Key words
Crocus sativus
- Iridaceae - solid lipid nanoparticles - Turbiscan analysis - antioxidant - cytotoxic
Abbreviations
AAPH:
2,2′-azobis(2-amidinopropane) dihydrochloride
CRO:
crocin
CRT:
crocetin
DR:
drug recovery
DSC:
differential scanning calorimetry
FL:
fluorescein
HPMC:
(hydroxypropyl) methyl cellulose
ORAC:
oxygen radical absorbance capacity
PCS:
photon correlation spectroscopy
PDI:
polydispersivity index
QESD:
quasi-emulsion solvent diffusion
SLNs:
solid lipid nanoparticles
TSI:
Turbiscan stability index
Introduction
Saffron, dry stigmas of Crocus sativus L. (Iridaceae) flowers, has been studied for many therapeutic applications including
antioxidant and cytotoxic properties [1], [2]. This spice contains bioactive constituents such as picrocrocin, CRO, and CRT, which
belong to the family of carotenoids [3].
Carotenoids are a group of lipophilic pigmented natural substances formed by carotenes
and their derivatives. The basic structure of carotenoids is made up of eight isoprenoids
combined with a series of double-bonded conjugates. Carotenoids can have vitaminic
activity, and can perform an immunomodulatory function in relation to the strong chain-breaking
antioxidants capacity. They act as free radical scavengers by donating hydrogen or
transferring a single electron to radicals. Through this action, carotenoids can strengthen
the immune system and also block the progress of precancerous diseases.
CRO and CRT ([Fig. 1]), in particular, possess interesting biological activities. CRT is an amphiphilic
low molecular weight carotenoid showing excellent antioxidant, anti-atherosclerotic,
anti-inflammatory, and anticancer activities in vitro and in vivo
[4], [5]. CRO, the water-soluble derivative of CRT, has high antioxidant activity [6], [7], [8] and shows beneficial effects on many organs such as the nervous, cardiovascular,
and immune systems and against cancer [9], [10], [11], [12], [13].
Fig. 1 Chemical structures of crocin and crocetin.
In spite of such beneficial activities of both CRT and CRO, their use is limited because
of their unfavorable physical and chemical features. CRO, for instance, is scarcely
stable. Indeed, it loses most of its functionality after exposure to heat, oxygen,
light, and acids [14], [15]. It shows a poor absorption and low bioavailability after oral administration, being
hydrolyzed by β-glucosidase and rapidly eliminated [16]. On the contrary, the biological application of CRT is mainly limited due to its
hydrophobic nature and poor aqueous solubility [17].
Nanoscale drug delivery systems are an important tool to overcome the aforementioned
limitations and to improve the pharmacokinetics and bioavailability of drugs and natural
active compounds [18].
SLNs, which are members of the lipid nanocarriers family, are able to solubilize hydrophilic
and lipophilic molecules in a physiologic environment, controlling their delivery
and protecting them from degradation [19]. Compared with other systems, SLNs have many benefits, such as ease of preparation
and low cost of raw materials, high-scale production, excellent physical stability,
and biodegradability of lipids [20]. These favorable properties of SLNs make them promising nanocarriers to overcome
or at least minimize some of the drawbacks of conventional formulations.
A general recognized limit of SLNs is represented by the entrapment of a hydrophilic
drug into the lipid matrix of the nanoparticle. This condition is often characterized
by a low payload and a scarce bioacceptability due to the use of organic solvents
during the formulation [21].
The aim of this study was the preparation and characterization of CRO and CRT containing
SLNs, and the evaluation of the antioxidant and antiproliferative effects of these
carotenoids once incorporated into lipid nanoparticles. For this aim, the ORAC assay
and the MTT test were used, respectively.
CRO-loaded SLNs, in spite of the hydrophilic character of CRO, were formulated by
using a modification of the QESD technique, a particular approach to the general solvent-change
method [22].
