1. Introduction
The rapid and uncontrolled proliferation of tumor cells to adjacent tissues and organs, as well as the lack of targeted therapies with reduced systemic toxicity, make cancer one of the diseases with the highest mortality rate [
1]. Several limitations associated with conventional treatments, specifically in the pharmacokinetic profile of the drug (e.g., reduced specificity, systemic toxicity and early metabolization and elimination) [
2], point to an urgent need to develop new therapeutic approaches that allow the safe use of anticancer drugs.
In recent years, many studies have been focused on the development of magnetoliposomes to overcome the limitations of conventional liposomes [
3,
4,
5]. These drug delivery systems, which consist of liposomes based on magnetic nanoparticles (MNPs), improve the therapeutic efficacy by the synergy between magnetic targeting and simultaneous hyperthermia and controlled drug delivery. In the presence of an external magnetic field gradient, the nanoparticles are responsible for guiding the nanosystem to the tumor site, where the content will be released on demand by using an alternating magnetic field (AMF) and/or a NIR laser light source. Moreover, this type of nanoparticle is able to produce a dual hyperthermia effect (magnetic and/or photothermal) [
6,
7,
8,
9].
Nanoparticles with superparamagnetic behavior, i.e., without coercivity, hysteresis or remanence, are ideal for biomedical applications. By adjusting their elemental composition and shape anisotropy, the magnetic properties of the MNPs can be optimized. Magnetite (Fe
3O
4) and maghemite (γ-Fe
2O
3) have been the most explored compositions for cancer therapy applications [
10,
11,
12,
13,
14]. Recently, a high interest in the study of ferrites composed of alkaline earth metals has emerged. In particular, calcium and magnesium ferrites show high cell viability, as these elements are easily metabolized. Therefore, greater biocompatibility and biodegradability are guaranteed, as reported by several research groups [
15,
16,
17,
18,
19]. Magnesium ferrite MNPs have been shown to have improved magnetic properties, acting as efficient hyperthermia agents [
20]. Hirazawa et al. [
21] reported that the partial replacement of Mg
2+ ions by Ca
2+ in magnesium ferrite nanoparticles (Mg
1-xCa
xFe
2O
4) favors biocompatibility, magnetization and thermal energy dissipation under an external magnetic field. The heating efficiency of nanoparticles with shape anisotropy, i.e., non-spherical, has shown clear advantages [
22,
23,
24,
25]. For instance, the easier magnetization along their longest axis gives them a longer blood circulation time and better magnetic and hyperthermia properties [
26]. Unlike isotropic MNPs, their magnetic behavior results from the interaction between shape anisotropy and magnetocrystalline anisotropy [
27,
28].
Liposomes are advantageous nanosystems for drug delivery applications. As the non-covalent forces associated with their self-organization are reversible, this allows a dynamic transition between several morphologies in response to different stimuli [
29]. At the tumor level, there are specific stimuli, which can be strategically used as triggers for controlled drug delivery [
30]. For instance, variations in the pH value can stimulate the release of encapsulated therapeutic agents into the cytoplasmic space of abnormal cells through the endocytic pathway, by lysosomes and endosomes, using pH-sensitive liposomes [
31]. As the bloodstream and healthy tissues have a neutral pH (approximately pH = 7.4), upon reaching the tumor microenvironment, which is more acidic (pH around 5), these types of formulations undergo a membrane destabilization, which helps the fusion of liposomes and drug release at acidic pH. At the endosomal level, these liposomes lose their initial stability due to their increased fusogenic potential, which prevents lysosomal degradation by enzymatic action, releasing the therapeutic compound into the cytosol [
30].
