Mucoadhesive liposomes as new formulation for vaginal delivery of curcumin
Katja Berginc1,2,4, Sabina Suljaković2, Nataša Škalko-Basnet3, Albin Kristl2
1
Lek d.d., Verovškova 57, 1000 Ljubljana, Slovenia.
2
Faculty of Pharmacy, University of Ljubljana, Aškerčeva 7, 1000 Ljubljana, Slovenia.
3
Drug Transport and Delivery Research Group, Department of Pharmacy, Faculty of Health
Sciences, University of Tromsø, Universitetsveinen 57, 9037 Tromsø, Norway.
4
To whom correspondence should be addressed (e-mail: katja.berginc@sandoz.com;
telephone: ++386 1 58 03 989)
ABSTRACT
Local delivery to the affected area represents the optimal means by which advantageous
pharmacological properties of curcumin may be fully exploited as currently, due to the
biopharmaceutical limitations associated with this polyphenol, its full beneficial effects
remain limited. Curcumin-containing liposomes coated with bioadhesive polymers of natural
and synthetic origin (chitosan, Carbopol) were evaluated in vitro. For these purposes, an in
vitro model of vaginal mucus was developed allowing the monitoring of curcumin
permeability in the conditions mimicking vaginal environment. The model was optimized by
varying the amounts of glycoproteins, as compared to the permeabilities determined through
isolated bovine mucus. The strength of bioadhesion was evaluated using the isolated bovine
mucosa. Both curcumin solution and non-coated curcumin liposomes served as controls.
Bioadhesive polymers enabled significantly higher (p < 0.05) curcumin permeability through
the artificial and isolated bovine mucus compared to the controls. Polymer coating of
liposomes resulted in an increase in their bioadhesiveness. Mucoadhesive liposomes can be
considered as potential novel drug delivery systems intended for vaginal administration of
curcumin.
Key words: curcumin, vaginal delivery, chitosan, Carbopol, mucoadhesive liposomes.
List of abbreviations:
BCS – Biopharmaceutical Classification System
CPGM – crude pig gastric mucins
MLV – multilamellar vesicles
PBS – phosphate buffer saline
PC – phosphatidylcholine
SVF – simulated vaginal fluid
1. INTRODUCTION
Curcumin represents one of 400 natural candidates that are currently being scrutinized as
potential chemo-preventive agents and this polyphenolic compound has shown promise in
clinical trials for a variety of cancer conditions including multiple myeloma, pancreatic, and
colon cancer (1, 2). Being sensitive to environmental conditions (i.e. stability issues) coupled
with poor absorption (low solubility, low permeability), and extensive pre-systemic
metabolism, attempts to deliver a systemically effective curcumin dose and to achieve healthbeneficial effects outside the gut mucosa have failed (3). Nevertheless, curcumin
pharmacological activities appear opportune intriguing a continuous search for a “super
curcumin” (2).
Various delivery systems have been proposed as a means to improve
therapuetic effects of curcumin. Among them, liposomes are able to incorporate poorly
soluble molecules and enable their aqueous medium-based administration seem to be among
the most promising delivery systems (4). Although several research groups focused on
development of liposomal delivery system for curcumin aiming at either intravenous or oral
administration (5 – 8), to the best of our knowledge only a very limited reports deal with
curcumin aiming at vaginal administration (9, 10, 11). This route enabled delivery of effective
curcumin amount directly to a diseased area when applied in delivery systems (i.e. liposomes)
that ensure satisfactory curcumin stability and solubility (10). As drug carriers liposomes have
shown good potentiality in the anticancer therapy. Owing to their advantages, antitumor
efficacy and tolerability have substantially improved while limitations pertaining to
conventional anticancer treatments (i.e. poor solubility, irritant properties, lack of stability,
rapid metabolism, non-selective drug distribution, poor patient compliance and quality of life
etc.) have elegantly been avoided (12). Curcumin and many of the anticancer drugs share
biopharmaceutical limitations thus; incorporating curcumin into liposomes seems a reasonable
approach to enhance its activity (9).
In our previous in vitro study (10), promising pharmacokinetic outcome was achieved when
curcumin was applied in liposomes to the vaginal mucosa. Depending on the liposomal size,
the concentration of curcumin in different layers of vaginal tissue was significantly higher
compared to curcumin applied in solution. Additionally, curcumin tissue retention (from
solution and liposomes) was significantly higher compared to the tissue retention of other
concurrently assayed highly and low permeable and highly soluble standards at the expense of
significantly lower permeability (i.e. curcumin permeability was approximately 1-2 log units
below the permeabilities determined for BCS1 standards). Therefore, we concluded that
curcumin shows negligible potential for systemic absorption when applied to vaginal mucosa,
eliminating the possibility of pharmacokinetic and/or pharmacodynamic interactions in comedicated patients (10).
The prerequisite for successful topical vaginal therapy is the prolonged residence time of
drug-containing formulation on vaginal site (13). In this study we thus focused on the
assessment of the importance of vaginal mucus and its effect on the limited contact time
between curcumin-containing liposomes and mucosa. For these purposes liposomes that
yielded the highest curcumin tissue retention (and the lowest curcumin tissue permeability)
(10) were coated with two bioadhesive polymers (i.e. chitosan and Carbopol) in 0.1 and 0.6%
(w/v) concentrations. Chitosan, a linear hydrophilic polymer made of copolymers of N-acetyl
glucosamine linked by β(1-4) glycosidic bonds and glucosamine has well known
mucoadhesive properties. Carbopol is the polymer of acrylic acid cross-linked with
polyalkenyl ethers or divinyl glycol and is found in various oral mucoadhesive drug delivery
systems due to its ability to interact with the mucus glycoprotein and to remain localized at a
specific site (14). The in vitro model mimicking vaginal mucus was optimized varying the
concentrations of isolated crude pig gastric mucins type II (CPGM) as a surrogate for vaginal
glycoproteins and the corresponding curcumin permeability was determined. Since vaginal
mucus is not just a simple aqueous dispersion of glycoproteins, curcumin permeability was
additionally determined through isolated bovine mucus, which’s structural and functional
properties resemble the conditions in the post-menopausal women. Finally, the strength of the
system’s bioadhesion on isolated bovine mucosa was monitored to further justify the
beneficial effects of mucoadhesive coating.