Results and Discussion
The methods used to formulate CRO- and CRT-loaded SLNs have proven to be valid and
highly reproducible despite the instability of the molecules. CRO-SLNs and CRT-SLNs
were prepared with Softisan 100 (hydrogenated coco-glycerides) and Pluronic F68 (poloxamer
188) as solid lipid and surfactant, respectively. We decided to use hydrogenated coco-glycerides
after a screening to identify the best lipid matrices for CRO and CRT incorporation.
These lipids are characterized by a low melting point (35 °C), a very important feature
to prepare the SLNs, thus avoiding thermal stress conditions. The formulation strategies
we used gave interesting results as confirmed by PCS analysis and drug loading determination
([Table 1]). SLNs prepared by the multiple emulsion technique showed a mean diameter of 270 nm,
a PDI value around 0.15, and a DR% of 82.1, while CRT-SLNs, formulated by the solvent
diffusion technique, showed acceptable results regarding dimensional analyzes: the
mean diameter was around 280 nm and the PDI value around 0.30, while the DR was about
94% ([Table 1]). As regards to zeta potential values, we obtained − 17.8 mV for CRT-SLNs and − 30.2
for the CRO-SLNs, predicting a good long-term stability for both formulations ([Table 1]).
Table 1 Particle size, polydispersivity index, and zeta potential values of different SLN-based
formulations.
|
Sample
|
Particle size (nm, mean ± SD)
|
Polidispersivity index (mean ± SD)
|
Zeta potential (-mV, mean ± SD)
|
Drug recovery (DR %)
|
|
Unloaded CRO-SLNs
|
248.6 ± 19.80
|
0.289 ± 0.02
|
19.3 ± 1.54
|
–
|
|
CRO-SLNs
|
273.3 ± 21.80
|
0.158 ± 0.01
|
30.2 ± 2.41
|
82.1
|
|
Unloaded CRT-SLNs
|
231.8 ± 18.54
|
0.258 ± 0.02
|
40.0 ± 3.20
|
–
|
|
CRT-SLNs
|
278.1 ± 22.20
|
0.305 ± 0.02
|
17.8 ± 1.42
|
94
|
In addition, we assessed the physical stability of all the samples using an optical
analyzer Turbiscan AGS, previously used to evaluate the stability of vesicles and
nanoparticle dispersion [23]. On the basis of TSI (Turbiscan stability index) values, we could infer that drug
incorporation increases the physical stability of SLNs. In fact, as shown in [Fig. 2], both CRO- and CRT-loaded SLNs have TSI values lower than the unloaded one.
Fig. 2 TSI values of unloaded (grey ■) and crocin-loaded SLNs (black ■) and unloaded (grey
●) and crocetin-loaded SLNs (black ●) stored at 25 °C.
[Fig. 3] shows the DSC data of the different raw materials. Blank and drug-loaded SLNs were
also prepared and characterized for comparison. Bulk Softisan 100 showed a sharp endothermic
event, due to the melting, around 42 °C. When this material was formulated as SLNs,
it appears broader (panel A) or in a sharp shape (panel B) and happened at a slightly
higher temperature. These differences are ascribed to the different composition of
the nanocarriers and to the experimental procedures employed to formulate them. The
loading of CRO and CRT seems to not provoke any considerable effect in the lipid matrix
thermal behavior.
Fig. 3 DSC thermograms of unloaded SLNs and crocin or crocetin- loaded SLNs.
[Fig. 4] shows the in vitro release profile of CRO and CRT from SLNs. The release profiles of the two drugs from
the nanoparticles show a similar trend and follows the well-known “two-step drug release”
related to distribution of active compounds in the SLN matrix [24]. At first, the surface bounded drug and the drug that is released immediately because
of a large diffusion in concentration gradient represents the burst release. The second
phase and the subsequent prolonged release could be due to the slow release of the
drug from the lipid core.
