Cyclodextrin inclusion complex and amorphous solid dispersions as formulation approaches for enhancement of curcumin’s solubility and nasal epithelial membrane permeation

Background Curcumin is a compound that occurs in the rhizomes of the turmeric plant ( Curcuma longa ) and has shown potential for the treatment of illnesses including certain neurodegenerative diseases. The bioavailability of curcumin is hindered by its extremely poor aqueous solubility. Results This study aimed to apply formulation strategies such as inclusion complex formation with hydroxypropyl-β-cyclodextrin (HPβCD), as well as amorphous solid dispersion (ASD) formation with poly(vinylpyrrolidone-co -vinyl acetate) (PVP VA64) and hydroxypropyl methylcellulose (HPMC) to increase curcumin’s solubility and thereby its nasal epithelial membrane permeation. The curcumin formulations were evaluated by means of DSC, TGA, FT-IR, XRPD, microscopic imaging, aqueous solubility and membrane permeation across nasal respiratory and olfactory epithelial membranes. The solubility of curcumin was substantially increased by the formulations from 8.4 µg/ml for the curcumin raw material to 79.0 µg/ml for the HPβCD inclusion complex, 256.4 µg/ml for the HPMC ASD and 314.9 µg/ml for the PVP VA64 ASD. The HPMC ASD only slightly changed the membrane permeation of curcumin, while the PVP VA64 ASD decreased the membrane permeation of curcumin. The HPβCD inclusion complex enhanced the nasal epithelial membrane permeation of curcumin statistically significantly across the olfactory epithelial tissue and extensively across the respiratory epithelial tissue. Conclusion Complexation of curcumin with HPβCD enhanced the solubility of curcumin and thereby also increased its permeation across excised nasal respiratory and olfactory epithelial tissue. This indicated high potential of the curcumin-HPβCD complex for nose-to-brain delivery of curcumin for treatment of neurodegenerative diseases by means of intranasal administration.


Background
Turmeric, a spice that originates from the rhizomes of the Curcuma longa plant, has been used for centuries in Asia and India as an alternative herbal medicine for many acute and chronic diseases (e.g., certain cancers, bacterial infections, inflammatory conditions and oxidative conditions) [1][2][3][4][5][6].The biologically active compound found in turmeric is a natural polyphenol identified as curcumin (1,7-bis[4-hydroxy-3-methoxyphenyl]-1,6-heptadiene-3,5-dion) [2,4,7].Curcumin extracted from C. longa is accompanied by two other closely related curcuminoids, namely demethoxycurcumin and bis-demethoxycurcumin [1].Curcumin has been proven to be safe and well-tolerated by many studies, even at doses as high as 8 g/day [3,4,6].In a review by Gupta et al. [3], curcumin's clinical relevance is made prominent with reference to the different diseases that can be treated with this compound, which includes neurodegenerative diseases.Although the exact pharmacological mechanism of action for the many therapeutic actions of curcumin is not yet fully understood, previous studies indicated the possibility of biochemical pathway modulation to suppress inflammation and various other cellular activities [4,8].
Curcumin's therapeutic potential is hindered by its extremely low solubility, and in addition, it also exhibits relatively low membrane permeability and is susceptible to rapid metabolism [4].Consequently, curcumin has very low oral bioavailability, which means that large oral doses are required to achieve the desired therapeutic outcomes.This necessitates the consideration of using a different route of administration than the oral route.The nasal route of drug administration is a favorable alternative as the nasal cavity is highly vascularized and bypasses first-pass metabolism [9].This route of drug administration also has the unique capability of bypassing the blood-brain barrier for improved drug delivery into the brain [10].The respiratory region of the nasal cavity contains the trigeminal nerve that is linked to the pons and cerebrum [9][10][11], through which molecules such as curcumin can reach the brain.More importantly, the olfactory region in the nasal cavity has been associated with direct noseto-brain drug delivery [9][10][11].This region is innervated by the olfactory nerves that directly connect the nasal cavity to the brain through the olfactory epithelium [9,12].Drug molecules can move to the brain from the olfactory region via different mechanisms including passive diffusion across the olfactory epithelium, along the olfactory neuron's axon or through the gaps between the olfactory nerves and ensheathing cells [9].
The concentration gradient required to drive the passive diffusion process can be aided by enhancing curcumin's solubility [13], which can be done by utilizing formulation strategies such as preparation of inclusion complexes with hydroxypropyl-β-cyclodextrin (HPβCD) as well as the formation of amorphous solid dispersions (ASD) with polymers.HPβCD is a cyclic oligosaccharide with a cone-shaped conformation that has a hydrophobic core and hydrophilic outer surface, making it an ideal carrier for hydrophobic molecules such as curcumin to enhance its solubility [14].ASDs are formulations where the active compound is dispersed within the matrix of a hydrophilic polymer, thereby altering the crystalline packing of the molecules, resulting in the stable amorphous state with increased solubility [15].