2. MATERIALS AND METHODS
2.1 Materials
Curcumin, crude pig gastric mucin type II (CPGM), and salts for the incubation saline were
from Sigma Aldrich (Diesenhofen, Germany). Human albumins were purchased at Slovenian
transfusion agency.
Chitosan® low molecular weight was obtained from Sigma-Aldrich, Chemie GmbH,
Steinheim, Germany. Lipoid S 100 (soybean lecithin, >94% phosphatidylcholine) was a gift
from Lipoid GmbH, Ludwigshafen, Germany. Carbopol®974P NF was purchased from
Noveon Inc., Cleveland, USA.
All chemicals used in this study were of the highest grade available.
2.2 Methods
2.2.1 Preparation and characterization of non-coated and coated liposomes
Liposome preparation
Liposomes containing curcumin, coated and non-coated were prepared as previously reported
(12). In brief, (PC; 200 mg) was dissolved in chloroform and mixed with curcumin (20 mg).
The solvent was evaporated using rotoevaporator system (Büchi rotavapor R-124 with
vacuum controller B-721, Büchi Vac® V-500, Büchi Labortechnik, Flawil, Switzerland) for
at least 1.5 hours at 50 mm Hg and 50 °C. The remaining film was then re-suspended in 10 ml
of PBS (pH = 7.4). Prior to coating, liposomal suspension was left overnight at 4 °C.
Coating of liposomes
The curcumin-containing liposomes were coated with 0.1% and 0.6% (w/v) chitosan solutions
prepared in 0.1% (v/v) glacial acetic acid, respectively (14). The coating was performed in the
presence of free curcumin. An equal volume of chitosan solution was added drop-wise to
liposomes under magnetic stirring at room temperature for 1 hour. The coated liposomal
suspensions were then placed in refrigerator overnight to stabilize. For the preparation of
Carbopol-coated liposomes, 0.1% and 0.6% (w/v) Carbopol®974P NF was dissolved in
phosphate buffer pH 7.4 (14). The rest of the coating procedure was as described for chitosancoated liposomes.
The entrapment efficiency was determined as the percentage of liposomally associated
curcumin in comparison to total amount of curcumin taken into the preparation and was
determined by HPLC (9).
Size distributions
Photon correlation spectroscopy (PCS) technique was employed to determine the vesicle size
and size distribution using a Nicomp model 380 particle sizing system (Nicomp Particle
Sizing Systems, Santa Barbara, CA). Samples preparation was done according to the
guidelines for particle size analysis given by the International Organization for
Standardization, and average size of liposomes was determined (15). Liposomal suspension
sample preparation was performed in a laminar airflow bench and each sample was analyzed
using data collection times (3 cycles of 5 minutes) sufficient to ensure statistically sound data
base (minimum 1000 counts in the channel). The measurements were performed in triplicates.
Zeta potential
Zeta potential measurements were performed on a Malvern Zetasizer Nano Z (Malvern,
Worcestershire, UK). The liposomal suspensions were diluted in 1:40 ratio in filtrated water
before measurements. All of the results were the average of at least three independent
measurements.
2.2.2 Buffers for in vitro experiments
Phosphate buffer pH 7.4 was prepared by dissolving NaCl (8 g/L), KCl (0.2 g/L), Na2HPO4
(1.44 g/L), and KH2PO4 (0.24 g/L) in MiliQ water and the pH was adjusted to 7.4 with 1M
HCl or 1M NaOH. On the day of the experiment human serum albumin was added to obtain
4% acceptor solution used in all in vitro permeability experiments due to poor curcumin
solubility and the tendency for adsorption to plastics. Also, albumin in the acceptor solution
was also used to maintained sink similar to in vivo situation.
For simulated vaginal fluid (SVF) NaCl (3.51 g/L), KOH (1.4 g/L), Ca(OH)2 (0.222 g/L),
acetic acid (1 g/L), lactic acid (2 g/L), glycerol (0.16 g/L), urea (0.4 g/L), and D-glucose (5
g/L) were dissolved in MiliQ water and pH was adjusted to 4.5 with 1M HCl (10).
Ringer buffer was prepared from NaHCO3 (2.1 g/L), NaH2PO4xH2O (0.055 g/L), NaCl (3.624
g/L), KCl (0.373 g/L), CaCl2x2H2O (0.176 g/L), MgCl2x6H2O (0.244 g/L), NaHPO4x2H2O
(0.285 g/L), and D-glucose (1.8 g/L) in MiliQ water and pH was adjusted with NaOH to 7.4
during bubbling with carbogen (O2/CO2 95/5) (16).
2.2.3 The in vitro permeability of curcumin
The concentrated liposome samples were sonicated prior to the experiments in an ultrasound
bath for 5-10 min. Donor solutions were prepared by diluting concentrated non-coated and
coated liposome samples with SVF to obtain identical curcumin donor (20 µM) and similar
phospholipid concentrations (0.075 – 0.082 mg/ml)). Curcumin stock solution (i.e. reference
formulation) was prepared by dissolving curcumin standard in DMSO and it was
subsequently diluted with SVF to 20µM to obtain the final DMSO concentration below 1%.