Fig. 4 In vitro release profiles of crocin and crocetin from CRO-SLNs and CRT-SLNs, respectively.
Each point represents the mean value ± SD (n = 3).
As previously reported, the release profiles of the two drugs are very similar, such
as the amount of drug released from the vehicles, about 60% of the loaded drug for
CRT and 68% for CRO. These small differences are probably due to the different chemical
characteristics of the two drugs and to the different procedures used to formulate
CRO-SLNs and CRT-SLNs.
In this work, we evaluated the antioxidant effect of CRO and CRT alone and once loaded
in SLNs. [Fig. 5] shows the results of the ORAC assay that we used as an analytical method to determine
the antioxidant potential of these carotenoids and the effect of SLNs in modulating
their activity. The first evidence we obtained from the results of the ORAC assay
is that CRT showed an antioxidant efficacy higher than CRO. We observed these results
although CRO and CRT belong to the same family and show a similar structure. We hypothesize
that the presence of the gentobiose sugar in the molecular structure of CRO modifies
the conjugation degree of the active compound and consequently could influence its
antioxidant activity ([Fig. 5]). Once encapsulated in SLNs, both CRO and CRT are able to preserve high fluorescence
compared to empty SLNs or the positive control, demonstrating that the antioxidant
activity of these active compounds is maintained for a longer time when it is associated
to SLNs. It is noteworthy that CRT guaranteed a higher antioxidant activity compared
to CRO, also when encapsulated in SLNs. This evidence demonstrated that SLNs are able
to control the release of these active compounds, but they did not interfere in modifying
their intrinsic activities.
Fig. 5 ORAC assays for CRO-SLN and CRT-SLN. Empty SLN-1 and empty SLN-2 represent the blank
of CRO-SLN and CRT-SLN, respectively. Data represent the mean of three independent
experiments ± SD; *p < 0.05. FL: Fluorescein; AAPH: 2,2′-Azobis(2-amidinopropane)
dihydrochloride; SLN: solid lipid nanoparticles; CRO: crocin; CRT: crocetin.
The in vitro cytotoxic effect of CRO and CRT was evaluated by the MTT assay. We choose the cancer
cell lines A375 (human melanoma) and sNF96.2 (malignant Schwann) since their use was
supported by scientific evidences that demonstrated the important role of CRO and
CRT in preventing and/or ameliorating some skin tumors [25] and soft tissue sarcomas in in vivo models [26]. The two cancer cells, A375 and SNF96.2, were treated with blank and drug-loaded
formulations, and incubated for 72 h. From the analysis of the data, the blank carriers
did not have any important effect on cell viability, indicating the nontoxic nature
of both SLN-based carriers (SLN1 and SLN2). Free drugs and SLN-based formulations
showed a different antiproliferative effect against A375 and SNF96.2 cell lines ([Figs. 6] and [7]). In particular, free CRO failed to induce an important cytotoxic effect with respect
to control formulation for both cell lines, while CRT exhibited an extremely significant
toxic effect in the same experimental conditions ([Figs. 6] and [7]).
Fig. 6 Antiproliferative activity of CRO-SLN and CRT-SLN against a human melanoma cell line
(A375) versus CRO and CRT solution in PBS (0.1% w/w). Empty SLN-1 and empty SLN-2
represent the blank of CRO-SLN and CRT-SLN, respectively. Data represent the mean
of three independent experiments ± SD; *p < 0.05; **p > 0.05. SLN: solid lipid nanoparticle;
CRO: crocin; CRT: crocetin.
Fig. 7 Antiproliferative activity of CRO-SLN and CRT-SLN against a malignant a Schwann cell
line (sNF96.2) versus CRO and CRT solution in PBS (0.1% w/w). Empty SLN-1 and empty
SLN-2 represent the blank of CRO-SLN and CRT-SLN, respectively. Data represent the
mean of three independent experiments ± SD; *p < 0.05; **p > 0.05. SLN: solid lipid
nanoparticle; CRO: crocin; CRT: crocetin.