Materials
The curcumin (98%) raw material used in this study was obtained from Toronto Research Chemicals Inc., Toronto, Canada.The purity of the curcumin raw material was provided on a Certificate of Analysis by the supplier and also determined by comparing the concentration of the same quantity of a curcumin reference standard by means of HPLC analysis.The hydroxypropylβ-cyclodextrin (HPβCD), poly(vinylpyrrolidone-co-vinyl acetate) (PVP VA64) and hydroxypropyl methylcellulose (HPMC) were sourced from DB Fine Chemicals, Johannesburg, South Africa.All solvents used in the preparation of the formulations, including ethanol and methanol, were purchased from ACE Chemicals, Johannesburg, South Africa.The ingredients to produce Krebs-Ringer Bicarbonate Buffer (KRB) used in the permeation studies and Lucifer yellow (LY) were acquired from Sigma-Aldrich, Johannesburg, South Africa.Consumables used, which include 100 µl crimp cells, were purchased from Mettler Toledo, Greifensee, Switzerland; 1-µm Acrodisc ® Glass Fiber membrane filters were sourced from Pall Corporation, Bengaluru; Costar ® 96-well plates were obtained from The Scientific Group, Randburg, South Africa; hydrochloric acid was sourced from Merck, Johannesburg, South Africa; acetic acid glacial and n-hexane were sourced from ACE Chemicals, Johannesburg, South Africa, while dimethylsulfoxide was purchased from Merck Chemicals (Pty) Ltd, Johannesburg, South Africa.The dialysis membranes (Membra-Cel ™ , 34 mm, 14 k Da) used in the compound release studies were sourced from Viskase ® Companies, Inc, Illinois, USA.

Analytical methods
Ultraviolet (UV) spectroscopic analytical method All samples were quantified for curcumin content by means of UV absorbance spectroscopy at a wavelength of 432 nm using a Spectramax ® Paradigm Multi-Mode microplate reader (Molecular Devices, San José, California, USA).The UV absorbance analytical method was validated in terms of linearity, repeatability (intra-day precision), intermediate precision (inter-day precision), specificity and accuracy [16].The limit of detection (LOD) and the limit of quantification (LOQ) were also determined.The stock solution of curcumin (42 µg/ml) used for the serial dilution was prepared in 40% v/v methanol (MeOH) in ultrapure water (UPH 2 O).
Fluorescence spectrophotometric analytical method Lucifer yellow containing samples collected during the membrane integrity experiment were analyzed using fluorescence spectroscopy with a Spectramax ® Paradigm Multi-Mode microplate reader with the excitation wavelength set to 485 nm and emission wavelength set at 525 nm [17].The fluorescence analytical method was also validated in terms of linearity, repeatability (intra-day precision), intermediate precision (inter-day precision) and accuracy, and the LOD and LOQ were also determined.

Curcumin-cyclodextrin inclusion complex preparation
The inclusion complex prepared between curcumin and HPβCD in this study was conducted according to a previously reported solvent evaporation method by Yadav et al. [18].In order to obtain a stoichiometric ratio of 2:1 for HPβCD:curcumin in solution, a quantity of 1.00 g curcumin was added to 400 ml of methanol, which was stirred until a clear solution was obtained and then a quantity of 6.12 g HPβCD was added to the curcumin solution and stirred until everything was dissolved.The solution was left at room temperature on a magnetic stirrer in a fume hood for 72 h to evaporate the solvent.The dry powder (i.e., curcumin-HPβCD inclusion complex) obtained was then passed through a sieve (aperture size: 850 µm), and the product was stored in an amber bottle in a desiccator.

Curcumin-polymer amorphous solid dispersion (ASD) preparation
The curcumin ASDs were prepared based on a solvent evaporation method published by Wegiel et al. [19].Two ASDs were prepared, each with a different polymer by employing a 1:4 weight ratio of curcumin:polymer.
For preparation of the curcumin ASD with PVP VA64, a quantity of 1.00 g curcumin was added to 400 ml ethanol, which was stirred until a clear solution was obtained.A quantity of 4.00 g PVP VA64 was then added to the curcumin solution and stirred until a clear solution was obtained.The solution was placed in a round-bottom flask and placed on a rotary evaporator (Rotavapor R-210 and heating bath B-491, Buchi Flawil, Switzerland) at 40 °C for 1 h.The residue was emptied into a glass baking dish that was placed in the fume hood (at room temperature) and left to evaporate all the remaining solvent overnight.The crystals obtained were ground into a fine powder with a mortar and pestle.
To prepare the ASD consisting of curcumin and HPMC, the following method was employed.A mixture of water and an organic solvent (ethanol) was used to ensure sufficient solubility of both curcumin and the polymer [13].A quantity of 1.99 g curcumin was allowed to dissolve in 85% v/v ethanol in water solution.When fully dissolved, a quantity of 8.00 g HPMC was gradually added, while stirring until no solid particles were visible.The solution was then transferred to a round-bottom flask and placed on a rotary evaporator (Rotavapor R-210 and heating bath B-491, Buchi Flawil, Switzerland) for 2 h at 40 °C.The residue was decanted into a glass baking dish and left overnight in a fume hood to evaporate all the solvent at room temperature.The powder (i.e., ASD) obtained after evaporation was ground into a fine powder with a mortar and pestle.

