Accepted Manuscript Title: PREPARATION AND CHARACTERIZATION OF QUERCETIN-LOADED LIPID LIQUID CRYSTALLINE SYSTEMS Author: A. Linkeviˇci¯ut˙e A. Misi¯unas E. Naujalis J. Barauskas PII: DOI: Reference:
S0927-7765(15)00076-4 http://dx.doi.org/doi:10.1016/j.colsurfb.2015.02.001 COLSUB 6891
To appear in:
Colloids and Surfaces B: Biointerfaces
Received date: Revised date: Accepted date:
15-10-2014 22-1-2015 1-2-2015
Please cite this article as: A. Linkeviˇci¯ut˙e, A. Misi¯unas, E. Naujalis, J. Barauskas, PREPARATION AND CHARACTERIZATION OF QUERCETIN-LOADED LIPID LIQUID CRYSTALLINE SYSTEMS, Colloids and Surfaces B: Biointerfaces (2015), http://dx.doi.org/10.1016/j.colsurfb.2015.02.001 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
PREPARATION AND CHARACTERIZATION OF QUERCETIN-LOADED
2
LIPID LIQUID CRYSTALLINE SYSTEMS
3
A. Linkevičiūtė1,2,*, A. Misiūnas2, E. Naujalis1,2, J. Barauskas3
5
1
6
LT-03225 Vilnius, Lithuania
7
2
9
cr
Department of Analytical and Environmental Chemistry, Vilnius University, Naugarduko 24,
State Research Institute, Center for Physical Sciences and Technology, A. Goštauto 11,
LT-01108 Vilnius, Lithuania 3
us
8
ip t
4
Biomedical Science, Faculty of Health and Society, Malmö University, SE-20506 Malmö,
Sweden
11
E-mail addresses: [email protected], [email protected], [email protected],
12
[email protected]
an
10
M
13
*Corresponding author: Ausra Linkeviciute, State Research Institute, Center for Physical
15
Sciences and Technology, A. Gostauto 11, LT-01108 Vilnius, Lithuania. Tel.: +370 5 2612758;
16
Fax. +370 5 2627123. E-mail addresses: [email protected]
17
te
d
14
Abstract
19
The aim of the present study was to investigate mixtures of soy phosphatidylcholine (SPC) and
20
glycerol dioleate (GDO) as encapsulation matrices for antioxidant quercetin. The effects of
21
quercetin loading into non-aqueous formulations, non-lamellar liquid crystalline phases and their
22
colloidal dispersions were studied by using synchrotron small angle X-ray diffraction, dynamic
23
light scattering, cryogenic electron microscopy and high performance liquid chromatography.
24
Quercetin incorporation is discussed in the context of lipid aggregation behavior, self-assembled
25
nanostructure and chemical stability. The obtained results show that SPC/GDO-based
26
formulations can incorporate relatively high amounts of quercetin and serve as liquid crystalline
27
delivery vehicles in the form of bulk phases or colloidal dispersions.
Ac ce p
18
28 29
Keywords:
Page 1 of 27
30
Quercetin, lipid liquid crystals, non-lamellar phases, lipid liquid crystalline nanoparticles, small
31
angle X-ray diffraction, cryo-TEM, HPLC.
Ac ce p
te
d
M
an
us
cr
ip t
32
Page 2 of 27
32
1.
Introduction
34
Flavonoids belong to the large and diverse group of polyphenolic compounds characterized by a
35
common benzo-γ-pyrone structure and are ubiquitously present in plants [1, 2]. During recent
36
years, these compounds are of growing interest due to broad pharmacological features, including
37
anti-thrombotic, anti-inflammatory, anticancer and immunostimulator activities [3-5]. Moreover,
38
in the human diet quercetin (QU) (3,3´,4´,5,7-pentahydroxyflavone) is one of the most abundant
39
natural flavonoid, which antioxidant activity is higher than for the other well-known antioxidant
40
molecules, i.e. ascorbylpalmitate, trolox and rutin [6, 7]. Besides antioxidant properties,
41
quercetin also possesses anticancer, antiviral and antiallergic features [8, 9]. The anticancer
42
properties of this natural antioxidant have been proved by in vivo and in vitro experiments which
43
demonstrated that QU has a significant role in inhibition of breast, colon, prostate, ovary,
44
endometrium, and lung tumor cancer cells [10-12].
45
The main limitation in the therapeutic employment of QU is its low water solubility and
46
instability in physiological medium which restricts its use to oral route of administration [13]. In
47
order to extend clinical applications and to solve the solubility, instability, and bioavailability
48
issues of QU it is necessary to develop an appropriate flavonoid delivery system by
49
entrapping/adsorbing QU into the more favorable environment. A variety of different QU
50
immobilization/incorporation strategies, such as using emulsions [14-16], nanosuspensions [17],
51
microcapsules [18, 19], solid lipid nanoparticles [20-24], have been suggested to overcome at
52
least some of these limitations. Much attention has been also made to a various types
53
phospholipid-based vesicles and liposomes as potential vehicles for entrapment and delivery of
54
QU [3, 25-30]. An advantage to use phospholipids is their amphiphilic nature that can modify the
55
solubility behavior and rate of drug release and enhance drug absorption across the biological
56
barriers [3, 31, 32]. However, complex preparation routines and low encapsulation capability of
57
poor water soluble molecules restricts the potential application of vesicles and liposomes for the
58
entrapment and delivery of QU.
59
It was known for years that many phospholipids, such as phosphatidylcholines, spontaneously
60
form flat lipid bilayers and fragmented vesicular particles in aqueous dispersions. It was early on
61
also observed that a variety of polar lipids and their mixtures can self-assemble into different
Ac ce p
te
d
M
an
us
cr
ip t
33
Page 3 of 27
non-lamellar liquid crystalline (LC) phases in water under solution conditions similar to that of
63
biological systems [33]. Non-lamellar LC phases, such as sponge, hexagonal, and cubic,
64
typically comprise of hydrophilic and hydrophobic, which may be continuous or discrete, thus
65
forming either mono- or bicontinuous networks, depending on the molecular nature of the lipid
66
or lipid mixture [34]. They have generally much higher surface area per volume than lamellar
67
structures, may better solubilize hydrophobic, hydrophilic and also amphiphilic molecules [35-
68
37], and can be used as delivery systems for peptides, proteins and food bioactives [38, 39].
69
Recently, a versatile drug delivery lipid system based on soy phosphatidylcholine (SPC) and
70
glycerol dioleate (GDO) has been developed. It offers optimal functional properties, such as
71
bioadhesion and controlled release, good drug loading ability, enhanced flexibility of self-
72
assembly that can be tuned in the range from lamellar to various reversed non-lamellar LC
73
phases [40, 41]. Furthermore, in the presence of polymeric particle stabilizer Polysorbate 80
74
(P80) SPC/GDO mixtures can be easily dispersed into non-lamellar liquid crystalline
75
nanoparticles (LCNPs) with controllable small size and inner morphology making such
76
nanoparticles suitable for parenteral administration routes [42-44].
77
Here we investigated the potential of SPC/GDO-based non-aqueous lipid formulations, non-
78
lamellar LC phases and LCNPs as delivery vehicles of QU. To our best knowledge, only one
79
study of QU entrapment into non-lamellar lipid-based LC system has been previously reported.
80
Murgia and co-workers [45] have been studied the incorporation of QU into glycerol monooleate
81
(GMO)-based bicontinuous cubic LCNPs and demonstrated that small amounts of QU does not
82
alter size, charge and inner nanostructure of the nanoparticles. In this study several mono-, di-
83
glyceride and SPC/GDO-based non-lamellar LC phase forming lipid compositions were explored
84
in the context of their ability to solubilize QU. The main objective of the study was to elucidate
85
the effects of QU solubilization on the nanostructure of lipid-based non-lamellar LC phases and
86
their LCNPs as a function of lipid composition and QU concentration by means of polarized
87
light microscopy, synchrotron small-angle X-ray diffraction (SAXD), dynamic light scattering
88
(DLS) and cryogenic transmission electron microscopy (cryo-TEM). Finally, chemical stability
89
of entrapped QU in different LC systems has been evaluated by using HPLC.
Ac ce p
te
d
M
an
us
cr
ip t
62
90
Page 4 of 27
90 91
2.
Materials and methods
2.1.
Materials
93
ip t
92
94
Quercetin (QU) (CAS-No.117-39-5; purity ≥ 98.0%) was purchased from Sigma-Aldrich.
96
Diglycerolmonooleate (DGMO) (Rylo PG 29) and glycerol dioleate (GDO) (Rylo PG 19) from
97
Danisco, commercial mixture of mono-, di- and triglycerides denoted as Capmul GMO-50 from
98
Abitec, polysorbate P80 (P80) from Croda, glycerylmonooleyl ether (GME) from Niko
99
Chemicals, and soy phosphatidylcholine (SPC) denoted as SPC S100 from Lipoid were used as
100
received. Methanol (MeOH) and acetonitrile (ACN) was purchased from Sigma-Aldrich, 99.7 %
101
ethanol (EtOH) from either Merck or Solveco, and formic acid from Scharlau Chemie. All
102
chemicals were of analytical grade. The water used was passed through the NANOpure Infinity
103
(Branstead) water purification system.
104
2.2.
Preparation of lipid-based formulations
107
Non-aqueous lipid formulations were prepared by mixing appropriate amounts of lipid
108
components (DGMO, GDO, Campul GMO-50, GME and SPC) without and in the presence of
109
co-solvent ethanol to facilitate mixing. Lipid mixtures were then placed on a roller mixer at room
110
temperature for 24 h until mixed completely. Prepared lipid formulations were kept at room
111
temperature (RT) until further use. QU containing formulations were prepared by weighing
112
appropriate amounts of QU and non-aqueous lipid formulations and placing on a roller mixer for
113
24–48 h at RT until homogeneous mixture was obtained. Visual inspection between cross-
114
polarizers and light microscopy were used to inspect samples for the presence of undissolved
115
QU. In this manuscript, the concentration of QU is always expressed as wt% of the total
116
formulation weight. Component ratios of lipid formulations are also always expressed in wt%.
Ac ce p
te
106
d
105
M
an
us
cr
95
117 118
2.3.
Preparation of bulk lipid liquid crystalline (LC) phases
119
Page 5 of 27
120
LC phases were prepared by injection of QU containing non-aqueous lipid formulations (about
121
300 mg) into water at the formulation/water weight ratio of 5/95. Samples were immediately
122
sealed and left to equilibrate at RT in still standing vials for at least 10 days before experiments.
123
2.4.
Preparation of lipid liquid crystalline nanoparticle (LCNP) dispersions
ip t
124 125
Non-aqueous lipid formulations for LCNP preparation were prepared by mixing appropriate
127
amounts of lipids (SPC/GDO = 35/65, 60/40 wt%), polymeric particle stabilizer P80, and ethanol
128
at fixed (SPC+GDO)/P80/EtOH weight ratio of 75/15/10. QU containing formulations were
129
prepared by weighing appropriate amounts of QU and mixing with non-aqueous formulations at
130
QU concentration of 0.5, 1.0, 2.0 and 4.0 wt%. Lipid mixtures were placed on a roller mixer for
131
24 h until mixing was complete, and were then dispersed in 82.5 wt% water (17.5 wt% LCNPs).
132
The aqueous dispersions were immediately sealed, shaken, and left to vortex for 72 h on a
133
mechanical mixing table at 300 rpm at RT. Prepared dispersions were stored at RT until further
134
use.
M
an
us
cr
126
136
2.5.
