressed as means ±S.D. (n = 3).
5
26 Page 26 of 30
1 2 3 c
4 O
5 6 7
d
a OH
b
e
4
OH OH
6
OH O
5 2
HO
3
e
O
OH
b
1
c d
Glucopyranose
Caffeic acid
a
(A)
8
4 2
9 10 11 12
5 6
13
3
14 15 16 17
e
18
b
cd
a
19 20
(B) 4 2
21 22 23
5
24 25
6
26
3
27 28 29 1
30 31
1
Fig. 7. H- H NOESY NMR spectra of CA/CD systems. (A) GM (CA/αCD=1/1), (B) GM (CA/βCD=1/1)
32 33
27 Page 27 of 30
1 2 3 4 5 6
(A)
(B)
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Fig. 8. Proposed structural images of CA/CDs complex. (A) side view of GM (CA/αCD), (B) side view of GM (CA/βCD)
23
28 Page 28 of 30
1 2 3
＊
4
＊
5
3.0
6
2.57
7 8
2.5 2.05
9
2.03
10
2.0
12 13 14
IC50 (μg/ml)
11
1.77 1.42
1.5
15 16
1.0
17 18
0.5
19 20 21
0.0
22 23 24 25 26
Fig. 9. IC50 of DPPH radical scavenging test of CA/CD systems.
27
Results are expressed as means ± S.D. (n = 3) ＊ :P<0.05 vs CA (Tukey’s test).
28
29 Page 29 of 30
1
Table 1 Apparent stability constant (Ks) of CA inclusion complexes. CA complex
Ks ± S.D. (M-1)
αCD
1547.5±30.9
βCD
371.4±6.22
γCD
391.2±25.8
2 3
30 Page 30 of 30