Characterization of curcumin-cyclodextrin inclusion complex and amorphous solid dispersions
Fourier transform infrared spectrometry (FT-IR) Fourier transform infrared spectrometry (FT-IR) was employed to identify if complex formation occurred between curcumin and HPβCD, as well as the formation of ASDs with the two selected polymers.FT-IR spectra were recorded for the curcumin-HPβCD inclusion complex, the curcumin-PVP VA64 ASD, the curcumin-HPMC ASD, simple physical mixtures of all the combinations and each of the raw materials individually, with a Shimadzu IR Tracer-100 spectrophotometer (Shimadzu, Kyoto, Japan), utilizing a spectral range of 750-4000 cm −1 , 64 scans and a 4 cm −1 resolution.
X-ray powder diffraction (XRPD) X-ray powder diffraction (XRPD) is an analytical technique commonly used to identify the presence of crystalline material within a sample of solid material.The sample is exposed to the radiation of X-rays, and the diffractometer then measures the intensity and the scatter provided by the sample.The diffractogram of a specific material is unique to its crystal structure [20].An XRPD pattern or diffractogram was generated using a PANalytical Empyrean diffractometer (PANalytical, Almelo, Netherlands) for each of the following: the curcumin-HPβCD-complex, the curcumin PVP VA64 ASD, the curcumin-HPMC ASD, physical mixtures of all the combinations and each raw material.The diffractograms were recorded under the following conditions: target, copper; voltage, 40 kV; current, 30 mA; divergence slit, 2 mm; anti-scatter slit; 0.6 mm; detector slit, 0.2 mm; monochromator; scanning speed, 2°/min (stepsize 0.025°; step time, 1.0 s).

Simultaneous thermal analysis (DSC/TGA)
To observe any heat flow endotherms and weight changes, simultaneous differential scanning calorimetry (DSC) and thermogravimetrically analysis (TGA) were performed.Thermograms of the curcumin-HPβCD inclusion complex, curcumin-PVP VA64 ASD, curcumin-HPMC ASD, physical mixtures of all the combinations and each raw material were recorded using a Mettler DSC 3 + (Mettler Toledo, Greifensee, Switzerland) connected to a computer equipped with Mettler STARe Default DB V14.00 software (V16.30a).A quantity of 5-6 mg of each material was transferred to crimp cells (100 µl) and placed in an automatic sampler, which transferred the cells into the furnace.The samples were heated to 195 °C at a heating rate of 10 °C/min and purged with nitrogen gas at a rate of 5 ml/min.An empty 100 µl crimp cell was implemented as the reference.
Visual inspection with polar and light microscopy Microscopic inspection of the samples was carried out using a Nikon Eclipse E4000 microscope equipped with a Nikon DS-Fi1 camera (Nikon, Japan).A small quantity of each material was centrally placed on a microscopic slide and inspected with the microscope at a magnification of 4× and 10×.The first comparison was drawn between the curcumin-HPβCD complex and a physical mixture of curcumin and HPβCD in the 1:2 molar ratio without the polarity filter.The second comparison was between the ASD and their physical mixtures, each containing 25% curcumin, assessed with the polarity filter.
Aqueous solubility Solubility studies were executed similar to a method published by Mangolim et al. [21].The solubility was determined over a period of 3 h in accordance with the time of exposure of each material in solution to the excised nasal epithelial membranes during the ex vivo permeation studies.For the aqueous solubility studies, a sufficient quantity of curcumin-HPβCD inclusion complex, curcumin-PVP VA64 ASD, curcumin-HPMC ASD, physical mixtures of all the combinations and curcumin raw material was added to 10 ml of ultrapure water in amber glass tubes to provide supersaturated solutions (5 mg/ml).The solubility studies were conducted in triplicate (n = 3).A water bath was preheated to 37 °C (± 2 °C) and the samples were then placed onto a rotary axis within the water bath.The tubes were rotated for 3 h whereafter it was removed, contents filtered (1-µm glass fiber membrane filter) and analyzed for curcumin concentration using UV-visible spectroscopy (wavelength = 432 nm).

Curcumin inclusion efficiency in hydroxypropyl-β-cyclodextrin complex
A method previously published by Mendes et al. [22] was used to determine the quantity of curcumin included in the complex formed with HPβCD.This method is based on the practical insolubility of cyclodextrin and its complexes in n-hexane [22].Approximately 5 mg of the complex was accurately weighed to which 0.1 ml acetic acid was added and it was then made up to 5 ml with n-hexane to remove the free curcumin molecules not included in the cyclodextrin complex.The supernatant was then removed and analyzed for curcumin content, which represented the free curcumin.The residue (curcumin-HPβCD complex) was dried at 35 °C for 20 min or until it was dry.After drying, 400 µl DMSO and 400 µl 2% v/v acetic acid were added to the powder.The mixture was allowed to precipitate over a period of 12 h.Thereafter, it was centrifuged at 3000 rpm for 10 min.The supernatant was diluted with MeOH (ratio: 1:1) and analyzed for curcumin content, which represented the curcumin included in the cyclodextrin complex.Analysis was performed using the validated UV absorbance analysis method on a Spectramax ® paradigm plate reader at a wavelength of 426 nm.The percentage curcumin included into the cyclodextrin complex (i.e., % inclusion efficiency) was calculated with the following equation:

Ex vivo permeation studies
Preparation of nasal epithelial tissue The sheep nasal epithelial tissue was excised from animals slaughtered at an abattoir and prepared for the ex vivo permeation studies as previously published by Haasbroek-Pheiffer et al. [17] and Gerber et al. [23].In brief, a reciprocating saw was used to remove the snout from a slaughtered sheep's head at an abattoir by a horizontal cut anterior to the eyes.The snout was then immersed in cold KRB (± 4 °C) and taken to the laboratory.At the laboratory, the skin was removed via blunt dissection and the snout was separated along the septal medial line using a bandsaw, allowing the (1) abstraction of the septum and exposing the respiratory tissue (ventral nasal conchae).First, a vertical incision across the most distal part of the ventral conchae was made slightly proximal to its round edge, followed by a second horizontal incision across the inferior portion of the ventral conchae parallel to the hard palate.This allowed the separation of the respiratory epithelial tissue from the underlying cartilage, providing a sheet of epithelial tissue.The olfactory epithelial tissue was removed from ethmoid conchae in a similar way by carefully detaching it from the cartilage.The dissected epithelial tissue was then cut into strips (approximately 1 cm × 2.5 cm) that were mounted onto the half-cells of a Sweetana-Grass diffusion chamber.After assembly of the half-cells of the Sweetana-Grass diffusion apparatus, a volume of 7 ml pre-heated KRB was added to each chamber, then placed into a heating block and connected to a carbogen source, containing 95% oxygen and 5% carbon dioxide.The mounted tissue was equilibrated at 34 °C for 30 min before the commencement of the permeation study.
Determining the integrity of the mounted nasal epithelial tissue Lucifer yellow (LY) was used as an exclusion marker molecule to provide evidence that the nasal epithelial tissues (i.e., respiratory and olfactory) mounted in between the Sweetana-Grass half-cells remained intact for the duration of the permeation study.LY permeation was determined as previously published by Haasbroek-Pheiffer et al. [17] and Gerber et al. [23].In brief, a LY stock solution (50 µg/ml) was prepared in pre-heated (34 °C) KRB and 7 ml of this solution was placed into each of the six apical chambers of the Sweetana-Grass diffusion chamber apparatus.The basolateral chamber contained pre-heated KRB with 5% v/v ethanol.Samples (180 µl) were withdrawn from the basolateral chamber at 20-min intervals over a period of 2 h.The withdrawn volume (180 µl) was replaced with an equal amount of preheated KRB containing 5% v/v ethanol.The samples were collected in a Costar 96-well plate, and the samples were analyzed for LY content by means of fluorescence spectroscopy (excitation: 485 nm, emission: 525 nm) using a Spectramax ® Paradigm Multi-Mode microplate reader as described above.
Curcumin permeation across nasal olfactory and respiratory epithelial tissue For the curcumin permeation studies done in the absorptive direction, experimental solutions were prepared for each curcumin material (i.e., curcumin raw material alone, curcumin-HPβCD complex, curcumin-PVP VA64 ASD and curcumin-HPMC ASD) in pre-heated KRB (34 °C) at a concentration of 3 mg/ml curcumin.A volume of 7 ml of each experimental solution was placed into each apical chamber, while a volume of 7 ml pre-heated ( Curcumin entrapment in nasal olfactory and respiratory epithelial tissue At the end of the permeation study (i.e., at 120 min), the nasal epithelial tissue was removed from the chamber and cut into small pieces with dissection scissors and placed in 1 ml methanol, after which it was ultrasonicated for 10 min, centrifuged and a sample was withdrawn from the supernatant to determine the quantity of curcumin retained in the tissue.These samples were collected in a clear Costar 96-well plate and analyzed for curcumin concentration with UV absorbance at a wavelength of 432 nm using a Spectramax ® Paradigm Multi-Mode microplate reader (Molecular Devices, San José, California).The permeation studies were conducted across respiratory and olfactory nasal tissues with six independent repeats (n = 6) for each experimental solution.

Curcumin release and permeation across synthetic dialysis membrane
To assess curcumin release from the cyclodextrin inclusion complex and ASD formulations (i.e., from the curcumin-HPβCD complex, curcumin PVP VA64 ASD and curcumin HPMC ASD), permeation studies were conducted across a synthetic dialysis membrane (Membra-Cel ™ , 34 mm, 14 k Da).Segments of the synthetic dialysis membrane, approximately 1 cm × 2.5 cm in size, were prepared and mounted on the half-cell of the Sweetana-Grass diffusion chamber apparatus.The experimental solutions were prepared following the same procedure for the permeation studies across the nasal epithelial tissue.The apical cham-ber was filled with the experimental solutions (n = 3), after which 180 µl samples were extracted from the basolateral chamber every 20 min for a duration of 2 h.These samples were then analyzed for curcumin content utilizing the validated UV absorbance method on a Spectramax ® Paradigm Multi-Mode microplate reader.