Polarizing light microscopy (PLM)
te
137
d
135
The texture and temperature-induced phase transitions of LC phases were examined by a
139
polarizing microscope Optiphot equipped with a digital camera DS-2Mv (Nikon) and a heating
140
table Analysa LTS350 (Linkam). A small specimen of sample was placed between two
141
microscope slides and their edges were immediately sealed with a thermo stable silicon grease to
142
prevent evaporation of water from the specimen. A stepwise increase of temperature (typically
143
5 °C in every step with a heating rate of 1 °C/min) was used to induce phase transitions. QU
144
loaded lipid LCNPs were examined at RT by a polarizing microscope DM 750 (Leica) equipped
145
with a digital camera MC170 (Leica).
Ac ce p
138
146 147
2.6.
Small-angle X-ray diffraction (SAXD)
148 149 150
The nanostructure of LC phases was studied using synchrotron SAXD measurements, which were performed at the I911-4beamline at MAX-lab (Lund University, Sweden), using a 1M
Page 6 of 27
PILATUS 2D detector containing a total of 981 x 1043 pixels. Bulk lipid LC samples were
152
mounted between kapton windows in a steel sample holder at the sample to detector distance of
153
1917 mm. Diffractograms were recorded with a wavelength of 0.91 Å and the beam size of 0.25
154
× 0.25 mm (full width at the half-maximum) at the sample. Temperature control within 0.1 °C
155
was achieved using computer controlled Julabo heating circulator F12-MC (Julabo Labortechnik
156
GMBH, Seelbach, Germany). The experiments were performed successively at 25, 35, 45, 55,
157
and 65 °C with a 60 s exposure time at each temperature and a wait of 10 minutes between
158
temperature steps. The resulting CCD images were integrated and analyzed using the Fit2D
159
software provided by Dr. A. Hammersley [http://www.esrf.fr/computing/scientific/FIT2D].
160
Silver behenate calibrated sample-to-detector distance and detector positions were used.
us
cr
ip t
151
162
2.7.
an
161
Particle size and zeta potential
163
LCNP particle size distributions and zeta potentials were measured using Zetasizer Nano ZS
165
analyzer from Malvern Instruments. For particle size distributions the disposable cuvette filled
166
with 1 mL of LCNP dispersion, which was first diluted to 99.5 wt% of water. The obtained data
167
were averaged from 30 measurements (10 s each). The refractive indices used for lipid particles
168
and water were 1.48 and 1.33, respectively. The particle size distributions were reported as
169
intensity-averaged. The surface charge of the particles was measured using disposable zeta cells
170
filled with 1 mL of LCNP dispersion which was first diluted to 99.5 wt% of water. The zeta
171
potential was calculated using the Smoluchowski approximation for dispersion in water with
172
viscosity of 0.8872 cP, refractive index of 1.33, and dielectric constant of 78.5.
173 174 175
2.8.
Ac ce p
te
d
M
164
Cryogenic transmission electron microscopy (Cryo-TEM)
176
Lipid LCNP dispersions for electron microscopy were prepared in a controlled environment
177
vitrification system to ensure stable temperature and to avoid loss of water during sample
178
preparation. The climate chamber temperature was kept at 25−28 °C, and the relative humidity
179
was kept close to saturation to prevent sample evaporation. The samples were prepared by
180
placing 5 μL of LCNP dispersion on lacey carbon filmed copper grids and gently blotted with
181
filter paper to obtain a thin liquid film (20−400 nm) on the grid. Immediately after blotting, the
Page 7 of 27
grids were rapidly plunged into the liquid ethane at −180 °C to vitrify the water-rich samples to
183
prevent ice crystal formation and to preserve the internal crystalline structure. The vitrified
184
specimens were stored in liquid nitrogen (−196 °C) until measurements. An Oxford CT3500
185
cryo-holder and its work station were used to transfer the samples into the electron microscope
186
(Philips CM120 BioTWINCryo) equipped with a post-column energy filter (Gatan GIF100). The
187
acceleration voltage was 120 kV, and the working temperature was kept below −180 °C. The
188
images were recorded digitally with a CCD camera under low electron dose conditions.
cr
ip t
182
189
2.9.
QU stability study
us
190 191
The chemical stability of QU entrapped in different lipid non-aqueous formulations, LC phases
193
and LCNP dispersions were monitored for 3 months. During stability study all samples were
194
kept at RT in the darkness. For the evaluation of the residual QU concentration a portion of the
195
sample (about 10-15 mg) was collected at predetermined time points (1, 15, 30, 60 and 90 days),
196
dissolved in ACN:MeOH (1:1 v/v) solvent mixture at the lipid sample to solvent weight ratio of
197
1:100 and immediately analyzed using HPLC. Each sample was analyzed in triplicate.
199 200
2.10.
High performance liquid chromatography (HPLC)
te
198
d
M
an
192
QU concentration was determined using an Agilent HPLC 1100 Series (Agilent Technologies,
202
USA) chromatography system equipped with a quaternary pump, a vacuum degasser module, a
203
manual injector with a 20 µL sample loop, a temperature controlled column compartment and a
204
diode array detector (DAD) set at 370 nm. Chromatographic separation was achieved using a
205
Zorbax Eclipse XDB–C18 (5 µm, 150×4.60 mm, Agilent Technologies, USA) reversed phase
206
column coupled with an Eclipse XDB–C18 guard column (5 µm, 12.6×4.6 mm, Agilent
207
Technologies, USA). Determination was carried out in a solvent system of methanol–formic
208
acid–water as previously described, with minor modifications [46]. Data were collected and
209
processed using a ChemStation software version B.01.03. The obtained values with standard
210
methanolic QU solutions showed linearity over the concentration range of 0.1–100 μg/g with a
211
correlation coefficient (r2) of 0.999. The quantification limit in the HPLC assay was 0.1 μg/g and
212
standard deviation under repeatability conditions was no more than 5.6% in all concentrations
Ac ce p
201
Page 8 of 27
213
tested. The calibration curves for quantification of QU in QU-loaded non-aqueous lipid
214
formulations, LC phases and LCNP dispersions were prepared using QU standard with
215
respective lipid compositions. The obtained results of the quantification of various formulations
216
are summarized in Table S1 (Supporting information).
Ac ce p
te
d
M
an
us
cr
ip t
217
Page 9 of 27
217 218
3.
Results and discussions
3.1.
Solubility of QU in lipid formulations
220
ip t
219
221
It is known that some poorly water soluble compounds like QU may also have low solubility in
223
lipid excipients due to different physical chemical reasons [31]. As shown in other lipid- and
224
amphiphile-based systems, such as microemulsions [19], liposomes [26], lecithin-chitosan
225
nanoparticles [23] and solid lipid nanoparticles [24], the maximum solubility of QU is quite low
226
and often is in the order of 0.5 wt%, which can be somewhat improved by using additional
227
polymeric surfactants and/or co-solvents [16].
228
Therefore, before further experiments several non-lamellar LC phase forming non-aqueous lipid
229
mixtures were investigated with respect to their ability to solubilize QU. First, binary mixtures of
230
DGMO/GDO (85/15 and 60/40) (here and everywhere in the text formulation compositions are
231
expressed in wt%), DGMO/Capmul GMO-50 (85/15 and 60/40), DGMO/GME (85/15 and
232
60/40) and SPC/GDO (50/50) were explored. The results have shown that independently from
233
lipid ratio the solubility of QU in all tested mono- and di-glyceride-based mixtures was very low
234
and varies between 0.3 and 0.4 wt%. The addition of up to 10 wt% of ethanol (in respect to total
235
lipid) did not improve the solubility of QU in neither of the formulations. The observed solubility
236
in mixtures of mono- and di-glycerides was in line with previous results in similar systems. The
237
solubility of QU in commercial mixtures of long chain monoglycerides (i.e., Miglyol 812,
238
Capmul MCM, Labrafil 1944) is usually well below 1 wt% [16].
239
Much higher solubility of QU was found for the mixtures of SPC and GDO. Since solvent free
240
SPC/GDO mixtures prepared at equal lipid weight ratio were difficult to mix and were very
241
viscous their ability to solubilize QU were only tested in the presence of small amount of ethanol
242
at fixed Lipid/EtOH weight ratio of 90/10. SPC/GDO composition in the formulations was
243
varied between 60/40 and 35/65. Independently on lipid ratio about 5 wt% of QU was soluble in
244
all formulations. From the solubility test it may be concluded that lipid mixtures composed of
245
exclusively mono- and di-glycerides with and without EtOH have very limited ability to
246
solubilize QU which may be increased by about 10-20 times with the introduction of SPC into
Ac ce p
te
d
M
an
us
cr
222
Page 10 of 27
247
the mixture. Considering these results, SPC/GDO-based formulations containing up to 4 wt% of
248
QU and 10 wt% EtOH were selected for further studies.
249 250
3.2.
SPC/GDO-based LC phases with entrapped QU
ip t
251
The aim of this part was to investigate the effects of QU on the aggregation behavior of fully
253
hydrated non-lamellar bulk LC phases of SPC/GDO. As shown in recent study, the aqueous
254
phase behavior of mixtures of SPC and GDO is rather complex [40]. At 25 °C with increasing
255
GDO content fully hydrated SPC/GDO mixtures in water form the following LC phase
256
sequence: lamellar (Lα) reversed 2D hexagonal (H2, up to 62.5/37.5) reversed micellar
257
cubic of Fd m space group (50/50 – 45/55) reversed 3D hexagonal of P63/mmc space group
258
(42/58 – 40/60) unresolved “intermediate” (39/61 – 37/63) Fd m (35/65 – 22.5/77.5)
259
reversed micellar solution (L2, from 20/80). In this study, the effects of QU on the nanostructure
260
of different LC phases were investigated at four fixed SPC/GDO weight ratios (60/40, 50/50,
261
40/60 and 35/65) and four temperatures (25, 35, 45 and 55 °C).
262
Figure 1 shows SAXD data of fully hydrated SPC/GDO (60/40 and 35/65) LC phases with
263
entrapped 0.0, 0.5, 1 and 4 wt% of QU as a function of temperature.
264
As seen from Figure 1a, at weight ratio of 60/40 and at 25 °C SPC/GDO forms H2 LC phase,
265
which features three strong Bragg reflections positioned in ratios 1: 3: 4. At 45 °C another
266
coexisting LC phase, most likely Fd m cubic, also starts to form what is evidenced by the
267
appearance of additional peaks. However, independently on temperature the appearance of all
268
diffractograms is practically unaffected by the presence of up to 4 wt% of QU (Figures 1b-1d).
269
In addition, the calculated lattice parameter (a) for this LC phase is constant regardless QU
270
concentration and temperature and only slightly varies between 6.6 and 6.7 nm. This shows that
271
H2 phase prepared at high SPC content is quite robust and can accommodate relatively high
272
amounts of hydrophobic QU without changing nanostructure.
273
At weight ratio of 35/65 and 25 °C SPC/GDO self-assembles into ordered reversed micellar
274
Fd m cubic phase which is clearly characterized by the appearance of first 9 Bragg peaks
275
located at relative positions in ratios 3: 8: 11: 12: 16: 19: 24: 27: 32 (Figure 1e).
276
Contrary to H2 phase, the structure of Fd m cubic phase prepared at low SPC content is rather
277
sensitive to QU entrapment and temperature. At 45 °C it starts to “melt” and at 55 °C completely
Ac ce p
te
d
M
an
us
cr
252
Page 11 of 27
transforms into unordered reversed micelles (L2), which is corroborated by the disappearance of
279
the Bragg diffraction peaks and their smearing into two broad diffuse diffraction features. The
280
entrapment of QU even further disturbs ordered structure of the Fd m cubic phase. At 0.5 and 1
281
wt% of QU ordered cubic phase is almost fully transformed into unordered phase at 45 °C
282
(Figures 1f and 1g). Moreover, only a fraction of the Fd m cubic arrangement with entrapped 4
283
wt% of QU is present already at 35 °C (Figure 1h). As shown in Figure S1 (SI), both the
284
entrapment of QU and temperature considerably enlarge the unit cell dimensions of the cubic LC
285
phase at the SPC/GDO ratio of 35/65. With increasing QU concentration from 0 to 4 wt% the
286
calculated a value increases from 14.9 to 15.6 nm at 25 °C. At elevated temperatures a further
287
increases and reaches maximum value of about 15.9 nm at 1 wt% of QU and 45 °C before
288
complete transformation of ordered Fd m structure into unordered L2 phase.