The nanoencapsulation strategy significantly ameliorated the cytotoxic activity of
CRO for both cell lines (p < 0.05; [Figs. 6] and [7]), while CRT-SLNs did not produce a significant antiproliferative activity with respect
to free CRT (p > 0.05; [Figs. 6] and [7]).
The different trend observed between the two compounds is clearly referable to a higher
chemical stability of CRT compared to CRO, while the similar cytotoxicity of free
CRT and CRT-SLNs cannot exclusively be a matter of free CRT or CRT released from SLNs
in the adjacent milieu, but it also depends upon selective uptake and entry of nanoparticles/bioactives
inside the cells and enhanced effectiveness of bioactives released in cellular proximity.
CRT-SLNs, while ensuring an interesting cytotoxic activity in both the evaluated cell
models, are not able to produce an increase in the intrinsic activity of CRT, probably
because they are not able to be internalized in a satisfactory way. Therefore, we
hypothesize that the observed cytotoxic effect of CRT-SLNs would depend on the release
of CRT in the milieu and on the ability of the SLN system to protect the encapsulated
CRT from the chemical degradation to which the compound is subjected. Besides, free
CRT would have been able to produce in both models a more pronounced cytotoxic effect
if it had greater chemical stability. Further studies have been planned to confirm
the results obtained in the present work.
Materials and Methods
Materials
CRO (batch number: BCBT4979) and HPMC were purchased from Sigma-Aldrich. CRO (purity
≥ 90%) was obtained by CRO hydrolysis following a method reported elsewhere [27]. Softisan 100 (hydrogenated coco-glycerides) was obtained by Sasol Germany. Pluronic
F68 was provided from BASF. Ethanol and water were purchased from Sigma-Aldrich.
Formulation of crocin-solid lipid nanoparticles and crocetin-solid lipid nanoparticles
CRO-SLNs were prepared by using a variation of the QESD method [22], [28]. Briefly, CRO (0.1 g) was solved in 5 mL of water heated to 50 °C. The solution
obtained was kept under constant magnetic stirring. The oil phase constituted by Softisan
100 (0.816 g), ethanol (18 mL), and water (4 mL) was warmed up to 50 °C and slowly
added to CRO aqueous solution. The resulting formulation was dispersed using a high-speed
homogenizer (Ultra-Turrax T25, IKA-Werke GmbH &Co) then added slowly into a hot (50 °C)
surfactant solution (9 mL of water in which 0.7 g of Pluronic F68 were dissolved).
Finally, SLNs were precipitated in an ice bath by quickly adding cold water (200 mL)
to the quasi-emulsion under high-speed homogenization. After a stirring of 30 min,
the suspension was filtered and stored at 4 °C for further characterization [28].
CRT-SLNs were prepared following the solvent diffusion method [21]. Blank and CRT-SLNs were obtained by adding CRO (0.064 g) to Softisan 100 (0.023 g)
previously dissolved in 4.6 mL of ethanol at 50 °C. The aqueous phase constituted
by HPMC (0.23 g), Pluronic F68 (0.23 g), soy lecithin (0.23 g), and water (23 mL)
was heated to 50 °C. Then, the hot lipid phase was added to the surfactant solution
and dispersed at 50 °C for 8 min using a high-speed stirrer (Ultra Turrax T25, IKA-Werke
GmbH & Co.). The obtained pre-emulsion was ultrasonified for 10 min by using a UP
400 S instrument (ultrasonic processor; Dr. Hielscher GmbH) and finally precipitated
in an ice bath by quickly adding cold water (100.6 mL). The obtained SLN dispersion
was filtered and later stored at 4 °C.
