Intranasal administration offers an alternative and promising approach for direct nose-to-brain delivery. Herein, we developed two chitosan (CHT)-coated (and uncoated) nanoformulations of BNN27 (a synthetic C-17-spiro-dehydroepiandrosterone analogue), liposomes (LIPs), and nanoemulsions (NEs), and compared their properties and brain disposition (in vitro and in vivo). LIPs were formulated by thin film hydration and coated with CHT by dropwise addition. BNN27-loaded NEs (BNEs) were developed by spontaneous emulsification and optimized for stability and mucoadhesive properties. Mucoadhesive properties were evaluated by mucin adherence. Negatively charged CHT-coated LIPs (with 0.1% CHT/lipid) demonstrated the highest coating efficiency and mucoadhesion. BNEs containing 10% w/w Capmul-MCM and 0.3% w/w CHT demonstrated the optimal properties. Transport of LIP or NE-associated rhodamine-lipid across the blood–brain barrier (in vitro) was significantly higher for NEs compared to LIPs, and the CHT coating demonstrated a negative effect on transport. However, the CHT-coated BNEs demonstrated higher and faster in vivo brain disposition following intranasal administration compared to CHT-LIPs. For both BNEs and LIPs, CHT-coating resulted in the increased (in vivo) brain disposition of BNN27. Current results prove that CHT-coated NEs consisting of compatible nasal administration ingredients succeeded in to delivering more BNN27 to the brain (and faster) compared to the CHT-coated LIPs.
Keywords: BNN27; intranasal delivery; mucoadhesive formulations; chitosan; LIPs; nanoemulsions
Dehydroepiandrosterone (DHEA) is one of the most abundant neuroactive steroids. It is synthesized in the human adrenal cortex as well as in the brain by neurons and glia [[
A new family of small synthetic C17-spiro DHEA derivatives named microneurotrophins have been synthesized [[
In most cases, treatments for CNS disorders are administered parenterally, reducing drug effectiveness and potency. Even if the lipophilicity of the drug does not impede its accessibility to the brain through the circulation, systemic clearance significantly reduces the drug bioavailability [[
Herein, we prepared two different formulation types, LIPs and nanoemulsions with the optimal physicochemical and mucoadhesive properties, and compared them for their potential to deliver BNN27 to the brain following nasal administration. To the best of our knowledge, this is the first time that such mucoadhesive formulations have been compared for their brain disposition after intranasal delivery.
1,2-Distearoyl-sn-glycerol-3-phosphatidylcholine (PC) and 1,2-distearoyl-sn-glycero-3-phospho-(19-rac-glycerol) (sodium salt) (PG), and Lissamine Rhodamine B phosphatidylethanolamine or Rhodamine-lipid (RHO), were purchased from Avanti Polar Lipids (Alabaster, AL). Capmul MCM was received as a gift sample from Abitec Corporation Limited (Columbus, OH, USA). Labrafac Lipophile WL 1349, Labrafac PG, and Transcutol HP were received as gift samples from Gattefosse (Lyon, France). Tween 80 was purchased from Fisher BioReagents and Tween 20 from BioChemica UK Ltd. Carbopol 974P was kindly provided by Chemix SA (Athens, Greece). Cholesterol, mucin from porcine stomach Type III bound sialic acid 0.5–1.5% (partially purified powder), low molecular weight chitosan (LMW-CHT, with a molecular weight of 50–190 kDa and 75–85% deacetylated) and medium molecular weight chitosan (MMW-CHT, with a molecular weight between 190–310 kDa and 75–85% deacetylated), and all other excipients for the nanoemulsion preformulation studies were purchased from Sigma-Aldrich or Merck. BNN27 was kindly provided by Bionature Ltd.
For the quantification of BNN27, an enzymatic method—a cholesterol kit purchased from Biotechnological applications LTD (Athens, Greece)—was used. All other chemicals were of analytical quality and were purchased from Sigma-Aldrich or Merck (Darmstadt, Germany).
For the measurement of BNN27 loading in the various formulations during formulation development, as a routine everyday quantification method, the CO/PAP enzymatic method [[
The measurement of BNN27 in ΝΕs or solutions of NE ingredients is not possible with this method due to interactions between the NE ingredients and reagent. Thereby, the BNN27 extraction was applied for the measurement of BNN27 in NEs or in NE ingredients (oils, surfactants and co-surfactants). For this, 1 mL of BNN27-containing NEs or solutions was vigorously mixed (by vortex) with 2 mL chloroform for (at least) 2 min. Then, after complete separation of the two phases, the aqueous phase was removed and the organic was evaporated. The drug was re-suspended in 200 μL of ethanol and mixed with 1 mL of reagent. After incubation at 37 °C for 15 min, the sample OD-510 nm was measured.