289
The observed disordering effect of QU may be explained by reasonable assumption that in the
290
LC phase water insoluble hydrophobic QU molecule is preferentially located between lipid
291
hydrocarbon chains. Accommodation of QU in the hydrophobic regions between ordered
292
reversed micelles increases lipid chain packing stress and distance between micelles resulting in
293
slightly larger a values. At some point, further accommodation of QU and unit cell dimension
294
increase is not possible without transformation into unordered L2 phase. At higher temperatures
295
this process occurs at lower QU concentrations due to additional temperature induced lipid chain
296
packing disorder. Similar phase behavior trends were also observed for the SPC/GDO Fd m
297
cubic phases with entrapped benzydamine, lidocaine and granisetron [41].
298
In addition, less extensive but similar disordering effects of QU were also observed for the LC
299
phases prepared at intermediate SPC/GDO weight ratios of 50/50 and 40/60 (Figure S2, SI).
300
Here, the entrapment of QU has little effect on the unit cell dimensions of the LC phases.
301
However, in both cases QU decreases phase transition temperatures of ordered LC phases into L2
302
phase, which is observed for temperatures higher than 45 °C. Here it may be concluded that
303
effect of QU on the nanostructure of the bulk SPC/GDO LC phases is lipid weight ratio-
304
dependent. Thus, at weight ratio of 60/40, H2 LC structure remains unaffected in the presence of
305
QU. In contrast, at weight ratio of 35/65 the structure of the cubic Fd m phase is very sensitive
306
to QU concentration increase and starts transformation into L2 phase already at 35 °C and 4 wt%
307
of QU. Finally, LC phases at the SPC/GDO weight ratios of 50/50 and 40/60 show moderate
308
sensitivity to QU.
Ac ce p
te
d
M
an
us
cr
ip t
278
Page 12 of 27
309 310
3.3.
SPC/GDO/P80-based LCNPs with entrapped QU
311
The aim of this part was to investigate the effects of QU on size and stability of the dispersed
313
SPC/GDO/P80-based liquid crystalline nanoparticles (LCNPs). Dispersions were prepared at
314
SPC/GDO weight ratios of 60/40 and 35/65 in the presence of polymeric stabilizer P80 at fixed
315
lipid to polymer ratio. Dispersions containing up to 4 wt% of QU were prepared and
316
characterized with regards to particle size, charge and morphology.
317
Figure 2 shows obtained particle size distributions as a function of QU concentration. As seen
318
from Figures 2a and 2b, at low QU concentrations all LCNP dispersions are well-defined
319
displaying monomodal size distributions with polydispersity indices ranging between 0.13 and
320
0.16 µm. At 2 and 4 wt% of QU additional larger aggregates of about 5 m appear in the
321
dispersions prepared at both SPC/GDO ratios. Sample examination under polarized light
322
microscope reveals that these larger aggregates are small needle-like crystals phase separated
323
from LCNPs (Figure S3, SI). Since crystals have distinct yellow color (seen in microscope under
324
nonpolarized light), there is good reason to assume that they are formed exclusively by QU. This
325
show that SPC/GDO/P80 LCNPs can homogeneously entrap only up to 2 wt% of QU when
326
compared to 4-5 wt% of QU in the bulk LC phases. This difference may be attributed to a very
327
large surface-to-volume ratio of the particles and exposure to the excess aqueous phase creating
328
more defects and crystallization centers for QU. In addition, one cannot exclude that presence of
329
P80 may also change QU solubilization properties and induce precipitation. Note however, that
330
QU crystallization and phase separation does not affect LCNP own size distribution
331
characteristics. The colloidal stability of the LCNP dispersions with entrapped QU is also good.
332
Only minor changes in the size distributions are observed after 3 months of storage at RT (Figure
333
2, dotted lines).
334
Figure 3 demonstrates that the obtained mean particle size clearly depends on both lipid ratio and
335
QU concentration. Thus, LCNPs at SPC/GDO ratio 60/40 are slightly larger with the mean
336
particle size increasing from 140 to 210 nm as QU concentration is increased from 0 to 4 wt%. In
337
contrast, LCNPs at SPC/GDO ratio 35/65 are smaller with the mean particle size ranging from
338
about 80 to 110 nm. Unfortunately, we cannot explain such an effect of QU on the particle size
339
considering only bulk phase behavior and nanostructural features of the LC phases. Since LCNP
Ac ce p
te
d
M
an
us
cr
ip t
312
Page 13 of 27
dispersions were prepared by mechanical agitation we believe that the entrapment of QU may
341
slightly increase cohesion forces within LC structure and/or influence monocrystalline domain
342
size. Therefore, more energy is required to brake LC phase with entrapped QU into smaller
343
particles. In addition, LCNP zeta potential remains unaffected by the entrapment of QU and is
344
about -16 and -10 mV for particles at SPC/GDO ratio 60/40 and 35/65, respectively. Considering
345
size and surface to volume ratio differences the surface charge density of the particles is similar
346
at both SPC/GDO ratios.
cr
347
ip t
340
Representative cryo-TEM images shown in Figure 4 together with the measured SAXD profiles (Figure S4, SI) give further insights into QU-loaded particle morphology and
349
nanostructure. Overall, the observed particle sizes are consistent with DLS measurements. At
350
SPC/GDO ratio 60/40 particles are larger and characterized by a denser inner core and a less
351
dense shell comprising lamellar and “sponge”-like petals (Figures 4a and 4b). The observed
352
particle morphology is comparable with those previously prepared at similar SPC/GDO weight
353
ratio and P80 content without QU [47]. The obtained LCNP SAXD profiles as a function of QU
354
concentration in Figure S4a (SI) are almost identical showing negligible effect of QU on particle
355
nanostructure and are in line with the bulk phase behavior observations (Figures 1a-1d). The data
356
show only weak diffuse scattering which is not surprising since it is know that dispersion agent
357
P80 has a disordering effect on the internal SPC/GDO LCNP nanostructure [43]. Only small
358
shoulder appearing at q value of about 1.1 nm-1 can be observed which position may be related to
359
the first reflection of the undispersed H2 phase (Figures 1a-1d). Regardless of this relation SAXD
360
data suggest that in dispersed state the long-range order of the H2 LC arrangement in the particle
361
core is almost or completely lost.
362
As shown in Figures 4c and 4d, particles prepared at SPC/GDO 35/65 ratio are smaller in size
363
with more or less even density and hardly any noticeable features in the interior. This is also
364
reflected in the SAXD profiles shown in Figure S4b where only weak scattering pattern is
365
observed without any indications of the Bragg diffraction. From inspection of several dozen of
366
cryo-TEM images it may be also concluded that GDO-rich particles, with entrapped QU do not
367
possess pronounced swollen coronas of multiply connected bilayers which are always present for
368
SPC/GDO LCNPs prepared at high fraction of P80 [47, 48]. The observed difference may be
369
related to the bulk phase behavior where even low concentrations of QU were able to induce
370
Fd m phase transformation towards more reversed L2 phase (Figures 1e-1h). Most likely, in the
Ac ce p
te
d
M
an
us
348
Page 14 of 27
371
dispersed LC state the addition of small amount of QU is enough to allow more homogeneous
372
distribution of components within the particle and to prevent the formation of lamellar-like
373
corona.
374
3.4.
376
LCNPs
QU stability in SPC/GDO non-aqueous formulations, bulk LC phases and dispersed
ip t
375
cr
377
One of the important aspects of delivery formulations is the chemical stability of active
379
substance. Therefore, HPLC was used to evaluate the chemical stability of QU entrapped in all
380
three studied SPC/GDO-based systems: non-aqueous formulations, LC phases and LCNP
381
dispersions. Samples were prepared at two different SPC/GDO weight ratios containing 0.5 and
382
1 wt% of QU in respect to dry formulation weight.
383
For up to 90 days QU showed very good chemical stability when solubilized in non-aqueous
384
SPC/GDO-based formulations (Figures 5a and 5d). About 85–90% of QU is retained
385
independently on lipid composition and QU concentration. Introduction of water and hydration
386
of the formulations into bulk LC phases (Figures 5b and 5e) and LCNP dispersions (Figures 5c
387
and 5f) had more pronounced effect on the stability of QU. Gradual but significant loss of QU
388
content upon storage was observed in both systems. In the LC phases prepared at SPC/GDO ratio
389
35/65 QU content decreased to about 65%, whereas only 25–40% of QU was found after the end
390
of the study in the LC phases prepared at lipid weight ratio of 60/40. Such substantial difference
391
may be explained by the fact that SPC used in this study contain considerable amounts of linoleic
392
(C18:2) fatty acid chains. It is known that degree of unsaturation plays positive role in the lipid
393
autoxidation processes, in which highly reactive primary and secondary lipid oxidation products
394
are produced [49]. In SPC-rich LC phases larger amounts of reactive species may be formed,
395
which interact and chemically degrade QU when compared to the GDO-rich LC phases where
396
oleic fatty acid residues are dominant.
397
The lowest stability of QU was found in the LCNP dispersions where only 30–45% was
398
recovered after 90 days. Here, GDO-rich LCNPs showed only slight improvement over SPC-rich
399
dispersions. Most likely, P80 which was used to stabilize LCNPs dispersions also influence lipid
400
autoxidation complex processes and further decrease stability of QU. On the other hand, water
Ac ce p
te
d
M
an
us
378
Page 15 of 27
penetration ability and diffusion of lipid components is less restricted within small particle when
402
compared to macroscopic bulk LC phase and may accelerate degradation of QU.
403
Overall, QU shows good chemical stability when loaded in SPC/GDO-based non-aqueous
404
formulations for up to 3 months of storage at room temperature. On the other hand, hydrated
405
formulations (bulk LC phases and LCNP dispersions) show reduced QU chemical stability and
406
therefore are not suited for a long-term shelf storage but can still serve as suitable delivery
407
systems if are formed via in situ hydration of non-aqueous formulation (in case of LC phases) or
408
freshly produced just prior application/administration (in case of LCNP dispersions). For real-life
409
application, further studies on parameter optimization (e.g., sterilization, oxygen concentration,
410
temperature, pH, additives and other) in order to extend long-term stability of the formulations
411
will be needed.
412
3.
an
us
cr
ip t
401
Conclusion
M
413
The present study demonstrates the potential of liquid crystalline structures forming SPC/GDO
415
lipid formulations as encapsulation and delivery matrices of QU. The results have shown that up
416
to 5 wt% of QU can be effectively solubilized in non-aqueous SPC/GDO formulations by using
417
small amounts of EtOH as solvent. Prepared formulations are stable and have minimal effect on
418
the chemical stability of QU for few months. Upon hydration non-aqueous formulations can
419
easily self-assemble into non-lamellar bulk LC matrices with different nanostructures. The effect
420
of QU on the nanostructure of LC phases is lipid composition dependent. At high SPC content,
421
the entrapment of QU has practically no effect on the nanostructure of H2 phase at
422
physiologically relevant temperatures. At low SPC content, QU slightly increases the unit cell
423
dimensions of the reversed micellar Fd m cubic phase and promotes the formation of reversed
424
micellar solution at elevated temperatures. Finally, concentrated and colloidally stable non-
425
lamellar SPC/GDO-based LCNP dispersions containing up to 2 wt% of entrapped QU can be
426
easily prepared in the presence of stabilizer P80. Dispersion particle size can be tuned in the
427
range of about 80 – 210 nm by changing lipid composition and entrapped QU concentration.