Data processing
The percentage permeation of curcumin across nasal epithelial tissue To calculate the percentage of curcumin permeated across the excised nasal epithelial tissue at each withdrawal time point of each experimental group, Eq. 2 was employed.
Furthermore, the apparent permeability coefficient (P app ) values for each experimental group have been calculated employing Eq. 3 [17].
where dc dt is the permeability rate (concentration/min, slope of the graph of percentage permeation as a function of time), A is the area of the tissue across which diffusion takes place (cm 2 ), and C 0 is the initial concentration (3 mg/ml) of curcumin applied to the apical chamber.

Statistical analysis
Statistical analysis of the permeation data was performed utilizing a one-way ANOVA (analysis of variance).Subsequently, the data were subjected to measurements of normality and homogeneity after which the post hoc tests were conducted using the Tukey adjustment method. (2) % Curcumin permeation = Curcumin concentration at time interval Curcumin concentration applied × 100 (3)

Analytical method validation
The validation results for the ultraviolet (UV) spectroscopic (for analysis of curcumin samples) and fluorescence spectrophotometric (for analysis of Lucifer yellow samples) analytical methods are portrayed in Table 1.

Characterization of inclusion complex and amorphous solid dispersions Fourier transform infrared spectrometry (FT-IR)
The FT-IR spectra for curcumin raw material (Cur RM), HPβCD raw material (CD RM), a physical mixture thereof (Cur-CD PM) and the prepared curcumin-HPβCD inclusion complex (Cur-CD complex) are presented in Fig. 1.
The first characteristic peak of the Cur RM spectra showcases the presence of free phenolic O-H stretching vibration with an absorption band located at 3505 cm −1 [25][26][27].Additionally, two absorption bands appear at 2967 cm −1 and 2841 cm −1 , resembling the asymmetric vibrations associated with the methyl group and methoxy group, respectively [27].Furthermore, the stretching seen at 1627 cm −1 is evidence of the keto-enol [17] located on the curcumin molecule and vibrations at 1510 cm −1 reveal C=O and C=C on the curcumin molecule.In the CD RM spectrum two absorption bands can be seen, which resemble the O-H and the C-H 2 molecules of the HPβCD, located at 3406 cm −1 and 2927 cm −1 , respectively [25].Another characteristic band on the CD RM spectra is assigned to the C-O-C bond of this molecule and exhibits stretching at 1033 cm −1 [25][26][27].In the spectrum for the physical mixture of curcumin and HPβCD, the absorption peaks of each of them are evident.On the other hand, the spectral data of the curcumin-HPβCD complex differ from those of the individual components.
The FT-IR results obtained for the curcumin-polymer ASDs are displayed in Fig. 2. The FT-IR spectrum for HPMC raw material exhibits a broad absorption band located at 3458 cm −1 .The same spectrum shows characteristic peaks of HPMC, at 1497 cm −1 and at 950 cm −1 .The physical mixture, containing curcumin and HPMC, showed a combination of the two component's spectra.A noteworthy peak broadening and shift from 1602 to 1587 cm −1 showcase a bond between the HPMC and curcumin.
The FT-IR spectrum of PVP VA64 presented with characteristic peaks at 1668 cm −1 and 1737 cm −1 .The physical mixture consisting of curcumin and PVP VA64 exhibited a combination of the peaks on the spectra of each individual material.When observing the FT-IR spectrum of curcumin-PVP VA64 ASD (Fig. 2b), an amine bending vibration can be seen at 1584 cm −2 .
Fig. 1 FT-IR spectra of curcumin raw material (Cur RM), HPβCD raw material (CD RM), a physical mixture of curcumin and HPβCD (Cur-CD PM), and the prepared inclusion complex between curcumin and HPβCD (Cur-CD Complex)

X-ray powder diffraction (XRPD)
The XRPD diffractogram of the respective raw materials and physical mixtures, the inclusion complex between curcumin and HPβCD (Cur CD complex) and the ASDs formed between curcumin and the two selected polymers, namely PVP VA64 (Cur PVP VA64 ASD) and HPMC (Cur HPMC ASD), are shown in Fig. 3.
In Fig. 3a, curcumin raw material exists in a crystalline state presenting a diffractogram with high-intensity peaks, whereas HPβCD produced a diffractogram showcasing its amorphous form [25].As expected, the simple physical mixture produced a diffractogram with a combination of the peaks in both its components, where these peaks were less potent and may be due to the dilution effect of HPβCD and curcumin.The prepared Cur-CD complex on the other hand produced the loss of some peaks that are present in the abovementioned materials such as 9.07°, 14.67°, 17.36°, 18.28° (2Θ).Furthermore, additional peaks appeared for the inclusion complex (14.2°, 26.01° and 27.14° (2Θ)) that did not previously exist in the other materials.
A striking difference can be observed in the X-ray powder diffractogram of the HPMC ASD prepared with HPMC and curcumin (Fig. 3b), and the same phenomenon can be observed when comparing the curcumin diffractogram to that of the curcumin PVP VA 64 ASD (Fig. 3c).