Characterization of solid lipid nanoparticle-based formulations
Mean particle size of the lipid dispersions was measured by PCS. A Zetamaster (Malvern
Instrument Ltd.), equipped with a solid-state laser having a nominal power of 4.5 mW
with a maximum output of 5 mW 670 nm, was employed. Analyses were performed using
a 90° scattering angle at 20 ± 0.2 °C. Samples were prepared by diluting 100 µL of
SLN suspension, previously filtered through a 0.45-µm Acrodisc LC 13 PVDF filter (Pall-Gelman
Laboratory), with 900 µL of deionized water. During the experiment, the refractive
index of the samples always matched the liquid (toluene) to avoid stray light. The
zeta (ξ) potential was automatically calculated from electrophoretic mobility based on the
Smoluchowski equation (Equation 1):
where v is the measured electrophoretic velocity, η is the viscosity, ε is the electrical permittivity of the electrolytic solution, and E is the electric field. The accuracy was 0.12 µm cm/V s for aqueous systems. Samples
were suspended in distilled water and the measurements were recorded at 25 °C.
Determination of drug loading
The percentage of CRO and CRT entrapped in the lipid matrix was determined as follows:
an amount of CRO-SLN and CRT-SLN dispersions were filtered using a Pellicon XL tangential
ultrafiltration system (Millipore) equipped with a polyethersulfone Biomax 1000 membrane
(Millipore) with a 1 000 000 Da molecular weight cutoff. An amount of retained material
was freeze-dried, dissolved in dichloromethane, and analyzed by UV spectrophotometry
at 440 nm (Lambda 52, PerkinElmer) [29]. Calibration curves for the validated UV assays of CRO and CRT were performed on
six solutions in the concentration range of 10 – 100 mg/mL. The correlation coefficient
was 0.99. The incorporation efficiency was determined as DR, that is (Equation 2):
Differential scanning calorimetry
Calorimetric analysis was affected by using a Mettler Toledo DSC 1 STARe system equipped
with a PolyScience temperature controller. The detection system was an HSS8 high-sensitivity
sensor (120 gold-gold/palladium-palladium thermocouples) and a ceramic sensor (Mettler
Full Range; FRS5) with 56 thermocouples. The signal time constant was 18 s and the
digital resolution of the measurement signal was less than 0.04 µW. Resolution and
sensitivity calorimetric determined by TAWN test was 0.12 and 11.9, respectively.
The sampling rate was 50 values/second. The sensitivity was automatically chosen as
the maximum possible by the calorimetric system, and the reference was an empty pan.
The calorimetric system was calibrated, in temperature and enthalpy changes, by using
indium and following the procedure of the DSC 1 Mettler TA STARe instrument. Empty
and CRO- or CRT-loaded SLNs were sealed in an aluminum pan and submitted to DSC analysis
to determinate the influence of the drug on the thermotropic parameters of the SLNs.
Each sample was submitted to heating and cooling cycles in the temperature range of
10 – 120 °C at a scanning rate of 5 °C/min (heating) and a scanning rate of 10 °C/min
(cooling). Only free drugs were submitted to heating and cooling cycles in the temperature
range of 10 – 200 °C at the same scanning rate. Transition temperature and enthalpy
changes (ΔH) were calculated from peak areas with the Mettler STARe Evaluation software system
(version 13.00) installed on an Optiplex 3020 DELL.
Turbiscan analysis
Turbiscan AGS (Formulaction, lʼUnion, France, a robot and storage station integrated
TurbiscanLAB) was used to examine the dispersion stability of the SLNs. This instrument
allows for the use of a simple technique to observe reversible and irreversible destabilization
phenomena in the sample without the need of dilution. Turbiscan AGS is useful to detect
destabilization phenomena much earlier and also in a simpler way than other methods.
The suspension (10 mL) was placed in a flat-bottomed cylindrical glass tube and was
placed in the instrument. The transmission of light from the suspensions was then
measured periodically (1 h) along the height at 25 and 37 °C. The experiments were
performed until 48 h. The sedimentation behavior of the suspensions was monitored
by measuring the backscattering and transmission of monochromatic near infrared (λ = 850/880 nm).