All donor solutions were protected from light and warmed to 37°C before placing them onto
Transwell filters. Donor solutions (0.5 ml of curcumin solution or dispersion of curcumincontaining liposomes) were added on top of the Transwell filters (upper compartment – donor
chamber) and 1.5 ml of acceptor solution (SVF) was applied in the lower compartment. The
experiments were done in quadruplets. Samples (150 µl) were withdrawn from the acceptor
solution every 0.5 h for up to 3 h and replaced with a fresh acceptor solution. Plates were
incubated in a light protected plate shaker with the mixing rate adjusted to 300 rpm (Vortemp
56EVC). Concentration of curcumin was determined as stated in section 2.2.7.
2.2.4 The in vitro permeability of curcumin as assessed on crude pig gastric mucus layer
Artificial mucus was prepared from the crude pig gastric mucins type II to yield glycoproteins
in 1, 3, 5, 7, and 9 % (w/w) concentration. To allow the full hydration and reconstitution of
glycoprotein microstructure, glycoproteins were suspended in SVF one day before the
experiment and left on a magnetic stirrer (speed 500 rpm) at room temperature for 3 h to
obtain a homogeneous dispersion. Glycoprotein dispersions were stored at 4 - 8°C for
maximum of 12 h. Before applying glycoprotein dispersion onto Transwell filters, they were
stirred for 1h to adjust to room temperature. The 160 µL of glycoprotein dispersion was
pipetted on the filters and mixed on a plate shaker at 37 °C for an additional hour to assure
sufficient homogeneous distribution and wetting of glycoproteins on the filter. Then the donor
solutions (0.5 ml; curcumin solution and non-coated liposomes) were placed on top and the
experiments continued as described in 2.2.3.Concentration of curcumin was determined as
stated in section 2.2.7.
2.2.5 The in vitro permeability of curcumin as assessed on cow vaginal mucus
Cow vaginal tissue was obtained immediately after animal sacrifice in a local (Slovenian)
slaughterhouse. The leakage of vaginal mucus and contamination with blood was prevented
by tying the ovarian tubes and vagina. The tissue was transported to the laboratory stored at
10°C. The vaginal wall was cut and the mucus was collected without using aggressive
mechanic techniques (i.e. scraping), which could destroy mucus microstructure. Cow mucus
was added on top of Transwell filters (160 µL) and the experiments were performed in the
manner described in 2.2.4. Concentration of curcumin was determined as stated in section
2.2.7.
2.2.6 The in vitro determination of mucoadhesion
After animal sacrifice in a local slaughterhouse vaginal tissue was preserved in the isotonic
NaCl solution. Vagina was cut open and no rinsing was applied to preserve mucus on the
surface of the tissue. Tissue that was contaminated with blood or feces was discarded. Then
the tissue was divided into smaller pieces, which were glued to the cork support by
cyanoacrylic glue. Tissue inserts were placed into the empty Transwell support and 1.5 ml of
Ringer buffer pH 7.4 with glucose was added to support tissue vitality and integrity during the
experiment. Concentrated liposome samples were added on the tissue and incubated at 37°C
in plate shaker (300 rpm) protected from the light. After 3h tissue inserts were rinsed with 10
ml of SVF containing 4% of human serum albumin and the concentration of curcumin in the
rinsed samples was determined (see 2.2.7). The % of adhered curcumin was calculated by
subtracting the amount of curcumin in the washed samples from the initial curcumin
concentration in the liposomes and normalized to initial donor curcumin concentration.
2.2.7 LC-MS/MS quantification of curcumin
Samples were precipitated with ice cold MeOH (1:3 vol/vol), vigorously vortexed and left at
−20°C for 48h. Afterwards, the samples were centrifuged (15 min at 4°C and 15000 × g), the
supernatant was transferred into autosampler 96–well plate tray for subsequent analysis by
LC-MS/MS. LC/MS/MS apparatus consisted of an Agilent 6460 triple quadrupole mass
spectrometer equipped with a JetStream interface and connected to an Agilent 1290 Infinity
UPLC (Agilent Techologies, Santa Clara, USA). For chromatographic separation, a
Phenomenex Kinetex 50 × 2.0 mm C-18 column with 2.6-mm particles was used
(Phenomenex, Torrance, USA). The injection volume was 2µL and the column temperature
was 50°C. The mobile phase consisted of a linear gradient of water (A) and acetonitrile (B),
both containing 0.1% formic acid. The gradient started with 10% B which increased linearly
to 80% B over 2 min and then returned to original conditions in 5s. The flow rate was 0.65
mL/min and the total time of analysis was 2.7 min. The mass spectrometer operated in
positive ionization mode with the following parameters: drying gas temperature: 275°C,
drying gas flow: 5 L/min, nebulizer: 45 PSI (3.1 × 105Pa), sheath gas temperature: 320°C,
sheath gas flow: 11 L/min, capillary entrance voltage: 4000 V, nozzle voltage: 1000 V, delta
EMV: 200 V. The MRM’s used for the quantification of curcumin was as follows: The
multiple reaction monitoring m/z transitions and collision energies used for the quantification
of curcumin was 369 →177 at 16 eV.
2.2.8 Data analysis
The apparent permeability coefficient (Papp) of curcumin through empty filter and mucus
layers was calculated according Eq. (1).