Typically, pH-sensitive liposomes are composed of phospholipids of phosphatidylethanolamine (PE) and derivatives containing carboxylic groups that promote the stabilization of liposomes at neutral pH [
32]. These types of natural and unsaturated phospholipids have fusogenic properties, as they have a good ability to adhere to cell membranes. This adhesion is ensured by the weak hydration of the small polar heads [
32]. Several studies refer to the use of dioleoylphosphatidylethanolamine (DOPE) for the formation of pH-sensitive lipid vesicles. This phospholipid has an inverted conical shape and packing parameter above 1, naturally forming inverted structures [
33,
34,
35]. The phospholipid dipalmitoylphosphatidylethanolamine (DPPE) has a conical geometry similar to DOPE; however, it is saturated, presenting a more rigid structure [
36]. Liposomes containing only DOPE or DPPE have a reduced stability. To overcome this problem, it is usual to add an amphiphilic molecule, which stabilizes the lipid bilayer, e.g., cholesterol and derivatives such as cholesteryl hemisuccinate (CHEMS) [
37]. In liposomes containing phospholipids with the PE group and CHEMS, the factor that dictates the behavior and conformation of the vesicle is the ionization state of CHEMS. At neutral pH, CHEMS is ionized (negatively charged), so the liposome presents a linear conformation as a result of electrostatic repulsions between the phosphate groups of DOPE/DPPE and the carboxylic group of CHEMS. At acidic pH, the CHEMS carboxylic group is protonated, which causes a change in the conformation of DOPE/DPPE, resulting in an inverted hexagonal phase and, consequently, the destabilization of vesicle membranes and the release of antineoplastic agents [
30]. As it acts to reduce the transition temperature between the lamellar and hexagonal phases and decreases the permeability of biomembranes, preventing the early release of encapsulated drugs, cholesterol (Ch) was also included in the final lipid composition [
30,
37].
The present work focuses on the development of pH-sensitive nanosystems with improved magnetic properties derived from anisotropic-shaped MNPs. The structural, magnetic and hyperthermia characterization confirmed the synthesis of anisotropic mixed ferrite nanoparticles (Mg0.75Ca0.25Fe2O4) with high magnetization and thermal energy dissipation (under AMF or NIR irradiation). These were effectively surrounded by a pH-sensitive lipid bilayer of DOPE:Ch:CHEMS (45:45:10), forming solid magnetoliposomes. Overall, promising results were obtained for a controlled delivery of DOX under pH trigger, which was corroborated by cell viability assays performed on the HepG2 adherent human liver hepatocellular carcinoma cell line. To our knowledge, the combination of shape anisotropy magnesium/calcium ferrite nanoparticles and pH-sensitive liposomes of this nanosystem was not previously reported, showing promising results for application in combined therapies.
2. Materials and Methods
All the solutions and synthesis procedures were prepared using ultrapure water of Milli-Q grade (MilliporeSigma, St. Louis, MO, USA) and spectroscopic grade solvents.
2.1. Preparation of Anisotropic Nanoparticles of Mg0.75Ca0.25Fe2O4
Shape anisotropic magnetic nanoparticles of magnesium ferrite with 25% replacement by calcium ions (Mg
0.75Ca
0.25Fe
2O
4) were prepared using an adapted protocol previously described by Cardoso et al. [
23]. To synthesize MNPs with anisotropic shape, several surfactants can be used as shape inducing agents [
38]. In this protocol, a binary surfactant system was applied using octadecene and oleic acid as the solvent and shaping agent, respectively. The last one acts as a stabilizer and promotes the formation of specific arrangements on the surface in the nucleation and crystal growth phases of non-spherical nanoparticles [
39].
For this purpose, a solution of 2 mM of iron (III) citrate tribasic monohydrate in 15 mL of octadecene was heated to 120 °C, under continuous magnetic stirring, until dissolution of the metallic precursor. Then, 0.25 mM of calcium acetate hydrate, 0.75 mM of magnesium acetate tetrahydrate and 3.1 mM of oleic acid were added to the previous solution. After 1 h, under magnetic stirring at 120 °C, a condenser was linked to the system to ensure a flow capable of keeping the molecules in the liquid state at controlled temperature. The mixture was heated at a rate of 1 °C per min, until reaching 200 °C. Thenceforth, a new increase in temperature was imposed at 5 °C per min until reaching the boiling point of octadecene (290 °C), remaining at this temperature for 1 h. The resulting MNPs were suspended in tetrahydrofuran (THF) and washed with water and ethanol in several cycles of magnetic decantation. Finally, half of the nanoparticles were calcined at 350 °C under an ultrapure nitrogen flow for 30 min to remove surface organic residues of octadecene and oleic acid. Both calcined and non-calcined MNPs were characterized to compare their properties. The application of this protocol is expected to yield cubic-shaped magnetic nanoparticles.