Ac ce p
te
d
414
428 429
Author’s contributions
430
Page 16 of 27
431
Linkeviciute A. executed the experiment, extracted the data and wrote the initial draft of the
432
manuscript. Barauskas J. supervised the research, analysis of data and interpreted the results
433
obtained from the models. Misiunas A. technical assistance, while Naujalis E. was co-
434
supervising the research.
436
ip t
435
Conflict of interest
438
cr
437
The authors declare that they have no conflict of interest.
us
439
Acknowledgements
441
The authors acknowledge the Swedish synchrotron X-ray facility MAX IV Laboratory for
442
allocated beamtime at the I911-4beamline and Ana Labrador for technical support during
443
experiments. Authors also thank Gunnel Karlsson and Viveka Alfredsson at Lund University for
444
their help with the cryo-TEM imaging. AL acknowledges financial support from the Research
445
Council of Lithuania "Promotion of Student Scientific Activities" (VP1-3.1-ŠMM-01-V-02-003).
M
d te Ac ce p
446
an
440
Page 17 of 27
446 447
FIGURES
448
Figure 1. SAXD profiles of SPC/GDO mixtures in excess water prepared at lipid weight ratios
450
of 60/40 (a-d) and 35/65 (e-h) as a function of QU concentration (0, 0.5, 1.0 and 4.0 wt%) and
451
temperature (25, 35, 45 and 55 °C). Arrows in (e) and (h) show indexing of the reflections from
452
the reversed Fd m cubic phase. An explanation is given in the text.
us
453
cr
ip t
449
Figure 2. Particle size distributions of freshly prepared (solid lines) and 3 months old (dotted
455
lines) SPC/GDO/P80 LCNPs at SPC/GDO weight ratios of 60/40 (a) and 35/65 (b) as a function
456
of QU concentration. Dispersions were prepared at fixed lipid/P80 weight ratio of 75/15 in
457
82.5% water.
M
an
454
458
Figure 3. Dependence of mean particle size of the SPC/GDO/P80 LCNPs on QU concentration.
460
Dispersions were prepared at SPC/GDO weight ratios of 60/40 (open circles) and 35/65 (filled
461
circles). The lines are drawn to guide the eye.
te
Ac ce p
462
d
459
463
Figure 4. Representative cryo-TEM images of SPC/GDO/P80 LCNPs with entrapped 1 wt% QU
464
prepared at SPC/GDO weight ratios of 60/40 (a and b) and 35/65 (c and d). Dispersions were
465
prepared at fixed lipid/P80 weight ratio of 75/15 in 82.5% water.
466 467
Figure 5. Chemical stability of QU (normalized to initial) entrapped in SPC/GDO-based non-
468
aqueous formulations (a and d), LC phases (b and e) and LCNP dispersions (c and f). Samples
469
were prepared at fixed SPC/GDO weight ratios of 60/40 (open symbols) and 35/65 (filled
470
symbols) containing 0.5 (a-c) and 1.0 wt% (d-f) of QU in respect to dry weight. Results are
471
represented as mean values ± standard deviation (n=3).
472
Page 18 of 27
472 473
References
474
te
d
M
an
us
cr
ip t
[1] W. Vermerris and R. Nicholson, Phenolic Compound Biochemistry, Springer, 2006. [2] S. Kumar and A.K. Pandey, Sci World J, (2013). [3] D. Singh, M.S.M. Rawat, A. Semalty and M. Semalty, Curr Drug Deliv, 9 (2012) 305-314. [4] F. Bonina, M. Lanza, L. Montenegro, C. Puglisi, A. Tomaino, D. Trombetta, F. Castelli and A. Saija, Int J Pharmaceut, 145 (1996) 87-94. [5] A.C. Santos, S.A. Uyemura, J.L.C. Lopes, J.N. Bazon, F.E. Mingatto and C. Curti, Free Radical Bio Med, 24 (1998) 1455-1461. [6] M.I. Kaldas, U.K. Walle, H. Van der Woude, J.M. McMillan and T. Walle, J Agr Food Chem, 53 (2005) 4194-4197. [7] N. Nuengchamnong, A. Hermans-Lokkerbol and K. Ingkaninan, Naresuan University Journal, 12 (2004) 25-37. [8] J.V. Formica and W. Regelson, Food Chem Toxicol, 33 (1995) 1061-1080. [9] Y. Zheng, I.S. Haworth, Z. Zuo, M.S.S. Chow and A.H.L. Chow, J Pharm Sci-Us, 94 (2005) 1079-1089. [10] R.K. Hansen, S. Oesterreich, P. Lemieux, K.D. Sarge and S.A.W. Fuqua, Biochem Bioph Res Co, 239 (1997) 851-856. [11] G. Elia, C. Amici, A. Rossi and M.G. Santoro, Cancer Res, 56 (1996) 210-217. [12] M. Koishi, N. Hosokawa, M. Sato, A. Nakai, K. Hirayoshi, M. Hiraoka, M. Abe and K. Nagata, Jpn J Cancer Res, 83 (1992) 1216-1222. [13] T. Pralhad and K. Rajendrakumar, J Pharmaceut Biomed, 34 (2004) 333-339. [14] R. Casagrande, S.R. Georgetti, W.A. Verri, M.F. Borin, R.F.V. Lopez and M.J.V. Fonseca, Int J Pharmaceut, 328 (2007) 183-190. [15] S. Jain, A.K. Jain, M. Pohekar and K. Thanki, Free Radical Bio Med, 65 (2013) 117-130. [16] T.H. Tran, Y. Guo, D.H. Song, R.S. Bruno and X.L. Lu, J Pharm Sci-Us, 103 (2014) 840-852. [17] L. Gao, G.Y. Liu, X.Q. Wang, F. Liu, Y.F. Xu and J. Ma, Int J Pharmaceut, 404 (2011) 231-237. [18] D.H. Lee, G.S. Sim, J.H. Kim, G.S. Lee, H.B. Pyo and B.C. Lee, J Pharm Pharmacol, 59 (2007) 16111620. [19] F.T.M.C. Vicentini, T.R.M. Simi, J.O. Del Ciampo, N.O. Wolga, D.L. Pitol, M.M. Iyomasa, M.V.L.B. Bentley and M.J.V. Fonseca, Eur J Pharm Biopharm, 69 (2008) 948-957. [20] Guo Chen-yua, Yang Chun-fena, Li Qi-luc, Tan Qia, Xi Yan-weia, Liu Wei-nad and Z. Guang-xia, Int J Pharmaceut, 430 (2012) 292– 298. [21] S. Bose and B. Michniak-Kohn, Eur J Pharm Sci, 48 (2013) 442-452. [22] Y.Y. Zhang, Y. Yang, K. Tang, X. Hu and G.L. Zou, J Appl Polym Sci, 107 (2008) 891-897. [23] Q. Tan, W.D. Liu, C.Y. Guo and G.X. Zhai, Int J Nanomed, 6 (2011) 1621-1630. [24] H.L. Li, X.B. Zhao, Y.K. Ma, G.X. Zhai, L.B. Li and H.X. Lou, J Control Release, 133 (2009) 238-244. [25] E. Alexopoulou, A. Georgopoulos, K.A. Kagkadis and C. Demetzos, J Liposome Res, 16 (2006) 17-25. [26] A. Priprem, J. Watanatorn, S. Sutthiparinyanont, W. Phachonpai and S. Muchimapura, NanomedNanotechnol, 4 (2008) 70-78. [27] A.A. Date, M.S. Nagarsenker, S. Patere, V. Dhawan, R.P. Gude, P.A. Hassan, V. Aswal, F. Steiniger, J. Thamm and A. Fahr, Mol Pharmaceut, 8 (2011) 716-726. [28] A.P. Landi-Librandi, T.N. Chrysostomo, A.E.C.S. Azzolini, C.M. Marzocchi-Machado, C.A. de Oliveira and Y.M. Lucisano-Valim, J Liposome Res, 22 (2012) 89-99.
Ac ce p
475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515
Page 19 of 27
545 546 547
te
d
M
an
us
cr
ip t
[29] P.G. Cadena, M.A. Pereira, R.B.S. Cordeiro, I.M.F. Cavalcanti, B.B. Neto, M.D.C.B. Pimentel, J.L. Lima, V.L. Silva and N.S. Santos-Magalhaes, Bba-Biomembranes, 1828 (2013) 309-316. [30] C. Caddeo, O. Diez-Sales, R. Pons, X. Fernandez-Busquets, A.M. Fadda and M. Manconi, Pharm ResDordr, 31 (2014) 959-968. [31] H.L. Mu, R. Holm and A. Mullertz, Int J Pharmaceut, 453 (2013) 215-224. [32] K. Sandeep, M. Mohanvarma and P. Veerabhadhraswamy, Acta Pharmaceutica Sinica B, 3 (2013) 361-372. [33] V. Luzzati, R. Vargas, P. Mariani, A. Gulik and H. Delacroix, J Mol Biol, 229 (1993) 540-551. [34] K. Larsson, P. Quinn, K. Sato and F. Tiberg, Vol. 19 (2006). [35] L. Sagalowicz and M.E. Leser, Curr Opin Colloid In, 15 (2010) 61-72. [36] K. Larsson and F. Tiberg, Curr Opin Colloid In, 9 (2005) 365-369. [37] A. Angelova, B. Angelov, R. Mutafchieva, S. Lesieur and P. Couvreur, Accounts Chem Res, 44 (2011) 147-156. [38] I. Amar-Yuli, D. Libster, A. Aserin and N. Garti, Curr Opin Colloid In, 14 (2009) 21-32. [39] S.B. Rizwan, B.J. Boyd, T. Rades and S. Hook, Expert Opin Drug Del, 7 (2010) 1133-1144. [40] F. Tiberg, M. Johnsson, M. Jankunec and J. Barauskas, Chem Lett, 41 (2012) 1090-1092. [41] J. Barauskas, L. Christerson, M. Wadsater, F. Lindstrom, A.K. Lindqvist and F. Tiberg, Mol Pharmaceut, 11 (2014) 895-903. [42] F. Tiberg and M. Johnsson, J Drug Deliv Sci Tec, 21 (2011) 101-109. [43] M. Wadsater, J. Barauskas, T. Nylander and F. Tiberg, Acs Appl Mater Inter, 6 (2014) 7063-7069. [44] M. Wadsater, J. Barauskas, T. Nylander and F. Tiberga, Soft Matter, 9 (2013) 8815-8819. [45] S. Murgia, S. Bonacchi, A.M. Falch, S. Lampis, V. Lippolis, V. Meli, M. Monduzzi, L. Prodi, J. Schmidt, Y. Talmon and C. Caltagirone, Langmuir, 29 (2013) 6673-6679. [46] A. Linkeviciute, R. Butkiene and E. Naujalis, Chemija, 24 (2013) 217-222. [47] D.P. Chang, M. Jankunec, J. Barauskas, F. Tiberg and T. Nylander, Langmuir, 28 (2012) 10688-10696. [48] D.P. Chang, M. Jankunec, J. Barauskas, F. Tiberg and T. Nylander, Acs Appl Mater Inter, 4 (2012) 2643-2651. [49] A. Morales, C. Dobarganes, G. Marquez-Ruiz and J. Velasco, J Am Oil Chem Soc, 87 (2010) 12711279.
Ac ce p
516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544
Page 20 of 27
547 548 549 550 551
We examine the mixtures of SPC and GDO as encapsulation matrices for quercetin (QU). SPC/GDO-based formulations can incorporate relatively high amounts of QU. At high SPC content, the loaded QU has no effect on the nanostructure of H2 phase. At low SPC content, QU slightly increases the unit cell dimensions of Fd m phase.