𝑑𝑐
𝑉
𝑃𝑎𝑝𝑝 = 𝑑𝑡 × 𝐴×𝑐
Eq. (1)
0
where dc/dt represent changes in concentration of the curcumin in the acceptor compartment
per unit time under steady state conditions, V is the volume of the acceptor compartment, A
the exposed surface area, and c0 the initial curcumin concentration.
To correct for the filter support, true curcumin permeabilities through mucus were calculated
according to the Eq. (2).
1
1
𝑒𝑥𝑝𝑒𝑟𝑖𝑚𝑒𝑛𝑡
𝑃𝑎𝑝𝑝
= 𝑃𝑚𝑢𝑐𝑢𝑠 +
𝑎𝑝𝑝
1
𝑓𝑖𝑙𝑡𝑒𝑟
𝑃𝑎𝑝𝑝
Eq. (2)
Where Pappexperiment represent average Papp values determined for each experiment and Pappfilter
represents average values determined for curcumin diffusion through an empty Transwell
filter. The error (sp – standard deviation) for Pappmucus was calculated as presented in Eq. (3)
2
2
(𝑛𝑓𝑖𝑙𝑡𝑒𝑟 −1)×𝑠𝑓𝑖𝑙𝑡𝑒𝑟
+(𝑒𝑥𝑝−1)×𝑠𝑒𝑥𝑝
𝑠𝑝 = √
𝑛𝑓𝑖𝑙𝑡𝑒𝑟 +𝑛𝑒𝑥𝑝 −1
Eq. (3)
Where nfilter, and nexp represent the number of replicate samples (usually 4) in experiments
using empty filter or different mucus, and sfilter, sexp standard deviations determined in the
corresponding experiments.
Results in Tables and Figures are presented as means ± standard deviations.
Statistical assessment was performed with one-way ANOVA using SPSS 17.0 (SPSS Inc.,
Chicago IL).
3. RESULTS
3.1 Liposome characteristics
Liposomes incorporated high curcumin content as previously reported by us (9, 10). The
coating in the presence of non-incorporated curcumin did not significantly increase the
incorporation efficiency, as the efficiency was already high. The coating with Carbopol
solutions yielded lower incorporation of curcumin as compared to the coating with chitosan
solutions (Table 1). We were not able to determine the differences in the entrapment
efficiency with respect to the polymer concentrations used in coating (0.1 and 0.6 %).
However, the 0.6 % Carbopol solution is rather viscous and can lead to the agglomeration of
coated vesicles (data not shown). We were clearly dealing with MLVs with rather large
polydispersities and their average size is therefore an overestimate.
3.2 The release and diffusion of liposomal curcumin through an empty Transwell filter
Diffusion of curcumin from solution through empty Transwell insert was found to be
significantly higher (p < 0.0001) at all-time points compared to the diffusion from
investigated liposomes (Figure 1), indicating that liposomes retained associated curcumin to a
greater extent. Also, the values for diffusion of curcumin in solution form suggest that there
was no significant resistance by the filter towards curcumin. There were no significant
differences in the release and/or diffusion rates between non-coated, chitosan (0.1 and 0.6 %)
and 0.1% Carbopol-coated liposomes, while 0.6% Carbopol-coated liposomes provided the
slowest curcumin release and subsequent diffusion.
3.3 In vitro permeability of curcumin through artificial and isolated bovine mucus
Permeabilities of curcumin through different mucus layers (artificial mucus – dispersions of
CPGM and isolated cow mucus) are presented in Figure 2.
A negative correlation between curcumin permeability from solution and content of
glycoproteins in the mucus layer was observed. Permeability was the highest (statistically
significant, p < 0.05) through 1 w/v% CPGM dispersions, followed by statistically lower
permeabilities determined through 3, 5, and 7 w/v% CPGM and the lowest (p < 0.05) through
mucus containing 9w/v% of CPGM. Changing the content of CPGM in the artificial mucus
had no significant effect on the curcumin permeability from non-coated liposomes, although a
non-significant trend of negative correlation was noted. When chitosan-coated liposomes
were incubated, a bell-shaped relationship between curcumin permeability and CPGM content
was detected. The highest permeability values were measured through dispersion of CPGM
with 5 w/v% glycoproteins. Increasing the amount of chitosan in the liposomal coating had no
beneficial impact on the permeability. Namely, permeability from 0.6% chitosan-coated
liposomes was similar or even lower than that from liposomes coated with 0.1% chitosan. In
the case of Carbopol-coated liposomes, the curcumin permeability was found to be directly
proportional to the content of glycoproteins (i.e. the higher the glycoprotein content, the
higher the permeability). The highest permeability values were measured when 7 and 9 w/v%
CPGM were dispersed in SVF and the formulation containing higher amount of Carbopol
provided slightly higher curcumin permeabilities at these higher glycoprotein concentrations
(Figure 2).
When isolated bovine mucus was applied onto Transwell filters, curcumin permeabilities
from all systems (i.e. solution and liposomes) were significantly higher than any of the
corresponding values determined through the artificial mucus. Based on the in vitro
permeabilities, the following rank order was established: 0.6% coated liposomes > solution 
0.1% coated liposomes > non-coated liposomes.
3.4 In vitro bioadhesion
The results from in vitro measurements of the bioadhesion strength with the corresponding
statistics are presented in Table 2. During experiments we observed unavoidable leakage of
non-coated liposomes from the surface of vaginal tissue into the incubating Ringer buffer
below the tissue inserts. Vaginal tissues were thus cut into the pieces in a way that the center
of slices, where liposomes were applied, was positioned lower than the edges of tissue slices
to prevent unintentional leakage. Despite these attempts, we could not completely prevent the
loss of non-coated liposomes during 3 h-incubation period. Thus, the values for the % of the
adhered curcumin presented in Table 2 for non-coated liposomes can be an overestimate and
needs to be taken into consideration. Nevertheless, based on the one-way ANOVA the type of
liposomes clearly affected the strength of bioadhesion. Except for the 0.1% Carbopol-coated
liposomes, bioadhesive polymers increased the retention of curcumin delivery systems
compared to non-coated liposomes (Table 2). A positive correlation between the % of
bioadhesive polymer used in liposome coating and % of the retained curcumin was observed.