2.2. Preparation of Liposomes
In this work, pH-sensitive liposomes were prepared. Different lipid formulations were studied using the lipids dioleoylphosphatidylethanolamine (DOPE, from Avanti Polar Lipids, Birmingham, AL, USA), dipalmitoylphosphatidylethanolamine (DPPE, from Sigma-Aldrich, St. Louis, MO, USA), cholesterol (Ch, from Sigma-Aldrich, St. Louis, MO, USA) and cholesteryl hemisuccinate (CHEMS, from Sigma-Aldrich, St. Louis, MO, USA) at the molar ratios DOPE:CHEMS (7:3), DOPE:Ch:CHEMS (45:45:10) and DPPE:Ch:CHEMS (45:45:10). Liposomes composed of these formulations were synthetized following the ethanolic injection method [
40], where an ethanolic lipid solution (1 × 10
−3 M) was initially prepared according to the desired proportion. The volume of the lipid solution equivalent to this concentration was evaporated under an ultrapure nitrogen flow and subsequently redissolved in absolute ethanol in the same volume. Finally, the ethanolic solution was injected, drop by drop and under vortexing, to 3 mL of ultrapure water, inducing the formation of liposomes. The injection was performed while ensuring that the aqueous medium was at a higher temperature than the main phospholipid phase transition temperature (−16 °C for DOPE and 63 °C for DPPE) [
41].
2.3. Preparation of Solid Magnetoliposomes
Solid magnetoliposomes loaded with doxorubicin were prepared following the method described in Ref [
23]. For this purpose, the lipids DOPE, Ch and CHEMS were used at the molar ratio of 45:45:10, respectively. According to Ref [
23], a thin lipid film with 2/3 of 1 mM of DOPE:Ch:CHEMS was prepared by solvent evaporation under a nitrogen flow. To this film, 3 mL of heptane was added, and this solution was ultrasonicated at 190 W for 10 min. In this step, reverse micelles with uniform sizes are formed. Afterward, 1 mM of dried MNPs was added, and the solution was again ultrasonicated for another 5 min to force its entry into the previously formed micelles. To promote the nanosystem precipitation, this solution was placed in the cold. Then, by magnetic decantation, reverse micelles containing the MNPs were purified, and the resultant pellet was completely dried through an ultrapure nitrogen flow. Finally, the pellet was resuspended in 3 mL of ultrapure water, and a solution containing the remaining lipid and DOX (1.13 × 10
−4 M) was added by ethanolic injection, under vortex, forming the second lipid layer and consequently, the solid magnetoliposomes.
2.4. Spectroscopic Measurements
Spectroscopic measurements were carried out at Photophysics Laboratory of the Physics Center of the University of Minho, Braga, Portugal. Absorption spectra of MNPs dispersions were acquired on a double-beam Shimadzu UV-Vis-NIR spectrophotometer, model UV-3600 Plus (Shimadzu Corporation, Kyoto, Japan). The fluorescence emission spectra of samples containing the SMLs were measured on a Fluorolog 3 spectrofluorometer (HORIBA Jobin Yvon IBH Ltd., Glasgow, UK) equipped with double monochromators in excitation and emission, exciting at 480 nm (DOX excitation).
2.5. Structural Characterization
The crystalline structure and phase identification of the shape anisotropic magnetic nanoparticles were determined by X-ray diffraction (XRD), using a PAN’alytical X’Pert PRO diffractometer (Malvern Panalytical Ltd., Malvern, UK) in a Bragg–Brentano configuration operating with Cu Kα radiation (λ = 0.154060 nm), at the Electron Microscopy Unit of the University of Trás-os-Montes and Alto Douro (UTAD), Vila Real, Portugal.
All images of Mg0.75Ca0.25Fe2O4 MNPs and SMLs were obtained by transmission electron microscopy (TEM), using a JEOL JEM-1010 high-contrast microscope operating at 100 kV (Centro de Apoio Científico-Tecnolóxico à Investigación (CACTI), Vigo, Spain). The samples were ultrasonicated and deposited on copper grids with carbon and Formvar. The TEM images processing was performed using ImageJ software (version 1.53t, National Institutes of Health (NIH), Bethesda, MD, USA). The manual selection of the nanoparticles’ diameter allowed estimating their average size. The histograms obtained by this analysis were fitted to a Gaussian distribution.