In vitro release study
The amount of CRO and CRT released from SLNs was determined as reported elsewhere
[1], [30]. Briefly, a weighted amount of CRT-SLN or CRO-SLN suspension was dispersed in a
phosphate buffer solution and then divided into different aliquots in glass vials.
The vials were placed in a water bath that maintained a constant temperature of 37 °C.
At predetermined intervals, the batch was centrifuged at 3000 rpm for 10 min to separate
the released drug from the SLN dispersion. After centrifugation, the sample was recovered
and analyzed by a UV spectrophotometer at 440 nm to determine the CRT or CRO content.
Oxygen radical absorbance assay
An ORAC assay was used to qualitatively determine the antioxidant activity of CRO-SLNs
and CRT-SLNs [8]. A fresh solution of CRO and CRT alone (25 µL) plus empty and loaded vehicles (25 µL)
were placed in 96-well tissue culture plates. One hundred and fifty µL of FL (10 nM)
was used as the probe to assess the antioxidant activity. Twenty-five µL of water-soluble
azo-compound AAPH (100 mM) was used as a radical initiator to generate free radicals
at a constant rate. A positive (FL solution containing AAPH) and negative control
(FL solution containing no AAPH) were run simultaneously in PBS. A timer was started
upon introduction of the free radical generator and the plate was stored in the dark
at 37 °C. At each specified time point, the fluorescence of the solution was measured
(excitation 492 nm, emission 535 nm) using a Wallac 1420 Victor 3 96-well plate reader
fluorimeter (Perkin Elmer) and plotted as a function of time with Origin 7 software
(Origin Lab Corporation).
In vitro cell studies
Cell Culture
The human melanoma A375 and malignant Schwann sNF96.2 cell lines were provided by
American Type Culture Collection and routinely maintained in DMEM (Sigma-Aldrich)
supplemented with 10% FBS (Invitrogen), 4 mM/L glutamine, 4500 mg/L glucose, 1500 mg/L
sodium bicarbonate, 100 U/mL penicillin, and 100 µg/mL streptomycin. Cells were incubated
at 37 °C in a humidified atmosphere of 95% air/5% CO2. A subcultivation ratio of 1 : 3 to 1 : 4 was performed twice weekly by using trypsin
(0.25%)-EDTA (1 mm) for the detachment. The medium was changed every 2 – 3 days.
Cell treatments
A day before the experiments, cells were trypsinized, counted, and seeded at an initial
density of 8 × 103 cells per microwell in flat-bottomed 96-well culture microplates (Thermo Scientific
Nunc) and incubated at 37 °C in a humidified atmosphere containing CO2 (5%) with treatment medium (without FBS) for 24 h. After that, the medium was removed
and replaced with the treatment medium containing CRO-SLNs, CRT-SLNs, a CRO solution,
a CRT solution, and a sample of empty SLNs. After 72 h incubation, the cells proliferation
was evaluated.
Cell viability evaluation
Cell proliferation was tested by the MTT assay based on the conversion by mitochondrial
dehydrogenases of a substrate containing a tetrazolium ring into blue formazan, detectable
spectrophotometrically. At different time points, the treatment media were removed
and a solution of MTT (0.5 mg/mL) was added to each microwell. After 3 h of incubation
at 37 °C, the supernatant was removed and replaced with DMSO (100 µL). The optical
density of each well sample was measured with a microplate spectrophotometer reader
(Digital and Analog Systems) at λ = 550 nm. Each sample was tested in quadruplicate (n = 12) and cell viability was
reported as percentage (%) compared to the controls (untreated, empty SLNs).
Statistical analysis
Statistical analysis of the data was performed using Studentʼs t-test. A probability
(p) of less than 0.05 was considered significant in this study. All statistical analyses
were performed using the statistical software package SYSTAT, version 13 (Systat Inc.).