The post-hoc comparison between Carbopol- and chitosan-coated liposomes indicated
significantly higher (p < 0.05) bioadhesiveness provided by 0.6% chitosan-coated liposomes.
4. DISCUSSION
Delivery of curcumin to vaginal site may provide an advantageous alternative to
oral/intravenous delivery of curcumin (9). Although numerous delivery systems have been
proposed as mean to improve bioavailability of curcumin (4), only two systems, namely
cream (11) and liposomes (9, 10) were reported for the vaginal administration. The fact that
liposomes offer the potential to solubilize poorly soluble curcumin, and their coating the mean
to improve the retention time at the administration site, guided us in development of the
mucoadhesive liposomal delivery systems for curcumin. The success of non-invasive drug
delivery via vaginal mucosa for the purposes of local or systemic therapy depends on the
interplay between the local conditions (in vivo), the drug and drug carrier system and their
physico-chemical properties. While certain physiological characteristics (i.e. large surface
area, abundant vascularization) favor drug absorption, cyclical fluctuations of estrogens and
progesterone plasma levels influence the conditions in the vaginal tract (i.e. the amount, pH,
and viscosity of vaginal mucus, thickness of vaginal mucosa, composition of vaginal
microflora) and thus prevent reliable and repeatable behavior of conventional dosage forms
(17, 16). Mucoadhesive drug delivery systems thus offer a potential for improvement in
therapy outcome (13) because the retention time at the affected/diseased area is increased,
allowing more drug to be released from the formulation and consequently increasing the
probability of more drug diffusing to the epithelial cells.
Vaginal mucus is a complex mixture of exudates from fallopian tubes, uterus, and cervix
mixed with blood transudate, desquamated cells, vaginal bacteria and other less represented
constituents with more specialized functions (i.e. enzyme inhibitors) (17). Based on published
data the production of vaginal mucus depends on hormonal balance, accompanying diseases,
sexual activity, co-medication etc. and is estimated to be 1 – 11 ml per day in the reproductive
period, followed by a 50% decline in postmenopausal women (18). In this study, a worst-case
in vivo scenario was simulated to monitor the behavior of liposomes in vitro. Therefore, the
amount of mucus applied on Transwell filters (150 L) was selected based on the data of
daily mucus production in the reproductive age (11 ml/day) and estimated surface area of the
vaginal tract (the height 7.5 – 12 cm, the diameter 4 cm) (19). The content of glycoproteins in
the simulated mucus was adjusted based on published data on the mucosal content of water
and glycoproteins in the reproductive stage. Namely, depending on the plasma levels of sex
hormones, the content of water (90 – 99 w/w %) increases towards the middle of the cycle,
corresponding to the ovulation (98 w/w % of water) and gradually declines to the original
values at the end (20). Glycoproteins (0.5 – 5 w/w %) determine the tridimensional
microstructure of the mucus. As with water, their quantitative composition depends on the
hormonal status with gradual increase in protein content during the menstrual cycle (21, 22),
while the qualitative composition remains insensitive to hormonal status and thus unchanged
(i.e. there are always the same glycoproteins present in the vaginal mucus such as MUC4,
MUC5A/C and MUC8) (23, 24). Based on these facts, five different mucosal dispersions
were prepared from isolated gastric pig glycoproteins (CPGM) in 1, 3, 5, 7, and 9 w/v % and
curcumin permeability through these layers was determined in vitro. The 7 and 9 % (w/v)
glycoprotein concentrations were tested to compensate for the changes in water content
during the menstrual cycle to values below 95 w/w%; there are no data available on the
preferable concentrations of other mucosal constituents mimicking the post-estrogen phase,
when the lowest water content is expected (17, 18).
There are no other commercially available glycoproteins of vaginal origin. Thus, curcumin
permeability was additionally monitored through the isolated bovine mucus, since it exerts
more complex composition than the pig gastric mucins in dispersion of SVF. It is also
expected that isolated bovine mucus retains its microstructure and shows several similar
properties as the mucus in the postmenopausal women (i.e. pH, mixture of more mucosal
components than just water and glycoproteins) (25, 26).
Permeability values obtained for curcumin solution and non-coated liposomes (determined at
20 µM curcumin concentrations) and the strength of bioadhesion for non-coated liposomes
served as reference values. While diffusion of curcumin through vaginal mucus is limited by
the equilibrium between the free and the protein-(i.e. glycoproteins) bound curcumin (10), an
additional delay in curcumin delivery to the vaginal mucosa when administered as liposomes
may be expected. Namely, the release kinetics of loaded curcumin from the phospholipid
milieu and through Carbopol/chitosan coating lags the onset of tissue contact with curcumin.
Additionally, mucosal microstructure may act as a size-excluding sieve for bigger liposomes
(Table 1; (10)), therefore increasing the distance for curcumin to diffuse through.
Nevertheless, increased solubility and curcumin stability ensured by liposomes supported by
the increased retention time on the target site could outweigh retarded curcumin release (see
results for curcumin diffusion through an empty Transwell filter – Figure 1).