Differential scanning calorimetry (DSC)
The DSC thermograms of the relevant raw materials, their physical mixture and the Cur-CD complex are shown in Fig. 4a.Curcumin exhibited an intensive endothermic peak at 182.24 °C.In contrast to this, the DSC thermogram of the CD RM did not exhibit any endotherm in the scanned temperature region.The DSC thermogram of the physical mixture showcased a combination of the individual components with a melting point of 182.24 °C, corresponding with that of curcumin.The Cur-CD complex exhibited a shift in the endothermic peak at 173.73 °C.
Figure 4b and c shows the DSC thermograms for the ASDs of curcumin prepared with PVP VA64 and HPMC, respectively.The polymers (HPMC and PVP VA64) exhibited no endotherm, characteristic of the amorphous state thereof.The physical mixture of the above-mentioned polymers and curcumin produced a thermogram that had a subtle curcumin melting peak, possibly due to the dilution effect of curcumin by the polymer.The curcumin-HPMC and -PVP VA64 ASDs, on the other hand, showed no endotherms, therefore indicating the amorphous state of these ASDs [28].
In the TGA thermogram Fig. 5a, curcumin did not show any weight loss at the temperatures scanned and the Cur-CD physical mixture had a 3.45% weight loss.The CD RM produced a 4.73% weight loss, whereas the Cur-CD complex produced a noteworthy difference with a 9.02% weight loss.
Displayed in Fig. 5b and c are the TGA thermograms of the prepared curcumin-HPMC ASD and curcumin-PVP VA64 ASD formulations, respectively.The total weight loss of the HPMC sample was 3.5% between 40 and 90 °C.The curcumin-HPMC ASD, PVP VA64 raw material and

Visual inspection with polar and light microscopy
In Fig. 6, the visual differences between the particles of the prepared Cur-CD inclusion complex powder and the physical mixture between curcumin and HPβCD can be seen.The HPβCD in the physical mixture appears as spherically shaped transparent particles and the curcumin appears as bright orange particles.The Cur-CD inclusion complex (Fig. 4b) appears as light, yellow-colored particles with uniformity in texture and no distinction could be made between the two components.
Figure 7a displays the polarized microscopic image of a physical mixture of curcumin and HPMC, while Fig. 7b displays the polarized microscopic image of the Cur-HPMC ASD.It is evident that in the physical mixture, crystalline curcumin particles are visible in the microscopic image, whereas the microscopic image of the curcumin-HPMC ASD particles indicates it exists in the amorphous state.The second ASD formulation showcases the curcumin PVP VA64 ASD (Fig. 7d) and its physical mixture (Fig. 7c), and the same birefringence can be seen for the crystalline curcumin inside the physical mixture with none in the ASD.

Cyclodextrin inclusion efficiency
An inclusion efficiency of 98.86% was achieved by the curcumin-HPβCD complex prepared in this study.
In this study, the solubility of curcumin raw material was found to be 8.4 µg/ml at 3 h in water at 37 °C.The aqueous solubility of curcumin was substantially enhanced by the formation of the Cur-CD inclusion complex (79.0 µg/ml), Cur-HPMC ASD (256.4 µg/ml) and Cur-PVP VA64 ASD (314.9 µg/ml).

Curcumin permeation across nasal olfactory and respiratory epithelial tissue as well as the dialysis membrane
The ex vivo permeation results (P app values) for curcumin across excised nasal respiratory and olfactory epithelial tissues after the application of curcumin raw material (RM) as well as the CD inclusion complex and two ASDs are presented in Fig. 8.
The permeation of curcumin raw material (Fig. 8) was extremely low across the excised sheep nasal respiratory epithelial tissue (P app = 7.1 × 10 -8 cm/s) as well as across the excised sheep nasal olfactory epithelial tissue (P app = 6.1 × 10 -8 cm/s).The permeation of curcumin, when applied as the HPMC ASD, was lower than that of curcumin raw material (P app = 6.2 × 10 -8 cm/s) across the respiratory epithelial tissue.On the other hand, it was slightly higher than curcumin raw material across the olfactory epithelial tissue (P app = 6.6 × 10 -8 cm/s).The permeation of curcumin when applied as the PVP VA64 ASD was significantly lower across both the respiratory epithelial tissue (P app = 2.7 × 10 -8 cm/s) and olfactory epithelial tissue (P app = 2.4 × 10 -8 cm/s) as compared to that of curcumin raw material.Curcumin released from the HPMC ASD (P app = 3.03 × 10 -8 cm/s) across the dialysis membrane was slightly higher than that of the PVP VA64 ASD, but it was still relatively low (Fig. 8).
Curcumin permeation from of the Cur-CD inclusion complex was significantly higher across the olfactory epithelial tissue (P app = 8.9 × 10 -8 cm/s) and markedly higher across the respiratory epithelial tissue (P app = 8.2 × 10 -8 cm/s) as compared to that of curcumin raw material.In addition, the HPβCD inclusion complex demonstrated the highest permeation (P app = 4.9 × 10 -8 cm/s) of the three formulations across the dialysis membrane.