The average hydrodynamic diameter and zeta potential (ζ) of the developed liposomal formulations and SMLs were measured on a dynamic light scattering (DLS) equipment LitesizerTM 500 from Anton-Paar (Anton-Paar GmbH, Graz, Austria) equipped with a laser diode of λ = 658 nm. For liposomal measurements, aqueous solutions of 0.3 mM were prepared in phosphate buffer pH = 7.4 and pH = 5. In the case of SMLs, aqueous solutions of 0.5 mM were measured in aqueous buffer solutions. Before measurement, all samples were sonicated and filtered using a Socorex DosysTM all-glass syringe attached to a hydrophilic polytetrafluoroethylene (PTFE) filter. The analysis was carried out at 25 °C using an Univette cell with an optical path of 10 mm. Three independent measurements were performed for each sample. For SMLs colloidal stability assays, aqueous solutions in PBS pH = 7.4 were prepared and stored at 4 °C for 7 days. Differences in the hydrodynamic diameter, PDI and zeta potential were recorded by DLS and ELS, as described above.
2.6. Magnetic Characterization
The magnetic properties of the mixed ferrite nanoparticles were evaluated on a MPMS3 Superconducting Quantum Interference Device (SQUID) magnetometer Quantum Design MPMS5XL (Quantum Design Inc., San Diego, CA, USA) at IFIMUP (University of Porto, Porto, Portugal). The hysteresis loop was obtained by measuring the magnetization of the samples as a function of the applied magnetic field (H), applying fields up to 5.5 T, at room temperature (300 K). For that, the temperature was fixed, and the magnetic moment (M) was measured at different values of H.
2.7. Hyperthermia Measurements
The heat capability generation of the nanoparticles was quantified through the specific absorption rate (SAR), described by Equation (1):
where
is the specific heat capacity of the medium (4.186 J g
−1 K
−1);
/
is the initial slope of the temperature curve as a function of time; and
and
are the mass of the solvent and of the magnetic material, respectively [
42]. As SAR depends on the frequency and intensity of the applied AMF, it does not allow comparing values with confidence. Thus, the intrinsic loss power (ILP, nH·m
2/kg) is the parameter designed for this purpose, according to Equation (2):
where
corresponds to the intensity of the magnetic field (kA/m), and
is the frequency (kHz) [
43].
2.7.1. Magnetic Hyperthermia
The heating capabilities of Mg0.75Ca0.25Fe2O4 nanoparticles, in the presence of an alternating magnetic field, were evaluated using a hyperthermia setup at the Faculty of Engineering of University of Porto (Porto, Portugal) working at f = 155 kHz and H = 8.5 kA/m. Temperature variations resulting from the application of an AMF were recorded for 30 min. In the absence of magnetic field, the cooling of the solutions was also recorded for half an hour. For the magnetic hyperthermia measurement, the SAR value was calculated following Equation (1) and using the initial linear slope method. This indicates the rate at which electromagnetic energy is absorbed per unit mass of MNPs.
2.7.2. Photothermal Hyperthermia
The potential of anisotropic MNPs in photothermal therapy (PTT), i.e., their ability to dissipate thermal energy under NIR radiation, was evaluated using an experimental irradiation setup at the Physics Centre of University of Minho. The developed setup consists of a sample holder, a laser light source with a wavelength of 808 nm and 1 W/cm2 of power density, and a T-type thermocouple connected to a digital multimeter (Agilent U1242A) for temperature measurement. Before each measurement, the temperature was stabilized at room temperature, and then, the samples were irradiated. Temperature variations were recorded over time for 30 min, and afterward, the laser was turned off, and cooling was recorded in the same way. The SAR calculation was performed as described in the previous section.