The BNN27 content was calculated from a calibration curve that was conducted by the same method (applied in each case) using known amounts of BNN27 mixed with empty LIPs, blank NEs, or NE components.
MLV LIPs composed of PC or PC/PG at 9:1 (mol/mol) and loaded with BNN27 were prepared by the thin-film hydration method [[
Separation of LIPs from the non-encapsulated BNN27 was achieved by centrifugation (
For the LIP coating, two types of CHT were used: low molecular weight (LMW) and medium molecular weight (MMW). CHT solutions of different concentrations were prepared in isotonic acetate buffer (pH = 4.40). Then, 0.5 mL of the LIP dispersion (in PBS) was mixed with the dropwise addition of an equal volume of the appropriate CHT solution to give CHT/lipid (w/w) ratios of 0.1, 0.3, or 0.5, under continuous stirring for 1 h at room temperature [[
For coating efficiency determination, the phospholipid content of the samples was measured by the Stewart assay and compared to the total phospholipid (before centrifugation), as previously reported [[
Coating Efficiency (CE) % = LPRE (lipid-in-precipitate)/LTOT (Total Lipid) × 100 (
The size distribution (mean hydrodynamic diameter and polydispersity index) and ζ-potential of the LIP dispersions were measured by dynamic light scattering (DLS) and laser Doppler electrophoresis (LDE), respectively, on a Nano-ZS (Nanoseries, Malvern Instruments), which measures the mass distribution of the particle size as well as the electrophoretic mobility of the dispersed particles. Measurements were made at 25 °C with a fixed angle of 173°. Sizes quoted are the z-average mean (dz) for the liposomal hydrodynamic diameter (nm). Calculation of ζ-potential (mV) was carried out by the instrument from electrophoretic mobility, which was measured in small volume disposable zeta cells and converted to zeta potential by in-built software that applies the Helmholtz–Smoluchowski equation. For measurements, samples were diluted to have a 0.4 mg/mL lipid concentration.
In order to formulate BNN27-loaded nanoemulsions (BNEs), we followed the previously reported methodologies for the optimal formulations of nanoemulsions intended for nose-to-brain delivery of various drugs such as risperidone [[
In order to select the optimal materials, BNN27 solubility studies were carried out. The solubility of BNN27 was measured in three different oils that are commonly used in NEs [[
After the selection of the best oil and co-surfactant (based of BNN27 solubility), and due to the fact that two of the tested surfactants, Tween 20 and Tween 80, demonstrated similar (very high) ability to solubilize BNN27, both surfactants were studied with the scope to finally select the one that conferred NEs with the maximum stability. Since it is known from the relevant literature that the best surfactant/co-surfactant ratio in the case when Tween 80 is used (as surfactant) and Transcutol HP/propylene glycol (1:1) (as co-surfactant) is 4:2 (w/w) [[
Finally, the globule sizes of the NEs constructed using Capmul MCM as the oil phase at 8% or 10% (w/w), and Tween 20 or Tween 80 as the surfactant (always mixed with Transcutol HP/propylene glycol (1:1) (as the co-surfactant) with 44% (w/w) Smix content), were compared in order to identify which of the two surfactants, Tween 20 or Tween 80, produced NEs with the lowest globule size and highest stability.
After the identification of the optimal NE composition, BNN27-loaded NEs were formulated by the spontaneous emulsification (titration) method. For this, a saturated solution of BNN27 in Capmul MCM was prepared by adding 40 mg/mL BNN27 in the oil phase and applying magnetic stirring. Then, in the BNN27 Capmul MCM phase, the Smix (containing Tween 80 as surfactant) was added until a clear mixture was produced. Finally, H
Mucoadhesive, chitosan (CHT) or carbopol (CAR) coated BNN27-loaded NEs (BNEs) were also prepared. For this, concentrated BNEs (using the minimum volume of the external phase) were initially prepared, and then mixed with the required volume of CHT or CAR aqueous solution to attain a final CHT or CAR concentration of 0.3% w/w. After the addition of CHT or CAR, the BNEs were allowed to homogenize by continuous stirring for 1 h.
The quality and stability of the various NE or BNE formulations constructed were evaluated by dilution tests, centrifugation tests, measurements of pH, transmittance, globule size distribution, and ζ-potential. The dilution test was performed by diluting 1 mL of NEs to 100 mL with d.d. H
The stability of the NEs (coated and non-coated) was evaluated by applying all of the methods above-mentioned at various time points (
The adsorption of mucin on the surface of LIPs or NEs was used as a method to assess the mucoadhesive properties of the BNN27-loaded LIPs and NEs [[
For LIPs, 1 mL of mucin aqueous solution (0.5 mg/mL) was mixed (vortexed) with an equal volume of each LIP dispersion (lipid concentration at 2 mg/mL) at room temperature and the dispersions were centrifuged at 15,000 rpm for 30 min. Free mucin was measured in the supernatant. PC and PC/PG (negatively charged) LIPs were studied, before and after coating. The same protocol was used for the measurement of the mucoadhesive properties of NEs, after the NEs were diluted in order to eliminate any turbidity that would affect the measurements.