As shown in Figure 2, curcumin concentration versus time profile in the vaginal mucosa
would be insensitive to cyclic changes of the physiological conditions inside vaginal tract
(simulated with different concentrations of CPGM) with non-coated liposomes. However, this
type of liposomes provided the lowest curcumin permeability through the artificial and
insolated bovine mucus and the poorest bioadhesion strength (Table 2). When vaginal mucus
was simulated by CPGM suspension, the aqueous solution of curcumin enabled the highest
permeability only through 1 w/w % glycoprotein, while with higher glycoprotein content, the
permeabilities from chitosan- (0.1 %) and Carbopol- (0.6 %) coated liposomes surpassed
those determined for the solution. Additionally, 0.6% chitosan- and 0.6% Carbopol-coated
liposomes enabled the highest curcumin permeability through the isolated bovine mucus.
These findings, coupled with the improved retention time at the site of administration as
assessed with the strength of bioadhesion in vitro, make bioadhesive liposomes the superior
drug delivery system for curcumin. Beneficial effects of bioadhesive coating for curcumin
pharmacokinetic behavior are in line with publications on drug delivery systems containing
the same polymers – chitosan-coated liposomes improved drug permeability owing to its
penetration enhancing effects (noticed on non-vaginal systems – through buccal (27), ocular
(28), nasal (29), and intestinal (30)) and mucoadhesion properties, while Carbopol-coated
formulations prolonged the retention time of per-orally administered dosage forms in pig
gastric mucosa in vitro (31), and increased paracellular drug diffusion by calcium chelation
(32, 33).
Dispersion of CPGM as an alternative to native pig gastric mucus, which normally contains
only 5 w/v % of mucins, has been used also for investigating the permeability of acyclovir
derivatives in vitro (34). The rigidity and a gel-like structure of the native gastric mucus were
simulated with 30 and 50 w/v% of CPGM, respectively. Although good correlation between
permeabilities determined through CPGM dispersion and through the native mucus were
obtained, Legen et al., (34) observed that the permeabilities through CPGM dispersion were
significantly lower than the corresponding values through the native mucus, which is in
accordance with our observations (Figure 2 - significantly higher curcumin permeabilities
through isolated bovine mucus than through CPGM dispersions). Failure to reconstitute gel
structure of native mucus in vitro with CPGM has been postulated as the main reason for the
observed discrepancies in absolute permeability values obtained with both types of mucus
layers. Although our study mimicked in vivo glycoprotein content with similar in vitro
glycoproteins concentrations, the interactions of curcumin with CPGM dispersions appear to
be much more pronounced than those with isolated bovine mucus. The strength of
bioadhesion depends on the concentration of both parties involved in the process of polymer
side-chain inter-penetration, therefore, further studies evaluating the optimal amount of
bioadhesion polymers in the liposomal coating on the permeability/bioadhesion strength
should be warranted. Namely, increasing the content of bioadhesive polymer in the coating
should increase the strength of liposome adhesion on the mucus and curcumin permeability up
to the optimal values. After exceeding the optimal concentration of bioadhesive polymer, a
decline in the abovementioned parameters is expected due to the limited surface of polymer
available for interaction with the surrounding mucosal glycoproteins. The choice of the
polymer type and concentration was based on the previous experiences with both polymers
(14). Additionally, an increase in the liposomal size, as well as change in zeta potential, are
expected with increasing amounts of the coating, therefore the major part of the polymer
would remain hindered in the deeper layers, therefore not participating in the bioadhesion
process until the uppermost polymer sheets would be removed by erosion or by detachment
from mucus (32, 33). A combination of chitosan and Carbopol, or introduction of other
mucoadhesive polymers such as pectin, might yield superior properties to the ones observed
in the current study.
5. CONCLUSION
Bioadhesive liposomes were shown to be a superior drug delivery system for curcumin
targeting vaginal mucosa. Both chitosan and Carbopol coatings significantly increased
curcumin permeability through vaginal mucus in vitro compared to curcumin solution and
non-coated liposomes. That, accompanied by improved formulation retention on the bovine
mucosa (as assessed by in vitro bioadhesion) would make liposomes as carriers for curcumin
a superior drug delivery system over conventional vaginal dosage forms.
6. FIGURES
2 .0  1 0 -09
S o lu tio n
C u r c u m in ( m o l)
N o n - c o a te d
1 .5  1 0 -09
0 .1 % C h ito s a n
0 .6 % C h ito s a n
1 .0  1 0 -09
0 .1 % C a r b o p o l
0 .6 % C a r b o p o l
5 .0  1 0 -10
0
0
1
2
3
tim e ( h )
F ig u r e 1 : T h e r e le a s e a n d d iffu s io n o f c u r c u m in fr o m ( n o n ) - c o a te d lip o s o m e s a n d
s o lu tio n th r o u g h e m p ty T r a n s w e ll filte r s .