2.8. Drug Encapsulation Efficiency
Doxorubicin encapsulation efficiency, EE(%), in magnetoliposomes was estimated by fluorescence spectroscopy. Initially, the fluorescence emission spectra of aqueous solutions of DOX with different concentrations were measured. Using the maximum fluorescence intensity of each spectrum, obtained with excitation at 480 nm, a calibration curve of fluorescence intensity was plotted as a function of DOX concentration, and then, a linear regression was fitted. To determine the value of EE(%), 1.5 mL of SMLs solution was placed on the top of Amicon
® Ultra centrifugal filter units (100 kDa), which were subjected to centrifugations, at 3000 rpm for 10 min. As a result, two phases were obtained: the solution retained above the membrane of the filters, corresponding to doxorubicin, which was effectively encapsulated in the SMLs, and the aqueous solution below the membrane, containing the non-encapsulated drug. Subsequently, and based on the calibration curve initially obtained, the concentration of non-encapsulated drug was quantified. Three independent measurements were performed, and the respective standard deviation (SD) was calculated. The last step consisted of calculating the EE(%) of DOX in SMLs, as described by Equation (3):
2.9. Drug Release Kinetics
The DOX release profile of the developed nanosystems was evaluated using two different media, one to simulate the acidic tumor microenvironment (pH = 5) and another for physiological conditions (pH = 7.4). For this, solid magnetoliposomes were diluted in phosphate buffer with the desired pH value, and these solutions were placed above the Amicon filter membrane. The volume below the cellulose membrane was filled with 7.5 mL of the corresponding buffer solution, and the Amicon filters were placed on an orbital shaker at 300 rpm. Aliquots of 200 µL were pipetted out from the solution placed below the membrane, always being replaced with 200 µL of PBS at the corresponding pH. This procedure was repeated for 48 h. For each aliquot, the fluorescence spectrum was measured, with excitation at 480 nm, and the concentration of the released doxorubicin was quantified using the calibration curves obtained in PBS 5 and 7.4, as described in
Section 2.8. Three independent measurements were performed. To better understand the kinetics, the cumulative DOX release curves were fitted to the Weibull and first-order models using Prism 8 software (GraphPad Software, La Jolla, CA, USA).
2.10. Cell Culture
The adherent human liver hepatocellular carcinoma HepG2 cell line (ATCC HB-8065—American type culture collection, Virginia, VA, USA) was cultured and maintained in T75 culture flasks with Roswell Park Memorial Institute (RPMI) 1640 medium, with GlutaMAX™ supplemented with 10% fetal bovine serum (FBS), 1% Penicillin-Streptomycin in a humidified incubator, at 37 °C, with 5% CO2 environment. When confluence was achieved, the sub-culturing process was performed by trypsinization, and the cell count was performed using Trypan blue and a Neubauer chamber.
2.11. Biological Studies
For viability assays, the AquaBluer assay (MultiTarget Pharmaceuticals, LLC) was used. HepG2 cells were seeded on 96-well plates, at a density of 7000 cells/well, and were incubated with supplemented RPMI-1640 medium overnight, at 37 °C, in the CO2 incubator. Afterward, the medium was aspirated, and the viability assay was conducted by adding 100 μL of AquaBluer solution (1:100 in supplemented RPMI-1640 medium) to each well and incubation for 2 h, at 37 °C, under 5% CO2 environment. Then, a microplate reader at λexc = 540 nm and λem = 590 nm wavelengths was used to evaluate the viability. This process was repeated after 48 h of adding DOX-loaded SMLs at two different DOX concentrations (5.65 × 10−5 M and 1.13 × 10−4 M), as well as free DOX with the respective concentrations. Untreated HepG2 cells were used as a control. For each condition, three independent measurements were performed.
4. Conclusions
In this work, shape anisotropic Mg0.75Ca0.25Fe2O4 magnetic nanoparticles were prepared and characterized in detail. The magnetic properties allowed characterizing the nanoparticles as superparamagnetic, an ideal behavior for application as magnetic hyperthermia agents. Magnetic hyperthermia (SAR = 4.98 W/g) and photothermal (SAR = 1.1 × 104 W/g) measurements confirmed the enhanced potential of nanorods in cancer treatment. The calcined nanoparticles, with rod-like shape, were selected as the magnetic component of pH-sensitive DOPE:Ch:CHEMS (45:45:10) SMLs. After the effective incorporation of the drug doxorubicin in the SMLs, with a high encapsulation efficiency, its release profile was studied, which was much more pronounced in acidic medium (55%). Biological assays on the HepG2 cell line confirmed the drug delivery capability of SMLs, reducing cancer cells’ viability to 26.8% using a low doxorubicin concentration, and the development of a safe and non-toxic nanosystem was confirmed.
Considering all the results, the combination of non-spherical magnetic nanoparticles with a pH-sensitive liposomal formulation emerges as a promising approach to exert a multimodal cancer therapy through pH-sensitive drug delivery combined with dual hyperthermia (magnetic and photothermia). To the best of our knowledge, the pH-sensitive DOPE:Ch:CHEMS (45:45:10) formulation is innovative in the composition of solid magnetoliposomes.