For measurement of the free mucin in the supernatants, the Bradford colorimetric method was used [[
Immortalized human brain microvascular endothelial cells (hCMEC/D3) as well as human embryonic kidney cells (HEK) were used. HEK cells were grown in a high glucose DMEM medium supplemented with 10% FBS and 1% antibiotic-antimycotic solution (Invitrogen, Carlsbad, CA, USA). The cells were cultured at 37 °C, 5% CO
hCMEC/D3 cells (passage 25–35) were obtained under license from the Institut National de la Sante et de la RechercheMedicale, INSERM, Paris, France and grown in EndoGRO medium (Merck, Darmstadt, DE) supplemented with 10 mM HEPES, 1 ng/mL basic FGF (bFGF), 1.4 μM hydrocortisone, 5 μg/mL ascorbic acid, penicillin-streptomycin, chemically defined lipid concentrate, and 5% ultralow IgG FBS. All cultureware was coated with 0.1 mg/mL rat tail collagen type I (BD Biosciences, Franklin Lakes, NJ, USA).
The cytotoxicity of the liposomal and nanoemulsion samples toward the hCMEC/D3 and HEK cells was evaluated with the MTT assay. Briefly, 25,000 cells were seeded in collagen pre-coated 24-well plates and after overnight incubation, the medium was replaced with the amount of each sample required to confer 1 μM BNN27, and incubated at 37 °C and in 5% CO
For the monolayer studies, hCMEC/D3 cells were seeded on Transwell filters (polycarbonate six-well, pore size 0.4 μm; Millipore Merck, Darmstadt, DE) pre-coated with type I collagen, at 5 × 10
The permeability of the samples was determined after the preparation of RHO-labeled nanoformulations. After adding the samples to the top of the permeation filter (0.2 μM RHO per filter), the transport of RHO was calculated by the fluorescence intensity measurements (540/585) of the samples taken from the basolateral portion at selected time points (
LIPs (0.5–1 mg/mL) were re-suspended in 10 mM HEPES (to eliminate potential artifacts from phosphate salts) while NEs were diluted in water. Then, all types of samples were negatively stained with 1% phosphotungstic acid in dH
For the in vivo study, C57BL/6J 8-week old mice were utilized. Animals were housed and maintained in a 12-h light/dark cycle and fed ad libitum. All procedures were performed according to the European Union policy (Directive 86/609/EEC) (carried out in compliance with Greek Government guidelines) and institutionally approved protocols (Veterinary Directorate of Prefecture of Heraklion (Crete) and FORTH ethics committee (License number: EL91-BIOexp-02)). Mice were anesthetized using an intraperitoneal injection of a ketamine (Nerketan 10, 100 mg/mL) and xylazine (Xylapan, 20 mg/mL) cocktail. Once the hind-limp was lost, the mice were fixed in the supine position, 25 µL of each formulation was administered to each mouse in 60 s intervals, via 1–2 µL dose alternatively into each nostril [[
After weighing, brain samples were homogenized in 300 µL ice-cold distilled water/methanol solution (25/75 v/v) and sonicated for 20 min at 4 °C. Next, three volumes of ice-cold acetonitrile were added, followed by sonication for 10 min and centrifugation at 14,000× g/15 min/4 °C. After an extra addition of 100 µL ice-cold acetonitrile to the supernatant, a 10 min-sonication and centrifugation at 14,000× g/15 min/4 °C were carried out. The supernatant was vacuum-dried at a SpeedVac without heating. Prior to analysis, samples were stored at −80 °C.
To quantify the BNN27 levels, the detection limit of the enzymatic method used for BNN27 quantification in the formulations (>2 ppm) was not low enough; therefore, a liquid chromatography mass spectrometry (LC-MSn) method was developed using deuteriated pregnenolone (pregnenolone 17,21,21,21-D4) as the internal standard (70 ng/mL). The analysis was performed on an LTQ-Orbitrap Velos mass spectrometer (MS) (Thermo Fisher Scientific, Bremen, Germany) connected to an Accela ultra-high-performance LC (UHPLC) system. An Acquity UPLC BEH C18 VanGuard pre-column (130 Å, 1.7 µm, 2.1 mm × 100 mm) coupled to an Acquity UPLC BEH C18 column (130 Å, 1.7 µm, 2.1 mm × 5 mm) was used. Quality control samples were prepared at three concentrations (low, medium, high) to monitor the instruments' performance and chromatographic integrity over time. Monitoring occurred in positive ion mode. The standard curve concentration range was 1–2000 ng/mL (Y = 0.000546519 + 0.000897604∙X; R2 = 0.9673; W:1/x). The injection volume was set at 5 µL, and the mobile phase flow rate was set at 0.2 mL/min. Mobile phase solvents were A (95% H
All results were expressed as mean ± S.D from at least three independent experiments. Most data were analyzed by using one way ANOVA followed by the Bonferroni post hoc test. p < 0.05 was considered statistically significant for all comparisons. When more factors were compared, two-way ANOVA was performed. The significance of comparisons is presented in the graphs as: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.