Ac ce p
te
d
M
an
us
cr
ip t
552
Page 21 of 27
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 22 of 27
Ac ce p
te
d
M
an
us
cr
ip t
Figure(s)
Page 23 of 27
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 24 of 27
Ac ce p
te
d
M
an
us
cr
ip t
Figure(s)
Page 25 of 27
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 26 of 27
Ac
ce
pt
ed
M
an
us
cr
i
*Graphical Abstract (for review)
Page 27 of 27
S0927-7765(15)00076-4 http://dx.doi.org/doi:10.1016/j.colsurfb.2015.02.001 COLSUB 6891
To appear in:
Colloids and Surfaces B: Biointerfaces
Received date: Revised date: Accepted date:
15-10-2014 22-1-2015 1-2-2015
Please cite this article as: A. Linkeviˇci¯ut˙e, A. Misi¯unas, E. Naujalis, J. Barauskas, PREPARATION AND CHARACTERIZATION OF QUERCETIN-LOADED LIPID LIQUID CRYSTALLINE SYSTEMS, Colloids and Surfaces B: Biointerfaces (2015), http://dx.doi.org/10.1016/j.colsurfb.2015.02.001 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
PREPARATION AND CHARACTERIZATION OF QUERCETIN-LOADED
2
LIPID LIQUID CRYSTALLINE SYSTEMS
3
A. Linkevičiūtė1,2,*, A. Misiūnas2, E. Naujalis1,2, J. Barauskas3
5
1
6
LT-03225 Vilnius, Lithuania
7
2
9
cr
Department of Analytical and Environmental Chemistry, Vilnius University, Naugarduko 24,
State Research Institute, Center for Physical Sciences and Technology, A. Goštauto 11,
LT-01108 Vilnius, Lithuania 3
us
8
ip t
4
Biomedical Science, Faculty of Health and Society, Malmö University, SE-20506 Malmö,
Sweden
11
E-mail addresses: [email protected], [email protected], [email protected],
12
[email protected]
an
10
M
13
*Corresponding author: Ausra Linkeviciute, State Research Institute, Center for Physical
15
Sciences and Technology, A. Gostauto 11, LT-01108 Vilnius, Lithuania. Tel.: +370 5 2612758;
16
Fax. +370 5 2627123. E-mail addresses: [email protected]
17
te
d
14
Abstract
19
The aim of the present study was to investigate mixtures of soy phosphatidylcholine (SPC) and
20
glycerol dioleate (GDO) as encapsulation matrices for antioxidant quercetin. The effects of
21
quercetin loading into non-aqueous formulations, non-lamellar liquid crystalline phases and their
22
colloidal dispersions were studied by using synchrotron small angle X-ray diffraction, dynamic
23
light scattering, cryogenic electron microscopy and high performance liquid chromatography.
24
Quercetin incorporation is discussed in the context of lipid aggregation behavior, self-assembled
25
nanostructure and chemical stability. The obtained results show that SPC/GDO-based
26
formulations can incorporate relatively high amounts of quercetin and serve as liquid crystalline
27
delivery vehicles in the form of bulk phases or colloidal dispersions.
Ac ce p
18
28 29
Keywords:
Page 1 of 27
30
Quercetin, lipid liquid crystals, non-lamellar phases, lipid liquid crystalline nanoparticles, small
31
angle X-ray diffraction, cryo-TEM, HPLC.
Ac ce p
te
d
M
an
us
cr
ip t
32
Page 2 of 27
32
1.
Introduction
34
Flavonoids belong to the large and diverse group of polyphenolic compounds characterized by a
35
common benzo-γ-pyrone structure and are ubiquitously present in plants [1, 2]. During recent
36
years, these compounds are of growing interest due to broad pharmacological features, including
37
anti-thrombotic, anti-inflammatory, anticancer and immunostimulator activities [3-5]. Moreover,
38
in the human diet quercetin (QU) (3,3´,4´,5,7-pentahydroxyflavone) is one of the most abundant
39
natural flavonoid, which antioxidant activity is higher than for the other well-known antioxidant
40
molecules, i.e. ascorbylpalmitate, trolox and rutin [6, 7]. Besides antioxidant properties,
41
quercetin also possesses anticancer, antiviral and antiallergic features [8, 9]. The anticancer
42
properties of this natural antioxidant have been proved by in vivo and in vitro experiments which
43
demonstrated that QU has a significant role in inhibition of breast, colon, prostate, ovary,
44
endometrium, and lung tumor cancer cells [10-12].
45
The main limitation in the therapeutic employment of QU is its low water solubility and
46
instability in physiological medium which restricts its use to oral route of administration [13]. In
47
order to extend clinical applications and to solve the solubility, instability, and bioavailability
48
issues of QU it is necessary to develop an appropriate flavonoid delivery system by
49
entrapping/adsorbing QU into the more favorable environment. A variety of different QU
50
immobilization/incorporation strategies, such as using emulsions [14-16], nanosuspensions [17],
51
microcapsules [18, 19], solid lipid nanoparticles [20-24], have been suggested to overcome at
52
least some of these limitations. Much attention has been also made to a various types
53
phospholipid-based vesicles and liposomes as potential vehicles for entrapment and delivery of
54
QU [3, 25-30]. An advantage to use phospholipids is their amphiphilic nature that can modify the
55
solubility behavior and rate of drug release and enhance drug absorption across the biological
56
barriers [3, 31, 32]. However, complex preparation routines and low encapsulation capability of
57
poor water soluble molecules restricts the potential application of vesicles and liposomes for the
58
entrapment and delivery of QU.
59
It was known for years that many phospholipids, such as phosphatidylcholines, spontaneously
60
form flat lipid bilayers and fragmented vesicular particles in aqueous dispersions. It was early on
61
also observed that a variety of polar lipids and their mixtures can self-assemble into different
Ac ce p
te
d
M
an
us
cr
ip t
33
Page 3 of 27
non-lamellar liquid crystalline (LC) phases in water under solution conditions similar to that of
63
biological systems [33]. Non-lamellar LC phases, such as sponge, hexagonal, and cubic,
64
typically comprise of hydrophilic and hydrophobic, which may be continuous or discrete, thus
65
forming either mono- or bicontinuous networks, depending on the molecular nature of the lipid
66
or lipid mixture [34]. They have generally much higher surface area per volume than lamellar
67
structures, may better solubilize hydrophobic, hydrophilic and also amphiphilic molecules [35-
68
37], and can be used as delivery systems for peptides, proteins and food bioactives [38, 39].
69
Recently, a versatile drug delivery lipid system based on soy phosphatidylcholine (SPC) and
70
glycerol dioleate (GDO) has been developed. It offers optimal functional properties, such as
71
bioadhesion and controlled release, good drug loading ability, enhanced flexibility of self-
72
assembly that can be tuned in the range from lamellar to various reversed non-lamellar LC
73
phases [40, 41]. Furthermore, in the presence of polymeric particle stabilizer Polysorbate 80
74
(P80) SPC/GDO mixtures can be easily dispersed into non-lamellar liquid crystalline
75
nanoparticles (LCNPs) with controllable small size and inner morphology making such
76
nanoparticles suitable for parenteral administration routes [42-44].
77
Here we investigated the potential of SPC/GDO-based non-aqueous lipid formulations, non-
78
lamellar LC phases and LCNPs as delivery vehicles of QU. To our best knowledge, only one
79
study of QU entrapment into non-lamellar lipid-based LC system has been previously reported.
80
Murgia and co-workers [45] have been studied the incorporation of QU into glycerol monooleate
81
(GMO)-based bicontinuous cubic LCNPs and demonstrated that small amounts of QU does not
82
alter size, charge and inner nanostructure of the nanoparticles. In this study several mono-, di-
83
glyceride and SPC/GDO-based non-lamellar LC phase forming lipid compositions were explored
84
in the context of their ability to solubilize QU. The main objective of the study was to elucidate
85
the effects of QU solubilization on the nanostructure of lipid-based non-lamellar LC phases and
86
their LCNPs as a function of lipid composition and QU concentration by means of polarized
87
light microscopy, synchrotron small-angle X-ray diffraction (SAXD), dynamic light scattering
88
(DLS) and cryogenic transmission electron microscopy (cryo-TEM). Finally, chemical stability
89
of entrapped QU in different LC systems has been evaluated by using HPLC.
Ac ce p
te
d
M
an
us
cr
ip t
62
90
Page 4 of 27
90 91
2.
Materials and methods
2.1.
Materials
93
ip t
92
94
Quercetin (QU) (CAS-No.117-39-5; purity ≥ 98.0%) was purchased from Sigma-Aldrich.
96
Diglycerolmonooleate (DGMO) (Rylo PG 29) and glycerol dioleate (GDO) (Rylo PG 19) from
97
Danisco, commercial mixture of mono-, di- and triglycerides denoted as Capmul GMO-50 from
98
Abitec, polysorbate P80 (P80) from Croda, glycerylmonooleyl ether (GME) from Niko
99
Chemicals, and soy phosphatidylcholine (SPC) denoted as SPC S100 from Lipoid were used as
100
received. Methanol (MeOH) and acetonitrile (ACN) was purchased from Sigma-Aldrich, 99.7 %
101
ethanol (EtOH) from either Merck or Solveco, and formic acid from Scharlau Chemie. All
102
chemicals were of analytical grade. The water used was passed through the NANOpure Infinity
103
(Branstead) water purification system.
104
2.2.
Preparation of lipid-based formulations
107
Non-aqueous lipid formulations were prepared by mixing appropriate amounts of lipid
108
components (DGMO, GDO, Campul GMO-50, GME and SPC) without and in the presence of
109
co-solvent ethanol to facilitate mixing. Lipid mixtures were then placed on a roller mixer at room
110
temperature for 24 h until mixed completely. Prepared lipid formulations were kept at room
111
temperature (RT) until further use. QU containing formulations were prepared by weighing
112
appropriate amounts of QU and non-aqueous lipid formulations and placing on a roller mixer for
113
24–48 h at RT until homogeneous mixture was obtained. Visual inspection between cross-
114
polarizers and light microscopy were used to inspect samples for the presence of undissolved
115
QU. In this manuscript, the concentration of QU is always expressed as wt% of the total
116
formulation weight. Component ratios of lipid formulations are also always expressed in wt%.
Ac ce p
te
106
d
105
M
an
us
cr
95
117 118
2.3.
Preparation of bulk lipid liquid crystalline (LC) phases
119
Page 5 of 27
120
LC phases were prepared by injection of QU containing non-aqueous lipid formulations (about
121
300 mg) into water at the formulation/water weight ratio of 5/95. Samples were immediately
122
sealed and left to equilibrate at RT in still standing vials for at least 10 days before experiments.
123
2.4.
Preparation of lipid liquid crystalline nanoparticle (LCNP) dispersions
ip t
124 125
Non-aqueous lipid formulations for LCNP preparation were prepared by mixing appropriate
127
amounts of lipids (SPC/GDO = 35/65, 60/40 wt%), polymeric particle stabilizer P80, and ethanol
128
at fixed (SPC+GDO)/P80/EtOH weight ratio of 75/15/10. QU containing formulations were
129
prepared by weighing appropriate amounts of QU and mixing with non-aqueous formulations at
130
QU concentration of 0.5, 1.0, 2.0 and 4.0 wt%. Lipid mixtures were placed on a roller mixer for
131
24 h until mixing was complete, and were then dispersed in 82.5 wt% water (17.5 wt% LCNPs).
132
The aqueous dispersions were immediately sealed, shaken, and left to vortex for 72 h on a
133
mechanical mixing table at 300 rpm at RT. Prepared dispersions were stored at RT until further
134
use.
M
an
us
cr
126
136
2.5.
Polarizing light microscopy (PLM)
te
137
d
135
The texture and temperature-induced phase transitions of LC phases were examined by a
139
polarizing microscope Optiphot equipped with a digital camera DS-2Mv (Nikon) and a heating
140
table Analysa LTS350 (Linkam). A small specimen of sample was placed between two
141
microscope slides and their edges were immediately sealed with a thermo stable silicon grease to
142
prevent evaporation of water from the specimen. A stepwise increase of temperature (typically
143
5 °C in every step with a heating rate of 1 °C/min) was used to induce phase transitions. QU
144
loaded lipid LCNPs were examined at RT by a polarizing microscope DM 750 (Leica) equipped
145
with a digital camera MC170 (Leica).