Curcumin entrapment in nasal olfactory and respiratory epithelial tissue
A comparison of the percentage of curcumin entrapped in the nasal olfactory and respiratory epithelial tissue at the end of the permeation experiment (120 min) for each experimental group can be seen in Fig. 9.
Curcumin PVP VA64 ASD produced a statistically significantly lower entrapment percentage of curcumin in the olfactory tissue (0.22%) when compared to that of the curcumin RM (1.26%).The curcumin-HPMC ASD presented with a significantly higher quantity of curcumin entrapped in the olfactory tissue (2.38%) when compared with the curcumin RM.The curcumin entrapped in the respiratory tissue (1.13%) was markedly higher than that of curcumin raw material (0.79%).
In line with the membrane permeation results obtained, the curcumin-HPβCD complex showed the highest membrane entrapment values, which were significantly higher than that of curcumin raw material in both nasal epithelial tissue types with 3.09% in the olfactory epithelial tissue and 2.51% in the respiratory tissue.

Discussion
The FT-IR spectral data of the curcumin-cyclodextrin inclusion complex and raw materials can be seen in Fig. 1.Cur RM spectra exhibited characteristic resemblances with previously published results [25,33].Seeing that the Cur RM did not exhibit a carbonyl stretch at ± 1700 cm −1 (characteristic of its di-keto form), it can be concluded that the keto-enol form of curcumin was used.The FT-IR spectrum of the CD RM also displayed similar characteristics as previously obtained [3,25].When referring to the physical mixture of curcumin and HPβCD, the absorption peaks of each of them are evident.On the other hand, the spectral data of the curcumin-HPβCD complex differ from those of the individual components.The overpowering absorption peaks seen are allocated to the CD RM, disguising the characteristic peaks associated with curcumin.The unique C-O and C-C stretching vibration seen at 1516 cm −1 confirms the conformation of a Cur-CD complex [25].
The FT-IR spectral data of the curcumin-polymer ASDs and raw materials can be seen in Fig. 2. The broad absorption band located at 3458 cm −1 in the HPMC raw material spectrum (Fig. 2a) is indicative of the O-H group found in HPMC [34].Characteristic peaks of HPMC are visible, at 1497 cm −1 , indicating vibration of the O-H group, and the peak at 950 cm −1 reveals a C-O stretching [29], which is in accordance with previously published literature [34].The physical mixture of curcumin and HPMC presented with a combination of the two component's spectra.The fingerprint zone found between 1500 and 750 cm −1 has a unique pattern for each compound normally consisting of bending vibrations [35].A noteworthy peak broadening and shift from 1602 to 1587 cm −1 showcase a bond between the HPMC and curcumin.This corresponds with a similar bond identified by Chai, Isa [36].The intermolecular interactions observed on the FT-IR spectrum of the prepared curcumin-HPMC ASD material and the disappearance of some crystalline peaks of curcumin are suggestive thereof that an ASD was indeed formed between curcumin and HPMC as polymer.
A recent article published by Dong et al. [37] noted that there are two characteristic peaks on the FT-IR spectrum of PVP VA64, namely the C=O located on the pyrrolidone ring at 1668 cm −1 and the other being the vinyl acetate at 1737 cm −1 .The physical mixture of curcumin and PVP VA64 presented with a combination of the spectra of each individual material.The PVP VA64 polymer within the curcumin-PVP VA64 ASD conceals most of the curcumin peaks except the amine bending vibration at 1584 cm −2 .From the FT-IR spectra, it can be concluded that the curcumin-PVP VA64 ASD has been formed.
The formation of an inclusion complex between curcumin and cyclodextrin has been confirmed by the XRPD diffractogram (Fig. 3), These findings align with results published in a previous study by Mangolim et al. [21].ASD formation with both HPMC and curcumin and PVP VA 64 and curcumin was confirmed by its characteristic amorphous state.The difference between the ASD diffractograms and the curcumin RM is due to a halo-pattern characteristic of the amorphous state of the ASD, which has been described in previously published work as a confirmation of ASD formation [38][39][40].A possible restricting factor of XRPD analysis is its inability to differentiate between the amorphous phases.The distinguishing of these phases is beneficial to identify phase separation (drug-rich or polymer-rich phases) in ASDs and the possibility of subsequent recrystallization [41].When DSC analysis is performed in conjunction with XRPD, this problem can be overcome [41].
DSC thermograms (Fig. 4) confirmed the endothermic peak for curcumin raw material at 182.24 C which correlates with the melting point of this compound [42].The CD RM did not exhibit an endotherm reaction, while the physical mixture presented with a melting point of 182.24 °C.A shift in the endothermic peak was observed at 173.73 °C by the Cur-CD complex, indicating an interaction between curcumin and HPβCD.This phenomenon was recently reported by Low et al. [43] as confirmation of the inclusion of curcumin into an HPβCD complex.It was not possible to identify the glass transition temperature (Tg) of the ASDs at the temperature range scanned.This aligns with the findings of Li et al. [44], who reported that when the curcumin content is below 25%, the ASD has no recrystallization or melting point and is therefore regarded as amorphous and protected from recrystallization at high temperatures.