All calibration curves constructed both in the absence and presence of empty LIPs were linear; proving that the quantification method applied resulted in the accurate determination of BNN27 in LIPs (see Supplementary Materials Figure S1).
The BNN27 content of each LIP type was calculated from the appropriate calibration curve, after mixing equal volumes of the liposomal sample and pure ethanol, and applying the method described above. The results of studies carried out to optimize BNN27 loading in LIPs can be seen in Figure 1a. As seen, BNN27 loading could not confer a D/L (mol/mol) ratio higher than 0.056 that was realized when using an initial D/L ratio equal to 0.1 (mol/mol). Furthermore the loading efficiency was not significantly modified by using different initial D/L ratios, between 0.05 and 0.167.
Concerning the lipid membrane composition effect on BNN27 loading in LIPs, as seen from the results in Figure 1b, the addition of negative charged lipids (PG) resulted in a significant decrease in BNN27 incorporation into the LIPs. It is well-known that the composition of LIP lipid membranes determines the lipophilic drug partitioning/incorporation in LIPs, since these drugs are incorporated in the lipid membrane. Regardless of the decrease in BNN27 loading, the addition of the negative charge is very important for liposomal surface modification such as coatings with chitosan, as demonstrated previously and verified by the results presented in the following section.
From the Figure 1b results, the nanosize of both types (lipid membrane compositions) of BNN27 loaded LIPs was confirmed as well as their narrow size distribution. As anticipated, the addition of PG in the LIP membrane conferred a significant increase or the vesicle negative zeta potential.
The coating efficiency of chitosan-LIPs (CHT-Lip), together with the physicochemical properties of the various LIP types prepared, is presented in Table 1. As demonstrated, the vesicle coating efficiency was substantially higher in the case of PC/PG LIPs compared to PC LIPs, irrespective of the amount/type of CHT used. This is in agreement with a previous report that the CHT coating of LIPs cannot be further increased above the 0.1 w/w chitosan/lipid ratio [[
Uncharged PC vesicles had much lower CHT coating efficiencies (Table 1) compared to the negatively charged PC/PG LIPs. It was previously explained that in addition to the electrostatic interactions between positively charged chitosan and negatively charged LIPs, the coating process is regulated by a number of other mechanisms such as the formation of hydrogen bonds between the hydrogen of the polysaccharide and the nitrogen groups of the polar head of PC [[
The solubility of practically insoluble (in water) BNN27 in selected potential ingredients of NEs is reported in Figure 2. As seen, between the oils tested (Figure 2a), the highest solubility was measured in Capmul MCM, so this oil was selected as the oil phase for NEs.
Concerning the solubility of BNN27 in surfactants (Figure 2b), the highest solubility was measured in Tween 20 and Tween 80; therefore, we selected to study both surfactants in order to select the surfactant that acquired the optimal NE properties.
Finally, between the potential co-surfactants tested (Figure 2c), very high BNN27 solubilities were found in both the Transcutol and Transcutol/propylene glycol (1:1) mixture. In fact, we selected to use the Transcutol/PG mixture, since it has been previously reported to impart high stability and lower globule size (compared to other co-surfactants) to the NEs [[
The type of NE formed depends on the properties of the oil, surfactant, and co-surfactant. Many surfactants cannot lower the oil–water interfacial tension sufficiently to form NEs, so co-surfactant addition is necessary. Co-surfactants additionally ensure that the interfacial film is flexible enough to deform readily around each droplet as their intercalation between the primary surfactant molecules decreases both the polar head group interactions [[
Previously, by the construction of ternary phase diagrams, it was demonstrated that when Tween 80 was used a surfactant with the same oil phase (Capmul MCM) and co-surfactant system (Smix), the optimal surfactant/co-surfactant (Smix) ratio was 4:2 [[
After confirming that the 4:2 ratio (for surfactant/co-surfactants) was also optimal when Tween 20 was used as the surfactant, several NEs using 8% Capmul MCM as the oil, and increasing concentrations of the surfactant (S) + co-surfactants (Smix) between 24 and up to 44%, were prepared in order to select the composition that acquired the optimal NE properties. As seen in Figure 3, composition B6 had the best properties with regard to the globule size distribution and transmittance (%).