1 .0  1 0
-5
0
1 .0  1 0
-5
1 .0  1 0
-5
m u c u s c o m p o s itio n
9
%
c
o
w
L
G
L
G
L
9
%
c
o
w
L
G
L
G
c
o
w
L
9
%
G
L
7
%
G
L
5
%
G
L
%
3
o
c
G
L
w
L
9
%
G
L
G
7
%
G
5
%
3
%
G
L
G
%
1
L
0
L
0
-5
G
-5
2 .0  1 0
%
1 .0  1 0
3 .0  1 0
-5
1
-5
0 .6 % C a rb o p o l
P a p p ( c m /s )
P a p p ( c m /s )
2 .0  1 0
%
m u c u s c o m p o s itio n
0 .1 % C a rb o p o l
3 .0  1 0
7
5
m u c u s c o m p o s itio n
-5
L
%
G
L
%
1
9
%
c
o
G
w
L
G
L
7
%
G
G
5
3
%
%
G
L
G
%
1
L
0
L
0
L
-5
-5
G
1 .0  1 0
2 .0  1 0
%
-5
3 .0  1 0
-5
3
2 .0  1 0
0 .6 % C h i t o s a n
P a p p ( c m /s )
P a p p ( c m /s )
0 .1 % C h i t o s a n
3 .0  1 0
7
1
m u c u s c o m p o s itio n
m u c u s c o m p o s itio n
-5
%
%
G
L
%
c
o
G
w
L
G
L
%
9
7
%
G
L
G
%
5
%
3
1
%
G
G
L
L
0
L
-5
-5
5
1 .0  1 0
2 .0  1 0
G
-5
%
2 .0  1 0
3 .0  1 0
-5
3
3 .0  1 0
N o n -c o a te d
P a p p ( c m /s )
P a p p ( c m /s )
S o lu t io n
-5
m u c u s c o m p o s itio n
F ig u r e 2 : C u r c u m in in v itr o p e r m e a b ility f r o m s o lu t io n a n d ( n o n ) - c o a te d lip o s o m e s t h r o u g h d if f e r e n t
m u c u s la y e r s ( 1 , 3 , 5 , 7 , a n d 9 w /v % C P G M o r is o la t e d b o v in e m u c u s - c o w ) .
7. TABLES
Table 1: Characteristics of non-and bioadhesive liposomes (n=3)
Type of liposomes
Size (nm)
Zeta potential
Curcumin concentration
(mV)
(μg/ml)
Non-coated
920*
2.09 ± 0.71
175.4 ± 4.0
Chitosan-coated (0.1 %)
1280*
7.01 ± 0.91
92.0 ± 2.6
Chitosan-coated (0.6 %)
1320*
20.09± 2.71
93.0 ± 3.7
Carbopol-coated (0.1 %)
1180*
-10.45± 1.91
83.6 ± 3.1
Carbopol-coated (0.6 %)
1260*
-15.33 ± 2.01
81.8 ± 2.9
The size is an estimate as the polydispersity index was between 0.7 and 0.9.
Table 2: The in vitro bioeadhesion of liposomally-associated curcumin
Non-coated
0.1%
0.6%
0.1% Carbopol 0.6% Carbopol
(Nc)
chitosan
chitosan
(Ca-1)
(Ca-6)
(Ch-1)
(Ch-6)
87.7 ± 7.4
89.6 ± 6.6
90.8 ± 4.8
97.5 ± 3.6
𝑥̅ ± 𝑆𝐷 (%)
81.2 ± 7.8
ANOVA
Lower than Lower than Higher than Higher than Nc Higher
post-hoc test
Ca-1, Ca-6, Ca-6
Nc
Ch-6
SD – standard deviation; 𝑥̅ – average

- Only statistically significant outcomes are reported (p < 0.05)
than
Nc, Ch-1, Ch-6
8. REFERENCES
(1) S. Shishodia, Molecular mechanisms of curcumin action: Gene expression. BioFactors, 39
(2013) 37–55.
(2) V. Anand, M. Gulati, P. Rastogi, J. Dixit, M.-A. Khan, N. Lal, Curcumin: A therapeutic
approach: A review, Spatula DD, 2 (2012) 117-125.
(3) M. Metzler, E. Pfeiffer, S.-I. Schulz, J.-S. Dempe, Curcumin uptake and metabolism., Int.
Union Biochem. Mol. Biol. 39 (2013) 14–20.
(4) P. Basnet, N. Škalko-Basnet, Curcumin: An anti-inflammatory molecule from a curry
spice on the path to cancer treatment, Molecules 16 (2011) 4567-4598.
(5) C. Chen, T.D. Johnston, H. Jeon, R. Gedaly, P.P. McHugh, T.G. Burke, D. Ranjan,
An in vitro study of liposomal curcumin: Stability, toxicity and biological activity in human
lymphocytes and Epstein-Barr virus-transformed human B-cells. Int. J. Pharm. (2009) 366,
133-139.
(6) L. Li, B. Ahmed, K. Mehta, R. Kurzrock, Liposomal curcumin with and without
oxaliplatin: Effects on cell growth, apoptosis, and angiogenesis in colorectal cancer. Mol.
Cancer Ther. (2007) 6, 1276-1282.
(7) R.L. Thangapazham, A. Puri, S. Tele, R. Blumenthal, R.K. Maheshwari, Evaluation of a
nanotechnology-based carrier for delivery of curcumin in prostate cancer cells. Int. J. Oncol.
(2008) 32, 1119-1123.
(8) N.K. Narayanan, D. Nargi, C. Randolph, B.A. Narayanan, Liposome encapsulation of
curcumin and resveratrol in combination reduces prostate cancer incidence in PTEN knockout
mice. Int. J. Cancer (2009) 125, 1-8.
(9) P. Basnet, H. Hussain, I. Tho, N. Škalko-Basnet, Liposomal delivery system enhances
anti-inflammatory properties of curcumin, J. Pharm. Sci.101 (2012) 598-609.
(10) K. Berginc, N. Škalko-Basnet, P. Basnet, A. Kristl, Development and evaluation of an in
vitro vaginal model for assessment of drug’s biopharmaceutical properties: Curcumin, AAPS
PharmSciTech 13 (2012) 1045-1053.
(11) P. R. Debata, M.R. Castellanos, J.E. Fata, S. Baggett, S. Rajupet, A. Szerszen, S. Begum,
A. Mata, V.V. Murty, L.M. Opitz, P. Banerjee, A novel curcumin-based vaginal cream
Vacurin selectively eliminates apposed human cervical cancer cells, Gynecologic Oncology
(2013) 129, 145 -153.