Ac ce p
138
146 147
2.6.
Small-angle X-ray diffraction (SAXD)
148 149 150
The nanostructure of LC phases was studied using synchrotron SAXD measurements, which were performed at the I911-4beamline at MAX-lab (Lund University, Sweden), using a 1M
Page 6 of 27
PILATUS 2D detector containing a total of 981 x 1043 pixels. Bulk lipid LC samples were
152
mounted between kapton windows in a steel sample holder at the sample to detector distance of
153
1917 mm. Diffractograms were recorded with a wavelength of 0.91 Å and the beam size of 0.25
154
× 0.25 mm (full width at the half-maximum) at the sample. Temperature control within 0.1 °C
155
was achieved using computer controlled Julabo heating circulator F12-MC (Julabo Labortechnik
156
GMBH, Seelbach, Germany). The experiments were performed successively at 25, 35, 45, 55,
157
and 65 °C with a 60 s exposure time at each temperature and a wait of 10 minutes between
158
temperature steps. The resulting CCD images were integrated and analyzed using the Fit2D
159
software provided by Dr. A. Hammersley [http://www.esrf.fr/computing/scientific/FIT2D].
160
Silver behenate calibrated sample-to-detector distance and detector positions were used.
us
cr
ip t
151
162
2.7.
an
161
Particle size and zeta potential
163
LCNP particle size distributions and zeta potentials were measured using Zetasizer Nano ZS
165
analyzer from Malvern Instruments. For particle size distributions the disposable cuvette filled
166
with 1 mL of LCNP dispersion, which was first diluted to 99.5 wt% of water. The obtained data
167
were averaged from 30 measurements (10 s each). The refractive indices used for lipid particles
168
and water were 1.48 and 1.33, respectively. The particle size distributions were reported as
169
intensity-averaged. The surface charge of the particles was measured using disposable zeta cells
170
filled with 1 mL of LCNP dispersion which was first diluted to 99.5 wt% of water. The zeta
171
potential was calculated using the Smoluchowski approximation for dispersion in water with
172
viscosity of 0.8872 cP, refractive index of 1.33, and dielectric constant of 78.5.
173 174 175
2.8.
Ac ce p
te
d
M
164
Cryogenic transmission electron microscopy (Cryo-TEM)
176
Lipid LCNP dispersions for electron microscopy were prepared in a controlled environment
177
vitrification system to ensure stable temperature and to avoid loss of water during sample
178
preparation. The climate chamber temperature was kept at 25−28 °C, and the relative humidity
179
was kept close to saturation to prevent sample evaporation. The samples were prepared by
180
placing 5 μL of LCNP dispersion on lacey carbon filmed copper grids and gently blotted with
181
filter paper to obtain a thin liquid film (20−400 nm) on the grid. Immediately after blotting, the
Page 7 of 27
grids were rapidly plunged into the liquid ethane at −180 °C to vitrify the water-rich samples to
183
prevent ice crystal formation and to preserve the internal crystalline structure. The vitrified
184
specimens were stored in liquid nitrogen (−196 °C) until measurements. An Oxford CT3500
185
cryo-holder and its work station were used to transfer the samples into the electron microscope
186
(Philips CM120 BioTWINCryo) equipped with a post-column energy filter (Gatan GIF100). The
187
acceleration voltage was 120 kV, and the working temperature was kept below −180 °C. The
188
images were recorded digitally with a CCD camera under low electron dose conditions.
cr
ip t
182
189
2.9.
QU stability study
us
190 191
The chemical stability of QU entrapped in different lipid non-aqueous formulations, LC phases
193
and LCNP dispersions were monitored for 3 months. During stability study all samples were
194
kept at RT in the darkness. For the evaluation of the residual QU concentration a portion of the
195
sample (about 10-15 mg) was collected at predetermined time points (1, 15, 30, 60 and 90 days),
196
dissolved in ACN:MeOH (1:1 v/v) solvent mixture at the lipid sample to solvent weight ratio of
197
1:100 and immediately analyzed using HPLC. Each sample was analyzed in triplicate.
199 200
2.10.
High performance liquid chromatography (HPLC)
te
198
d
M
an
192
QU concentration was determined using an Agilent HPLC 1100 Series (Agilent Technologies,
202
USA) chromatography system equipped with a quaternary pump, a vacuum degasser module, a
203
manual injector with a 20 µL sample loop, a temperature controlled column compartment and a
204
diode array detector (DAD) set at 370 nm. Chromatographic separation was achieved using a
205
Zorbax Eclipse XDB–C18 (5 µm, 150×4.60 mm, Agilent Technologies, USA) reversed phase
206
column coupled with an Eclipse XDB–C18 guard column (5 µm, 12.6×4.6 mm, Agilent
207
Technologies, USA). Determination was carried out in a solvent system of methanol–formic
208
acid–water as previously described, with minor modifications [46]. Data were collected and
209
processed using a ChemStation software version B.01.03. The obtained values with standard
210
methanolic QU solutions showed linearity over the concentration range of 0.1–100 μg/g with a
211
correlation coefficient (r2) of 0.999. The quantification limit in the HPLC assay was 0.1 μg/g and
212
standard deviation under repeatability conditions was no more than 5.6% in all concentrations
Ac ce p
201
Page 8 of 27
213
tested. The calibration curves for quantification of QU in QU-loaded non-aqueous lipid
214
formulations, LC phases and LCNP dispersions were prepared using QU standard with
215
respective lipid compositions. The obtained results of the quantification of various formulations
216
are summarized in Table S1 (Supporting information).
Ac ce p
te
d
M
an
us
cr
ip t
217
Page 9 of 27
217 218
3.
Results and discussions
3.1.
Solubility of QU in lipid formulations
220
ip t
219
221
It is known that some poorly water soluble compounds like QU may also have low solubility in
223
lipid excipients due to different physical chemical reasons [31]. As shown in other lipid- and
224
amphiphile-based systems, such as microemulsions [19], liposomes [26], lecithin-chitosan
225
nanoparticles [23] and solid lipid nanoparticles [24], the maximum solubility of QU is quite low
226
and often is in the order of 0.5 wt%, which can be somewhat improved by using additional
227
polymeric surfactants and/or co-solvents [16].
228
Therefore, before further experiments several non-lamellar LC phase forming non-aqueous lipid
229
mixtures were investigated with respect to their ability to solubilize QU. First, binary mixtures of
230
DGMO/GDO (85/15 and 60/40) (here and everywhere in the text formulation compositions are
231
expressed in wt%), DGMO/Capmul GMO-50 (85/15 and 60/40), DGMO/GME (85/15 and
232
60/40) and SPC/GDO (50/50) were explored. The results have shown that independently from
233
lipid ratio the solubility of QU in all tested mono- and di-glyceride-based mixtures was very low
234
and varies between 0.3 and 0.4 wt%. The addition of up to 10 wt% of ethanol (in respect to total
235
lipid) did not improve the solubility of QU in neither of the formulations. The observed solubility
236
in mixtures of mono- and di-glycerides was in line with previous results in similar systems. The
237
solubility of QU in commercial mixtures of long chain monoglycerides (i.e., Miglyol 812,
238
Capmul MCM, Labrafil 1944) is usually well below 1 wt% [16].
239
Much higher solubility of QU was found for the mixtures of SPC and GDO. Since solvent free
240
SPC/GDO mixtures prepared at equal lipid weight ratio were difficult to mix and were very
241
viscous their ability to solubilize QU were only tested in the presence of small amount of ethanol
242
at fixed Lipid/EtOH weight ratio of 90/10. SPC/GDO composition in the formulations was
243
varied between 60/40 and 35/65. Independently on lipid ratio about 5 wt% of QU was soluble in
244
all formulations. From the solubility test it may be concluded that lipid mixtures composed of
245
exclusively mono- and di-glycerides with and without EtOH have very limited ability to
246
solubilize QU which may be increased by about 10-20 times with the introduction of SPC into
Ac ce p
te
d
M
an
us
cr
222
Page 10 of 27
247
the mixture. Considering these results, SPC/GDO-based formulations containing up to 4 wt% of
248
QU and 10 wt% EtOH were selected for further studies.
249 250
3.2.
SPC/GDO-based LC phases with entrapped QU
ip t
251
The aim of this part was to investigate the effects of QU on the aggregation behavior of fully
253
hydrated non-lamellar bulk LC phases of SPC/GDO. As shown in recent study, the aqueous
254
phase behavior of mixtures of SPC and GDO is rather complex [40]. At 25 °C with increasing
255
GDO content fully hydrated SPC/GDO mixtures in water form the following LC phase
256
sequence: lamellar (Lα) reversed 2D hexagonal (H2, up to 62.5/37.5) reversed micellar
257
cubic of Fd m space group (50/50 – 45/55) reversed 3D hexagonal of P63/mmc space group
258
(42/58 – 40/60) unresolved “intermediate” (39/61 – 37/63) Fd m (35/65 – 22.5/77.5)
259
reversed micellar solution (L2, from 20/80). In this study, the effects of QU on the nanostructure
260
of different LC phases were investigated at four fixed SPC/GDO weight ratios (60/40, 50/50,
261
40/60 and 35/65) and four temperatures (25, 35, 45 and 55 °C).
262
Figure 1 shows SAXD data of fully hydrated SPC/GDO (60/40 and 35/65) LC phases with
263
entrapped 0.0, 0.5, 1 and 4 wt% of QU as a function of temperature.
264
As seen from Figure 1a, at weight ratio of 60/40 and at 25 °C SPC/GDO forms H2 LC phase,
265
which features three strong Bragg reflections positioned in ratios 1: 3: 4. At 45 °C another
266
coexisting LC phase, most likely Fd m cubic, also starts to form what is evidenced by the
267
appearance of additional peaks. However, independently on temperature the appearance of all
268
diffractograms is practically unaffected by the presence of up to 4 wt% of QU (Figures 1b-1d).
269
In addition, the calculated lattice parameter (a) for this LC phase is constant regardless QU
270
concentration and temperature and only slightly varies between 6.6 and 6.7 nm. This shows that
271
H2 phase prepared at high SPC content is quite robust and can accommodate relatively high
272
amounts of hydrophobic QU without changing nanostructure.
273
At weight ratio of 35/65 and 25 °C SPC/GDO self-assembles into ordered reversed micellar
274
Fd m cubic phase which is clearly characterized by the appearance of first 9 Bragg peaks
275
located at relative positions in ratios 3: 8: 11: 12: 16: 19: 24: 27: 32 (Figure 1e).
276
Contrary to H2 phase, the structure of Fd m cubic phase prepared at low SPC content is rather
277
sensitive to QU entrapment and temperature. At 45 °C it starts to “melt” and at 55 °C completely
Ac ce p
te
d
M
an
us
cr
252
Page 11 of 27
transforms into unordered reversed micelles (L2), which is corroborated by the disappearance of
279
the Bragg diffraction peaks and their smearing into two broad diffuse diffraction features. The
280
entrapment of QU even further disturbs ordered structure of the Fd m cubic phase. At 0.5 and 1
281
wt% of QU ordered cubic phase is almost fully transformed into unordered phase at 45 °C
282
(Figures 1f and 1g). Moreover, only a fraction of the Fd m cubic arrangement with entrapped 4
283
wt% of QU is present already at 35 °C (Figure 1h). As shown in Figure S1 (SI), both the
284
entrapment of QU and temperature considerably enlarge the unit cell dimensions of the cubic LC
285
phase at the SPC/GDO ratio of 35/65. With increasing QU concentration from 0 to 4 wt% the
286
calculated a value increases from 14.9 to 15.6 nm at 25 °C. At elevated temperatures a further
287
increases and reaches maximum value of about 15.9 nm at 1 wt% of QU and 45 °C before
288
complete transformation of ordered Fd m structure into unordered L2 phase.