Thereby, we decided to use 44% concentration of S + Smix. Comparison of various NE formulations with Tween 80 or Tween 20 as the surfactant was then realized (after preparing various formulations) for the identification of the best surfactant. The results of their physicochemical properties and short-term physical stability study are presented in Figure 4.
As seen in Figure 4a, the NEs with Tween 80 had a significantly lower globule size for all oil concentrations tested, although their transmittance (%) was not affected by the different surfactants. Furthermore, Tween 80-containing NEs demonstrated higher stability (Figure 4c) during a preliminary 8 day at 25 °C stability study compared to the corresponding Tween 20-containing-NEs (Figure 4b), where a significant increase (p < 0.001) in globule size was observed after 8 day of storage at room temperature. In both types of NEs (with Tween 80 and Tween 20), when the oil concentration was increased so did the NE globule size. The last experiment showed that Tween 80-containing NEs had better properties and stability, and thereby Tween 80 was used as the surfactant in the following studies. For the latter decision, it was also taken into consideration that Tween 80 belongs to the class of non-ionic surfactants and is widely used since it is less toxic compared to ionic surfactants, and additionally, it is less affected by pH and ionic strength [[
After the selection of the ingredients and the optimal S + Smix percent, BNN27-loaded NEs (BNEs) were formulated using the method described above. Two types of BNN27-loaded NEs were prepared: one with an oil phase of 8% (w/w) and the other of 10% (w/w), which were evaluated for their properties and stability.
In general, a NE exhibits the characteristics of its external phase. There are several techniques for identifying the type of emulsion. Dilution studies are based on the fact that emulsions are soluble only in the liquid that forms their continuous phase. When diluted with water, no change was observed in the BNEs' droplet size and clarity, indicating that the BNEs are oil-in-water emulsions. Additionally, neither phase separation nor creaming was observed after centrifugation of the NEs, suggesting the stability of the systems. The physicochemical properties and quantitative test data of the formulated BNEs are shown in Figure 5a. The pH of all BNEs is between 4.05 and 5.73, which is within the previously considered normal pH range of nasal fluid (3.5–6.4) [[
The high degree of transparency of the non-coated BNEs verified that clear dispersions were formulated, while the CHT or CAR coated NEs had lower transmittance percentages due to the contribution of these components to turbidity. The low PDI of the non-coated BNEs indicates that they are a monodispersed system (Figure 5a).
All types of BNEs were also tested for drug content and found to demonstrate high BNN27 loading, ranging between approx. 90% and 99% of the amount of BNN27 used for their preparation (Figure 5a).
In the stability studies, the BNEs exhibited no precipitation of drug, creaming, phase separation, or flocculation on visual observation, and were found to be stable after centrifugation. When stored at 25 °C as well as at 4 °C (Figure 5b–e), only negligible changes in the quantitative parameters of the BNEs containing the 10%-oil phase were observed after 2 months of storage (Figure 5d,e). In contrast, the BNEs with the 8%-oil phase exhibited significant increases in their mean globule size (Figure 5b,c), especially during storage at 25 °C.
Similar stability studies were also carried out for the CHT- and CAR-coated BNEs (see Supplementary Materials Figure S3), where it was demonstrated that the BNEs with the 10%-oil phase were more stable for both types of coated BNEs compared with the corresponding BNE types with the 8%-oil phase. For this reason, BNN27-loaded NEs with 10% (w/w)-oil content (Capmul MCM) were used for the next in vitro and in vivo studies. The higher amount of oil phase in the BNEs would also provide the capability to load higher amounts of BNN27. Furthermore, it was previously reported the Capmul improved the NE-loaded drug permeation through the BBB barrier [[
Additionally, the drug content of all BNE types studied for their stability was measured at all time points, and no significant decrease was demonstrated in the BNN27 content (or else no drug leakage occurred) in any of the tested BNE types (see Supplementary Materials Figure S4).
The cytotoxicity of the prepared NEs toward the hCMEC/D3 and HEK-294 cells was evaluated by the MTT method. As seen in Figure 6, no toxicity was observed in any of the cell lines studied after 48 h of incubation with BNEs.
Transmission electron microscopy was performed to complete the characterization of the mucoadhesive nanoformulations. in Figure 7a, a polymeric chitosan membrane formed around the liposomal vesicles can be observed. It is interesting to highlight the different LIP mean-diameters observed by TEM and measured by DLS (which measures the hydrodynamic diameter), the last being more than two times smaller. For the BNEs, the TEM micrographs confirmed the differences in size of CHT-BMNEs with different percentages of the oil phase, as seen in Figure 7b,c. The latter differences are in agreement with the DLS measurements of the corresponding NE-types (Figure 5a).