(12) M. Slingerland, H.-J. Guchelaar, H. Gelderblom. Liposomal drug formulations in cancer
therapy: 15 years along the road. Drug Discovery Today 17 (2012) 160–166.
(13) Ž. Vanić, N. Škalko-Basnet,
Nanopharmaceuticals for improved topical vaginal
delivery: Can they deliver? Eur. J. Pharm. Sci. 50 (2013) 29-41.
(14) P. R. Karn, Ž. Vanić, I. Pepić, N. Škalko-Basnet, Mucoadhesive liposomal delivery
systems: the choice of coating material, Drug Devel. Ind. Pharm. (2011) 37, 482–488.
(15) M. do Cagno, J. Styskala, J. Hlavač, M. Brandl, A. Bauer-Brandl, N. Škalko-Basnet,
Liposomal solubilization of new 3-hydroxy-quinolinone derivatives with promising
anticancer activity: a screening method to identify maximum incorporation capacity, J. Lip.
Res. (2011) 21, 272 – 278.
(16) K. Berginc, J. Trontelj, N. Škalko-Basnet, A. Kristl, Physiological barriers to the oral
delivery of curcumin, Die Pharmazie, 67 (2012) 518-524. (17) R.-R. de Araujo Pereira, M.-L.
Bruschi, Vaginal mucoadhesive drug delivery systems. Drug Dev. Ind. Pharm. 38 (2011) 643652.
(17) G.-H. Deutcher, G80-537 Reproductive trace anatomy and physiology of the cow.
Historical materials from University of Nebrasca-Lincoln Extension. 1980, paper 315.
(18) G. Preti, G.-H. Huggins, Cyclical changes in volatile acidic metabolites of human vaginal
secretions and their relation to ovulation, J. Chem. Ecol. 3 (1975) 361-376.
(19) U. Hani, R.-S. Bhat, R. Sisodiya, H.-G. Shivakumar, Novel vaginal drug delivery
systems. A review. Curr. Drug Ther. 5 (2010) 95-104.
(20) K. Khanvilkar, M.-D. Donovan, D.-R. Flanagan, Drug transfer through mucus. Adv.
Drug Del. Rev. 48 (2001) 173-193.
(21) D.-P. Wolf, J.-E. Sokoloski, M. Litt, Composition and function of human cervical mucus.
Biochim. Biophy. Acta 630 (1980) 545-558.
(22) R.-J. van Kooij, H.-J. Roelofs, G.-A. Kathmann, M.-A. Kramer, Human cervical mucus
and its mucous glycoprotein during the menstrual cycle. Fertil. Steril. 34 (1980) 226-233.
(23) I. Carlstedt, H. Lindgren, J.-K. Sheehan, U. Ulmsten, L. Wingerup, Isolation and
characterization of human cervical-mucus glycoproteins. Biochem. J. 211 (1983) 13-22.
(24) E. Baloglu, Z.-A. Senyigit, S.-Y. Karavana, A. Bernkop-Schnurch, Strategies to prolong
the intravaginal residence time of drug delivery systems. J. Pharm. Pharm. Sci. 12 (2009),,
312-336.
(25) G.-H. Deutscher, G80-537 Reproductive trace anatomy and physiology of the cow.
Historical materials from University of Nebrasca-Lincoln Extension. 1980, paper 315.
(26) H. Takeuchi, Y. Matsui, H. Yamamoto, Y. Kawashima, Mucoadhesive properties of
carbopol or chitosan-coated liposomes and their effectiveness in the oral administration of
calcitonin to rats, J. Control. Release 86 (2003) 235-242.
(27) S. Senel, M.-J. Kremer, S. Kas, P.-W. Wertz, A.-A. Hincal, C.-A. Squier, Enhancing
effect of chitosan on peptide drug delivery across buccal mucosa. Biomaterials. 21 (2000)
2067-2071.
(28) G. Di Colo, Y. Zambito, S. Burgalassi, I. Nardini, M.-F. Saettone, Effect of chitosan and
of N-carboxymethylchitosan on intraocular penetration of topically applied ofloxacin, Int. J.
Pharm. 273 (2004) 37-44.
(29) R. Fernandez-Urrusuno, P. Calvo, C. Remunan-Lopez, J.-L. Vila-Jato, M.-J. Alonso,
Enhancement of nasal absorption of insulin using chitosan nanoparticles, Pharm. Res. 16
(1999) 1576-1581.
(30) M. Thanou, J.-C. Verhoef, H.-E. Junginger, Chitosan and its derivatives as intestinal
absorption enhancers, Adv. Drug Deliv. Rev. 50 (2001) S91-S101.
(31) V. Grabovac, D. Guggi, A. Bernkop-Schnurch, Comparison of the mucoadhesive
properties of various polymers. Adv. Drug Deliver. ver. 57 (2005) 1713-1723.
(32) A. Ahuja, R.-K. Khar, J. Ali, Mucoadhesive drug delivery systems, Drug Dev. Ind.
Pharm. 23 (1997) 489-515.
(33) F.-J.-O. Varum, E.-L. McConell, J.-J.-S. Sousa, F. Veiga, A.-W. Basit, Mucoadhesion
and the gastrointestinal tract. Crit. Rer. Ther. Drug Carrier Syst. 25 (2008) 207-258.
(34) I. Legen, A. Kristl, Comparative permeability of some acyclovir derivatives through
native mucus and crude mucin dispersions, Drug Dev. Ind. Pharm. 27 (2001) 669-674.