289
The observed disordering effect of QU may be explained by reasonable assumption that in the
290
LC phase water insoluble hydrophobic QU molecule is preferentially located between lipid
291
hydrocarbon chains. Accommodation of QU in the hydrophobic regions between ordered
292
reversed micelles increases lipid chain packing stress and distance between micelles resulting in
293
slightly larger a values. At some point, further accommodation of QU and unit cell dimension
294
increase is not possible without transformation into unordered L2 phase. At higher temperatures
295
this process occurs at lower QU concentrations due to additional temperature induced lipid chain
296
packing disorder. Similar phase behavior trends were also observed for the SPC/GDO Fd m
297
cubic phases with entrapped benzydamine, lidocaine and granisetron [41].
298
In addition, less extensive but similar disordering effects of QU were also observed for the LC
299
phases prepared at intermediate SPC/GDO weight ratios of 50/50 and 40/60 (Figure S2, SI).
300
Here, the entrapment of QU has little effect on the unit cell dimensions of the LC phases.
301
However, in both cases QU decreases phase transition temperatures of ordered LC phases into L2
302
phase, which is observed for temperatures higher than 45 °C. Here it may be concluded that
303
effect of QU on the nanostructure of the bulk SPC/GDO LC phases is lipid weight ratio-
304
dependent. Thus, at weight ratio of 60/40, H2 LC structure remains unaffected in the presence of
305
QU. In contrast, at weight ratio of 35/65 the structure of the cubic Fd m phase is very sensitive
306
to QU concentration increase and starts transformation into L2 phase already at 35 °C and 4 wt%
307
of QU. Finally, LC phases at the SPC/GDO weight ratios of 50/50 and 40/60 show moderate
308
sensitivity to QU.
Ac ce p
te
d
M
an
us
cr
ip t
278
Page 12 of 27
309 310
3.3.
SPC/GDO/P80-based LCNPs with entrapped QU
311
The aim of this part was to investigate the effects of QU on size and stability of the dispersed
313
SPC/GDO/P80-based liquid crystalline nanoparticles (LCNPs). Dispersions were prepared at
314
SPC/GDO weight ratios of 60/40 and 35/65 in the presence of polymeric stabilizer P80 at fixed
315
lipid to polymer ratio. Dispersions containing up to 4 wt% of QU were prepared and
316
characterized with regards to particle size, charge and morphology.
317
Figure 2 shows obtained particle size distributions as a function of QU concentration. As seen
318
from Figures 2a and 2b, at low QU concentrations all LCNP dispersions are well-defined
319
displaying monomodal size distributions with polydispersity indices ranging between 0.13 and
320
0.16 µm. At 2 and 4 wt% of QU additional larger aggregates of about 5 m appear in the
321
dispersions prepared at both SPC/GDO ratios. Sample examination under polarized light
322
microscope reveals that these larger aggregates are small needle-like crystals phase separated
323
from LCNPs (Figure S3, SI). Since crystals have distinct yellow color (seen in microscope under
324
nonpolarized light), there is good reason to assume that they are formed exclusively by QU. This
325
show that SPC/GDO/P80 LCNPs can homogeneously entrap only up to 2 wt% of QU when
326
compared to 4-5 wt% of QU in the bulk LC phases. This difference may be attributed to a very
327
large surface-to-volume ratio of the particles and exposure to the excess aqueous phase creating
328
more defects and crystallization centers for QU. In addition, one cannot exclude that presence of
329
P80 may also change QU solubilization properties and induce precipitation. Note however, that
330
QU crystallization and phase separation does not affect LCNP own size distribution
331
characteristics. The colloidal stability of the LCNP dispersions with entrapped QU is also good.
332
Only minor changes in the size distributions are observed after 3 months of storage at RT (Figure
333
2, dotted lines).
334
Figure 3 demonstrates that the obtained mean particle size clearly depends on both lipid ratio and
335
QU concentration. Thus, LCNPs at SPC/GDO ratio 60/40 are slightly larger with the mean
336
particle size increasing from 140 to 210 nm as QU concentration is increased from 0 to 4 wt%. In
337
contrast, LCNPs at SPC/GDO ratio 35/65 are smaller with the mean particle size ranging from
338
about 80 to 110 nm. Unfortunately, we cannot explain such an effect of QU on the particle size
339
considering only bulk phase behavior and nanostructural features of the LC phases. Since LCNP
Ac ce p
te
d
M
an
us
cr
ip t
312
Page 13 of 27
dispersions were prepared by mechanical agitation we believe that the entrapment of QU may
341
slightly increase cohesion forces within LC structure and/or influence monocrystalline domain
342
size. Therefore, more energy is required to brake LC phase with entrapped QU into smaller
343
particles. In addition, LCNP zeta potential remains unaffected by the entrapment of QU and is
344
about -16 and -10 mV for particles at SPC/GDO ratio 60/40 and 35/65, respectively. Considering
345
size and surface to volume ratio differences the surface charge density of the particles is similar
346
at both SPC/GDO ratios.
cr
347
ip t
340
Representative cryo-TEM images shown in Figure 4 together with the measured SAXD profiles (Figure S4, SI) give further insights into QU-loaded particle morphology and
349
nanostructure. Overall, the observed particle sizes are consistent with DLS measurements. At
350
SPC/GDO ratio 60/40 particles are larger and characterized by a denser inner core and a less
351
dense shell comprising lamellar and “sponge”-like petals (Figures 4a and 4b). The observed
352
particle morphology is comparable with those previously prepared at similar SPC/GDO weight
353
ratio and P80 content without QU [47]. The obtained LCNP SAXD profiles as a function of QU
354
concentration in Figure S4a (SI) are almost identical showing negligible effect of QU on particle
355
nanostructure and are in line with the bulk phase behavior observations (Figures 1a-1d). The data
356
show only weak diffuse scattering which is not surprising since it is know that dispersion agent
357
P80 has a disordering effect on the internal SPC/GDO LCNP nanostructure [43]. Only small
358
shoulder appearing at q value of about 1.1 nm-1 can be observed which position may be related to
359
the first reflection of the undispersed H2 phase (Figures 1a-1d). Regardless of this relation SAXD
360
data suggest that in dispersed state the long-range order of the H2 LC arrangement in the particle
361
core is almost or completely lost.
362
As shown in Figures 4c and 4d, particles prepared at SPC/GDO 35/65 ratio are smaller in size
363
with more or less even density and hardly any noticeable features in the interior. This is also
364
reflected in the SAXD profiles shown in Figure S4b where only weak scattering pattern is
365
observed without any indications of the Bragg diffraction. From inspection of several dozen of
366
cryo-TEM images it may be also concluded that GDO-rich particles, with entrapped QU do not
367
possess pronounced swollen coronas of multiply connected bilayers which are always present for
368
SPC/GDO LCNPs prepared at high fraction of P80 [47, 48]. The observed difference may be
369
related to the bulk phase behavior where even low concentrations of QU were able to induce
370
Fd m phase transformation towards more reversed L2 phase (Figures 1e-1h). Most likely, in the
Ac ce p
te
d
M
an
us
348
Page 14 of 27
371
dispersed LC state the addition of small amount of QU is enough to allow more homogeneous
372
distribution of components within the particle and to prevent the formation of lamellar-like
373
corona.
374
3.4.
376
LCNPs
QU stability in SPC/GDO non-aqueous formulations, bulk LC phases and dispersed
ip t
375
cr
377
One of the important aspects of delivery formulations is the chemical stability of active
379
substance. Therefore, HPLC was used to evaluate the chemical stability of QU entrapped in all
380
three studied SPC/GDO-based systems: non-aqueous formulations, LC phases and LCNP
381
dispersions. Samples were prepared at two different SPC/GDO weight ratios containing 0.5 and
382
1 wt% of QU in respect to dry formulation weight.
383
For up to 90 days QU showed very good chemical stability when solubilized in non-aqueous
384
SPC/GDO-based formulations (Figures 5a and 5d). About 85–90% of QU is retained
385
independently on lipid composition and QU concentration. Introduction of water and hydration
386
of the formulations into bulk LC phases (Figures 5b and 5e) and LCNP dispersions (Figures 5c
387
and 5f) had more pronounced effect on the stability of QU. Gradual but significant loss of QU
388
content upon storage was observed in both systems. In the LC phases prepared at SPC/GDO ratio
389
35/65 QU content decreased to about 65%, whereas only 25–40% of QU was found after the end
390
of the study in the LC phases prepared at lipid weight ratio of 60/40. Such substantial difference
391
may be explained by the fact that SPC used in this study contain considerable amounts of linoleic
392
(C18:2) fatty acid chains. It is known that degree of unsaturation plays positive role in the lipid
393
autoxidation processes, in which highly reactive primary and secondary lipid oxidation products
394
are produced [49]. In SPC-rich LC phases larger amounts of reactive species may be formed,
395
which interact and chemically degrade QU when compared to the GDO-rich LC phases where
396
oleic fatty acid residues are dominant.
397
The lowest stability of QU was found in the LCNP dispersions where only 30–45% was
398
recovered after 90 days. Here, GDO-rich LCNPs showed only slight improvement over SPC-rich
399
dispersions. Most likely, P80 which was used to stabilize LCNPs dispersions also influence lipid
400
autoxidation complex processes and further decrease stability of QU. On the other hand, water
Ac ce p
te
d
M
an
us
378
Page 15 of 27
penetration ability and diffusion of lipid components is less restricted within small particle when
402
compared to macroscopic bulk LC phase and may accelerate degradation of QU.
403
Overall, QU shows good chemical stability when loaded in SPC/GDO-based non-aqueous
404
formulations for up to 3 months of storage at room temperature. On the other hand, hydrated
405
formulations (bulk LC phases and LCNP dispersions) show reduced QU chemical stability and
406
therefore are not suited for a long-term shelf storage but can still serve as suitable delivery
407
systems if are formed via in situ hydration of non-aqueous formulation (in case of LC phases) or
408
freshly produced just prior application/administration (in case of LCNP dispersions). For real-life
409
application, further studies on parameter optimization (e.g., sterilization, oxygen concentration,
410
temperature, pH, additives and other) in order to extend long-term stability of the formulations
411
will be needed.
412
3.
an
us
cr
ip t
401
Conclusion
M
413
The present study demonstrates the potential of liquid crystalline structures forming SPC/GDO
415
lipid formulations as encapsulation and delivery matrices of QU. The results have shown that up
416
to 5 wt% of QU can be effectively solubilized in non-aqueous SPC/GDO formulations by using
417
small amounts of EtOH as solvent. Prepared formulations are stable and have minimal effect on
418
the chemical stability of QU for few months. Upon hydration non-aqueous formulations can
419
easily self-assemble into non-lamellar bulk LC matrices with different nanostructures. The effect
420
of QU on the nanostructure of LC phases is lipid composition dependent. At high SPC content,
421
the entrapment of QU has practically no effect on the nanostructure of H2 phase at
422
physiologically relevant temperatures. At low SPC content, QU slightly increases the unit cell
423
dimensions of the reversed micellar Fd m cubic phase and promotes the formation of reversed
424
micellar solution at elevated temperatures. Finally, concentrated and colloidally stable non-
425
lamellar SPC/GDO-based LCNP dispersions containing up to 2 wt% of entrapped QU can be
426
easily prepared in the presence of stabilizer P80. Dispersion particle size can be tuned in the
427
range of about 80 – 210 nm by changing lipid composition and entrapped QU concentration.
Ac ce p
te
d
414
428 429
Author’s contributions
430
Page 16 of 27
431
Linkeviciute A. executed the experiment, extracted the data and wrote the initial draft of the
432
manuscript. Barauskas J. supervised the research, analysis of data and interpreted the results
433
obtained from the models. Misiunas A. technical assistance, while Naujalis E. was co-
434
supervising the research.