Mucoadhesive properties were calculated as the percentage of mucin attached to LIPs or NEs. As seen in Figure 8, uncoated vesicles and NEs showed a small nonspecific mucoadhesion. As anticipated, PC/PG CHT-LIP is characterized by a strong positive surface charge (Table 1), which is why they exhibited significantly higher mucoadhesive properties compared to the PC CHT-LIPs. The uncoated PC/PG-LIPs were negatively charged and exhibited low mucoadhesive properties.
The adhesion process is rather complex and several theories have been proposed to explain the adhesion of polymeric materials [[
Concerning the mucoadhesive properties of the coated BNEs, as seen in Figure 8b, both polymers, CHT and CAR, resulted in a significant increase in the mucoadhesive properties of non-coated BNEs, however, the mucoadhesive properties of the CHT-coated BNEs were about two times higher than those of the corresponding CAR-coated BNEs. Indeed, it has also been previously observed that the mucosal capacity of CHT is significantly higher than the relevant properties of CAR [[
The permeability of the formulation-incorporated RHO across a cellular model of the BBB as well as the brain disposition of BNN27 following the intranasal administration of BNN-loaded formulation to mice were evaluated in the last part of the current study, in order to compare the various types of formulations. It should be clarified at this point that the hCMEC/D3 monolayer is considered as a model of "intact" BBB (with tight connections between the cells), while the nasal CNS barrier is proposed to be more "leaky" because of continuous neuron turnover [[
For the permeability study, formulations incorporating 1 mol% RHO as a lipophilic drug model were used. LIP (PC/PG) and CHT-coated LIP (coated with MMW CHT) as well as NE and CHT-NE with 10% Capmul were prepared; the physicochemical properties of the formulations are presented in Figure 9a.
During monolayer formation, the TEER of the monolayers was measured and was found to gradually increase from 31.5 Ω × cm
In preliminary cytotoxicity studies, it was confirmed that all the formulations used were non-cytotoxic toward the hCMEC/D3 cells following 2 h incubation with the cells at a final RHO-lipid concentration of 0.2 µM (see Supplementary Materials Figure S5), as also seen for the BNN27-incorporating formulations at a 1 µM concentration after 48 h (Figure 6). However, the substantial difference in the cytotoxicity of the two formulation types is a matter that needs to be highlighted. Indeed, RHO-incorporating LIPs (non-coated and coated) did not confer any cytotoxicity toward the hCMEC/D3 cells at 5 µM concentration (of RHO) after 48 h (Figure S5a), whereas the RHO-incorporated NEs diminished the cells even at 10 times a lower concentration (of RHO) and only 2 h incubation (Figure S5b).
Concerning the comparison between LIPs and NEs, interestingly, the translocation of NE-associated RHO and the corresponding permeability values were approx. three times higher for the corresponding values of LIPs. This is a very interesting finding, especially since such differences between LIPs and NEs have not been previously reported (to the best of our knowledge).
Concerning the effect of coating, the transport of the coated-nanoformulations (Figure 9b) as well as the corresponding permeability of the formulation-associated RHO (Figure 9c) were found to be lower than the corresponding values of the control (non-coated) nanoformulations, for both formulation types (LIPs and NEs) (Figure 9b), although in some cases, the differences noted were not statistically significant. This fact can be attributed to the adhesion of the CHT LIPs or NE droplets to the monolayer, resulting in delayed permeability. Free RHO (micellar solution) did not show any difference compared to LIP-RHO (not shown), while compared to the NEs, its permeability was significantly lower, which is in good agreement with the previously reported results [[
The calibration curve used for the calculation of the BNN27 concentration in the brain samples is presented in Figure S6 (see Supplementary Materials). As seen in Figure 10, the brain disposition of BNN27 after intranasal administration of BNN27 nanoformulations to C57BL/6J mice showed that CHT-BNEs achieved the highest levels of BNN27 in the brain compared to all of the other formulations used. Specifically, the BNN27 levels in the brain 1 h after nasal administration of CHT-BNE were more than 3.5 times higher than the corresponding concentrations after the administration of LIPs (coated or non-coated) or non-coated BNEs. At 2 h post-administration, the differences in brain disposition were lower, since all other formulations demonstrated increased brain concentrations of BNN27 at this time point (compared to the 1 h time point), except for CHT-BNE, but the BNN27 concentration after administration of CHT-BNEs was still more than 2-fold higher compared to all other nanoformulations (Figure 10b).
As demonstrated, 2 h post-administration, the non-coated BNEs achieved a higher BNN27 brain concentration compared to non-coated LIPs (although the difference was not statistically significant); however, nasal administration of CHT-coated nanoformulations resulted in higher brain concentrations of BNN27 in most cases, with the exception of LIPs at 1 h post-administration. Another very interesting observation from the data of Figure 10b is the very rapid brain disposition of CHT-BNE-associated-BNN27, which reached the highest BNN27 brain concentration already 1 h post-administration, in contrast to all of the other nanoformulations that demonstrated the highest BNN27 brain concentrations 2 h post-administration.
It has been reported that the natural biodegradable polymer chitosan may enhance the penetration and absorption of drugs through the nasal mucosa and may also delay mucociliary clearance [[
Herein, two types of nanoformulations, LIPs and NEs, were developed using biocompatible ingredients, optimized for drug loading, stability, and mucoadhesive properties, and evaluated as nanoformulations for the nose-to-brain delivery of BNN27. The effect of the CHT-coating was evaluated for both formulation types.
The results showed that NEs (consisted of Capmul MCM, Tween 80, Transcutol, and propylene glycol) had significantly higher in vitro BBB-crossing capability compared to LIPs (consisted of PC and PG) (Figure 9); however, the CHT-coating resulted in a slight decrease in the BBB monolayer transport for both formulation types, perhaps due to the adhesion of CHT-coated vesicles/globules on the apical side of the monolayer, thus preventing their transport to the basal side. In agreement with the monolayer permeability results, BNEs demonstrated significantly higher brain disposition of BNN27 compared to the BNN27-loaded LIPs (Figure 10). This result may be attributed to the previously reported effect of Capmul MCM to enhance drug permeability across the nasal mucosa and thus the brain disposition of drugs [[
A direct comparison of LIPs and NEs for nose-to-brain delivery of drugs has never been evaluated before; therefore, we cannot discuss the current results with respect to previous studies. We identified two studies reporting a direct comparison of LIPs and NEs for other drug delivery applications: one that compared the skin delivery of retinyl palmitate [[
In summary, the CHT-BNE formulation developed herein was demonstrated to confer faster and higher nose-to-brain delivery of BNN27 compared to CHT-LIPs. Such NEs could be considered as alternative systems for the brain delivery of lipophilic drugs following intranasal administration. Nevertheless, extended biocompatibility and toxicity studies are required to exclude any potential toxicity issues in relation to the cytotoxicity differences between the LIPs and NEs above-mentioned due to the high surfactant content of NEs.
Graph: Figure 1 (a) Effect of increasing the initial amount of BNN27 (or D/L mol/mol ratio) on the amount of BNN27 loaded in PC small unilamellar LIPs. (b) Physicochemical properties of BNN27-loaded LIPs. Each value is the mean from at least three independent samples.
Graph: Figure 2 Solubility of BNN27 in various potential ingredients of NEs, expressed as mg/mL. Each value is the mean of at least three independent samples. (a) Solubility in oils. (b) Solubility in surfactants. (c) Solubility in co-surfactants. *: p ≤ 0.05; **: p ≤ 0.01; ***: p ≤ 0.001.
Graph: Figure 3 Physicochemical properties of the NEs composed of Tween 20 (as surfactant), with 8% Capmul MCM as the oil and various S + Smix percentages (from 24% up to 44%).
Graph: Figure 4 (a) Physicochemical properties of NEs composed of Tween 20 or Tween 80 (as the surfactant). (b) Stability of Tween 20 containing NEs (for 8 days at 25 °C). (c) Stability of Tween 80-containing NEs (for 8 days at 25 °C).
Graph: Figure 5 (a) Physicochemical properties of the BNN27-loaded NEs (BNEs) composed of Tween 80 (as the surfactant) and oil content of 8% or 10%. Some BNEs were coated with CHT (CHT-BNEs) or Carbopol (CAR-BNEs). (b,c) Stability of BNE with the 8% oil-phase during 60 days of storage at 4 °C and 25 °C, respectively. (d,e) Stability of BNE with the 10% oil-phase during 60 days of storage at 4 °C and 25 °C, respectively.
Graph: Figure 6 Cell viability of the hCMEC/D3 and HEK-294 cells following the 48 h incubation of various BNN27 nanoformulation types at a BNN27 concentration of 1 µM.
Graph: Figure 7 TEM micrographs of the CHT-coated (a) BNN27-loaded LIPs (PC/PG LIPs coated with 0.1% MMW CHT), (b) BNEs with 8%-oil phase, and (c) BNE with 10%-oil phase.
Graph: Figure 8 Mucoadhesive properties (expressed as the capability to adsorb mucin (% adsorbed)) of: (a) CHT-coated LIPs and (b) CHT- or CAR-coated NEs. Non-coated formulations were evaluated under identical conditions (in all cases) for comparison. *: p ≤ 0.05; **: p ≤ 0.01; ***: p ≤ 0.001; ****: p ≤ 0.0001.
Graph: Figure 9 (a) Physicochemical properties of formulations used in the BBB-model permeability study. (b) Transport of formulation-associated RHO (%) across the monolayer with time (following incubation with LIP and NE formulations at 0.2 µM RHO concentration). (c) Permeability values of formulations (formulation-associated RHO permeability) across the hCMEC/D3 monolayer (calculated from the results of Figure 9b). ***: p ≤ 0.001; ****: p ≤ 0.0001.