436
ip t
435
Conflict of interest
438
cr
437
The authors declare that they have no conflict of interest.
us
439
Acknowledgements
441
The authors acknowledge the Swedish synchrotron X-ray facility MAX IV Laboratory for
442
allocated beamtime at the I911-4beamline and Ana Labrador for technical support during
443
experiments. Authors also thank Gunnel Karlsson and Viveka Alfredsson at Lund University for
444
their help with the cryo-TEM imaging. AL acknowledges financial support from the Research
445
Council of Lithuania "Promotion of Student Scientific Activities" (VP1-3.1-ŠMM-01-V-02-003).
M
d te Ac ce p
446
an
440
Page 17 of 27
446 447
FIGURES
448
Figure 1. SAXD profiles of SPC/GDO mixtures in excess water prepared at lipid weight ratios
450
of 60/40 (a-d) and 35/65 (e-h) as a function of QU concentration (0, 0.5, 1.0 and 4.0 wt%) and
451
temperature (25, 35, 45 and 55 °C). Arrows in (e) and (h) show indexing of the reflections from
452
the reversed Fd m cubic phase. An explanation is given in the text.
us
453
cr
ip t
449
Figure 2. Particle size distributions of freshly prepared (solid lines) and 3 months old (dotted
455
lines) SPC/GDO/P80 LCNPs at SPC/GDO weight ratios of 60/40 (a) and 35/65 (b) as a function
456
of QU concentration. Dispersions were prepared at fixed lipid/P80 weight ratio of 75/15 in
457
82.5% water.
M
an
454
458
Figure 3. Dependence of mean particle size of the SPC/GDO/P80 LCNPs on QU concentration.
460
Dispersions were prepared at SPC/GDO weight ratios of 60/40 (open circles) and 35/65 (filled
461
circles). The lines are drawn to guide the eye.
te
Ac ce p
462
d
459
463
Figure 4. Representative cryo-TEM images of SPC/GDO/P80 LCNPs with entrapped 1 wt% QU
464
prepared at SPC/GDO weight ratios of 60/40 (a and b) and 35/65 (c and d). Dispersions were
465
prepared at fixed lipid/P80 weight ratio of 75/15 in 82.5% water.
466 467
Figure 5. Chemical stability of QU (normalized to initial) entrapped in SPC/GDO-based non-
468
aqueous formulations (a and d), LC phases (b and e) and LCNP dispersions (c and f). Samples
469
were prepared at fixed SPC/GDO weight ratios of 60/40 (open symbols) and 35/65 (filled
470
symbols) containing 0.5 (a-c) and 1.0 wt% (d-f) of QU in respect to dry weight. Results are
471
represented as mean values ± standard deviation (n=3).
472
Page 18 of 27
472 473
References
474
te
d
M
an
us
cr
ip t
[1] W. Vermerris and R. Nicholson, Phenolic Compound Biochemistry, Springer, 2006. [2] S. Kumar and A.K. Pandey, Sci World J, (2013). [3] D. Singh, M.S.M. Rawat, A. Semalty and M. Semalty, Curr Drug Deliv, 9 (2012) 305-314. [4] F. Bonina, M. Lanza, L. Montenegro, C. Puglisi, A. Tomaino, D. Trombetta, F. Castelli and A. Saija, Int J Pharmaceut, 145 (1996) 87-94. [5] A.C. Santos, S.A. Uyemura, J.L.C. Lopes, J.N. Bazon, F.E. Mingatto and C. Curti, Free Radical Bio Med, 24 (1998) 1455-1461. [6] M.I. Kaldas, U.K. Walle, H. Van der Woude, J.M. McMillan and T. Walle, J Agr Food Chem, 53 (2005) 4194-4197. [7] N. Nuengchamnong, A. Hermans-Lokkerbol and K. Ingkaninan, Naresuan University Journal, 12 (2004) 25-37. [8] J.V. Formica and W. Regelson, Food Chem Toxicol, 33 (1995) 1061-1080. [9] Y. Zheng, I.S. Haworth, Z. Zuo, M.S.S. Chow and A.H.L. Chow, J Pharm Sci-Us, 94 (2005) 1079-1089. [10] R.K. Hansen, S. Oesterreich, P. Lemieux, K.D. Sarge and S.A.W. Fuqua, Biochem Bioph Res Co, 239 (1997) 851-856. [11] G. Elia, C. Amici, A. Rossi and M.G. Santoro, Cancer Res, 56 (1996) 210-217. [12] M. Koishi, N. Hosokawa, M. Sato, A. Nakai, K. Hirayoshi, M. Hiraoka, M. Abe and K. Nagata, Jpn J Cancer Res, 83 (1992) 1216-1222. [13] T. Pralhad and K. Rajendrakumar, J Pharmaceut Biomed, 34 (2004) 333-339. [14] R. Casagrande, S.R. Georgetti, W.A. Verri, M.F. Borin, R.F.V. Lopez and M.J.V. Fonseca, Int J Pharmaceut, 328 (2007) 183-190. [15] S. Jain, A.K. Jain, M. Pohekar and K. Thanki, Free Radical Bio Med, 65 (2013) 117-130. [16] T.H. Tran, Y. Guo, D.H. Song, R.S. Bruno and X.L. Lu, J Pharm Sci-Us, 103 (2014) 840-852. [17] L. Gao, G.Y. Liu, X.Q. Wang, F. Liu, Y.F. Xu and J. Ma, Int J Pharmaceut, 404 (2011) 231-237. [18] D.H. Lee, G.S. Sim, J.H. Kim, G.S. Lee, H.B. Pyo and B.C. Lee, J Pharm Pharmacol, 59 (2007) 16111620. [19] F.T.M.C. Vicentini, T.R.M. Simi, J.O. Del Ciampo, N.O. Wolga, D.L. Pitol, M.M. Iyomasa, M.V.L.B. Bentley and M.J.V. Fonseca, Eur J Pharm Biopharm, 69 (2008) 948-957. [20] Guo Chen-yua, Yang Chun-fena, Li Qi-luc, Tan Qia, Xi Yan-weia, Liu Wei-nad and Z. Guang-xia, Int J Pharmaceut, 430 (2012) 292– 298. [21] S. Bose and B. Michniak-Kohn, Eur J Pharm Sci, 48 (2013) 442-452. [22] Y.Y. Zhang, Y. Yang, K. Tang, X. Hu and G.L. Zou, J Appl Polym Sci, 107 (2008) 891-897. [23] Q. Tan, W.D. Liu, C.Y. Guo and G.X. Zhai, Int J Nanomed, 6 (2011) 1621-1630. [24] H.L. Li, X.B. Zhao, Y.K. Ma, G.X. Zhai, L.B. Li and H.X. Lou, J Control Release, 133 (2009) 238-244. [25] E. Alexopoulou, A. Georgopoulos, K.A. Kagkadis and C. Demetzos, J Liposome Res, 16 (2006) 17-25. [26] A. Priprem, J. Watanatorn, S. Sutthiparinyanont, W. Phachonpai and S. Muchimapura, NanomedNanotechnol, 4 (2008) 70-78. [27] A.A. Date, M.S. Nagarsenker, S. Patere, V. Dhawan, R.P. Gude, P.A. Hassan, V. Aswal, F. Steiniger, J. Thamm and A. Fahr, Mol Pharmaceut, 8 (2011) 716-726. [28] A.P. Landi-Librandi, T.N. Chrysostomo, A.E.C.S. Azzolini, C.M. Marzocchi-Machado, C.A. de Oliveira and Y.M. Lucisano-Valim, J Liposome Res, 22 (2012) 89-99.
Ac ce p
475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515
Page 19 of 27
545 546 547
te
d
M
an
us
cr
ip t
[29] P.G. Cadena, M.A. Pereira, R.B.S. Cordeiro, I.M.F. Cavalcanti, B.B. Neto, M.D.C.B. Pimentel, J.L. Lima, V.L. Silva and N.S. Santos-Magalhaes, Bba-Biomembranes, 1828 (2013) 309-316. [30] C. Caddeo, O. Diez-Sales, R. Pons, X. Fernandez-Busquets, A.M. Fadda and M. Manconi, Pharm ResDordr, 31 (2014) 959-968. [31] H.L. Mu, R. Holm and A. Mullertz, Int J Pharmaceut, 453 (2013) 215-224. [32] K. Sandeep, M. Mohanvarma and P. Veerabhadhraswamy, Acta Pharmaceutica Sinica B, 3 (2013) 361-372. [33] V. Luzzati, R. Vargas, P. Mariani, A. Gulik and H. Delacroix, J Mol Biol, 229 (1993) 540-551. [34] K. Larsson, P. Quinn, K. Sato and F. Tiberg, Vol. 19 (2006). [35] L. Sagalowicz and M.E. Leser, Curr Opin Colloid In, 15 (2010) 61-72. [36] K. Larsson and F. Tiberg, Curr Opin Colloid In, 9 (2005) 365-369. [37] A. Angelova, B. Angelov, R. Mutafchieva, S. Lesieur and P. Couvreur, Accounts Chem Res, 44 (2011) 147-156. [38] I. Amar-Yuli, D. Libster, A. Aserin and N. Garti, Curr Opin Colloid In, 14 (2009) 21-32. [39] S.B. Rizwan, B.J. Boyd, T. Rades and S. Hook, Expert Opin Drug Del, 7 (2010) 1133-1144. [40] F. Tiberg, M. Johnsson, M. Jankunec and J. Barauskas, Chem Lett, 41 (2012) 1090-1092. [41] J. Barauskas, L. Christerson, M. Wadsater, F. Lindstrom, A.K. Lindqvist and F. Tiberg, Mol Pharmaceut, 11 (2014) 895-903. [42] F. Tiberg and M. Johnsson, J Drug Deliv Sci Tec, 21 (2011) 101-109. [43] M. Wadsater, J. Barauskas, T. Nylander and F. Tiberg, Acs Appl Mater Inter, 6 (2014) 7063-7069. [44] M. Wadsater, J. Barauskas, T. Nylander and F. Tiberga, Soft Matter, 9 (2013) 8815-8819. [45] S. Murgia, S. Bonacchi, A.M. Falch, S. Lampis, V. Lippolis, V. Meli, M. Monduzzi, L. Prodi, J. Schmidt, Y. Talmon and C. Caltagirone, Langmuir, 29 (2013) 6673-6679. [46] A. Linkeviciute, R. Butkiene and E. Naujalis, Chemija, 24 (2013) 217-222. [47] D.P. Chang, M. Jankunec, J. Barauskas, F. Tiberg and T. Nylander, Langmuir, 28 (2012) 10688-10696. [48] D.P. Chang, M. Jankunec, J. Barauskas, F. Tiberg and T. Nylander, Acs Appl Mater Inter, 4 (2012) 2643-2651. [49] A. Morales, C. Dobarganes, G. Marquez-Ruiz and J. Velasco, J Am Oil Chem Soc, 87 (2010) 12711279.
Ac ce p
516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544
Page 20 of 27
547 548 549 550 551
We examine the mixtures of SPC and GDO as encapsulation matrices for quercetin (QU). SPC/GDO-based formulations can incorporate relatively high amounts of QU. At high SPC content, the loaded QU has no effect on the nanostructure of H2 phase. At low SPC content, QU slightly increases the unit cell dimensions of Fd m phase.
Ac ce p
te
d
M
an
us
cr
ip t
552
Page 21 of 27
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 22 of 27
Ac ce p
te
d
M
an
us
cr
ip t
Figure(s)
Page 23 of 27
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 24 of 27
Ac ce p
te
d
M
an
us
cr
ip t
Figure(s)
Page 25 of 27
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 26 of 27
Ac
ce
pt
ed
M
an
us
cr
i
*Graphical Abstract (for review)
Page 27 of 27