Graph: Figure 10 (a) Physicochemical properties of the formulations used in the in vivo disposition study. (b) Brain disposition of BNN27 1 and 2 h post-intranasal administration of the BNN27 loaded liposomal and NE formulations. Each value is the mean of at least three independent samples. ****: p ≤ 0.0001.
Table 1 Physicochemical properties of the CHT-coated LIPs. Effect of lipid membrane composition and CHT type and amount.
Lipos CHT MW-CHT/LIP ( Mean Diameter (nm) PDI ζ-Potential (mV) Coating Efficiency (%) PC Non-coated 78.1 ± 1.15 0.281 −3.48 ± 0.49 - Low-0.1 123.7 ± 2.17 0.398 3.65 ± 0.27 3.5 ± 0.2 Med.-0.1 158.8 ± 6.27 0.405 4.57 ± 0.18 4.8 ± 0.6 PC/PG Non-coated 83.01 ± 1.36 0.263 −12.4 ± 0.689 - Low-0.1 822 ± 21.54 0.452 15.2 ± 0.47 79.9 ± 1.5 Med.-0.1 1147 ± 86.27 0.447 23.1 ± 1.41 85.1 ± 4.5 Med.-0.3 2563.5 ± 109.6 0.592 24.4 ± 1.273 78.2 ± 0.7 Med.-0.5 5155 ± 51.24 0.617 29.4 ± 0.689 81.4 ± 1.8
M.K.: Experiments of formulation development, Analysis of Results; Preparation of spreadsheets of results and graphs; Participation in writing and in Conceptualization; K.K.: In vivo experiment execution, Preparation of samples for preservation; T.K.: Analysis of in vivo study samples, Analysis and drafting of results, Supervision; E.S.: Analysis of in vivo sample execution, preparation of samples for analysis; A.M.: Experiments for in vitro cell studies, Supervision, Data (in vitro cell studies) analysis; P.K.: Execution and supervision in physicochemical characterization studies, Data Analysis; T.C.: Conceptualization and design of analysis of in vivo samples, Writing of draft paper, Supervision, Funding; I.C.: Conceptualization and Design of in vivo studies, Supervision, Funding, draft paper correcting; S.G.A.: Conceptualization and design of studies, writing of final draft correction, supervision during experiment execution (formulation development and in vitro studies), Funding. All authors have read and agreed to the published version of the manuscript.
The animal study protocol was approved by Veterinary Directorate of Prefecture of Heraklion (Crete) and approved by FORTH ethics committee (License number: EL91-BIOexp-02; Approval Code: 360667).
Not relevant with this study.
The authors declare no conflict of interest.
Ach: achetylocholine; BBB: blood–brain barrier; CHT: chitosan; CNS: central nervous system; D/L: drug to lipid ratio; DHEA: dehydroepiandrosterone; DLS: dynamic light scattering; FBS: fetal bovine serum; GABA: γ-amino butyric acid; HEK: human embryonic kidney cells; hCMEC/D3: immortalized human cerebral microvascular endothelial cells; HMW: high molecular weight; LDE: laser Doppler electrophoresis; LIP: liposomes; LMW: low molecular weight; MMW: medium molecular weight; LY: Lucifer yellow; MTT: 3-(4,5- dimethylthiazol-2-yl)-2,5 diphenyltetrazolium bromide; MW: molecular weight; NE: nanoemulsion; NGF: nerve growth factor; PBS: phosphate buffered saline; PC: 1,2-distearoyl-sn-glycerol-3-phosphatidylcholine; PG: 1,2-distearoyl-sn-glycero-3-phospho-(1′-rac-glycerol) (sodium salt); p75NTR: p75 neurotrophin receptor; RHO: Lissamine Rhodamine B phosphatidylethanolamine; TEER: transendothelial electrical resistance; TEM: transmission electron microscopy; TrkA: tropomyosin receptor kinase A.
The authors are thankful to Chemix SA, Athens Greece, Noveon distributors, for providing Carpobol 974P and Zygouropoulos SA, Athens Greece, Gattefosse distributors (in Greece) for providing the formulation excipients. The help provided in the TEM studies by Mary Kollia, Laboratory of Electron Microscopy and Microanalysis (L.E.M.M.), Faculty of Natural Sciences, University of Patras, is highly acknowledged.
The following supporting information can be downloaded at: https://
By Maria Kannavou; Kanelina Karali; Theodora Katsila; Eleni Siapi; Antonia Marazioti; Pavlos Klepetsanis; Theodora Calogeropoulou; Ioannis Charalampopoulos and Sophia G. Antimisiaris
Reported by Author; Author; Author; Author; Author; Author; Author; Author; Author