Weight loss determined by the thermogravimetric analysis (TGA) of the physical mixture of Cur-CD can be ascribed to the loss of water [42].The weight loss presented by the HPMC RM can be attributed to sorbed water in the polymer matrix McPhillips et al. [45].In addition, water was used as a solvent in the preparation of the HPMC ASD.It is well-established in the literature that PVP molecules exhibit a pronounced affinity for water absorption [46]; therefore, the TGA thermogram in Fig. 5c presented a greater weight loss for the PVP VA64 raw material, which may be due to curcumin molecules replacing water molecules.
Light microscope images of a physical mixture of curcumin and HPβCD and a curcumin-HPβCD (Cur-CD) inclusion complex align with results previously published by Benucci et al. [47], who concluded that the lightercolored powder particles were inclusion complexes that enhanced the solubility of curcumin.Polarized light microscopy is a technique used to confirm the crystallinity of a compound in solid state.As discussed by Chasse et al. [48], it utilizes a polarizing filter that results in crystalline materials exhibiting polarization colors also known as birefringence and the amorphous state of a drug will appear dark.
The solubility of curcumin was substantially enhanced by the formation of the Cur-CD inclusion complex, Cur-HPMC ASD and Cur-PVP VA64 ASD.When curcumin is entrapped in the hydrophilic cavity of HPβCD, it causes a solubilization effect and thereby increases the apparent solubility of curcumin [49].Prepared ASDs have potential energy, which was released during the dissolution process, enabling curcumin to reach a supersaturated state [50].The increased solubility of curcumin could provide a higher concentration gradient across the membrane, which could act as a thermodynamic driving force to propel the curcumin molecules across the membrane [13].Another advantage described for ASD formulations is the ability to form liquid-liquid phase separation, which can produce small curcumin-rich droplets in solution that will serve as a reservoir and replace the curcumin in the solution that permeates across the membrane [51].
Finally, the curcumin RM, CD inclusion complex, Cur-HPMC ASD and Cur-PVP VA64 ASD were evaluated in terms of permeability across excised ovine respiratory, olfactory and dialysis membrane.Permeation from the HPMC ASD was lower than that of curcumin raw material across the respiratory epithelial tissue and only slightly higher across olfactory epithelial tissue, while permeation from the PVP VA64 ASD was significantly lower than the curcumin RM across both respiratory and olfactory tissue.
The permeation of curcumin from the HPMC ASD showcased a higher permeation when compared to the PVP VA64 ASD across both epithelial tissue types, which can possibly be explained by an interaction between the charged chains of HPMC and the phospholipid bilayer of the epithelial membrane.This may have contributed to the increased membrane permeability, due to the phospholipid bilayer disruption [52,53].Although the HPMC ASD showcased an increased permeation when compared to that of the PVP VA64 ASD, it did not increase the permeation of curcumin when compared to the raw material (control), probably due to the relatively slow release of the curcumin molecules as observed in the results obtained from the dialysis membrane permeation study.
Yu et al. [54] explained that the in vitro release of the compound from the ASD is directly proportional to the compound/polymer ratio, where a higher compound/polymer ratio (e.g., 1:30) resulted in a slower drug release rate.In the PVP VA64 ASD, a 1:25 curcumin:polymer ratio was used and it is hypothesized that this has caused a slow release rate of curcumin from the polymer matrix in the ASD, which restricted the membrane permeation of the curcumin molecules.Maincent, Williams [50] mentioned that the diffusion pathway of compound molecules from the polymer matrix in ASDs decreases as the polymer swells, thereby restricting the diffusion of compound molecules from the polymer matrix.Another possibility was proposed by Schittny et al. [55] namely that the polymer formed a gel-like layer on its outer surface through which the compound will have to diffuse before it is released in order to become available to permeate across the membrane.This explanation was substantiated by the results obtained from the curcumin release study conducted across the dialysis membrane from the PVP VA64 ASD (incorporated in Fig. 8).Given that the combined molecular size of the ASD containing curcumin and the PVP VA64 polymer exceeded the 14 kDa cutoff limit, curcumin molecules that were not released into the solution were unable to traverse the membrane, which resulted in the relatively low P app value of 2.64 × 10 -8 cm/s.It can therefore be concluded that the poor release of curcumin from the PVP VA64 polymer could have contributed to the poor membrane permeation.
The increase in permeation was possibly due to the increased concentration gradient as a result of the solubility increase of curcumin when formulated as a cyclodextrin inclusion complex [56].Another characteristic of the curcumin-HPβCD complex is that the compound is directly deposited onto the cell membrane after dissociation from the complex [57], resulting in an additional factor to increase the concentration gradient and therefore contributing to a higher passive diffusion rate [56].
From all the formulations investigated, the curcumin molecules were therefore released at the highest rate from the HPβCD complex.Although the permeation of curcumin across the dialysis membrane (without a phospholipid bilayer) was not as high as across the nasal epithelial tissue from the HPβCD complex, it can possibly be attributed to the interaction of cyclodextrin with the cholesterol composition of the biological membrane that increased the permeation of curcumin.According to dos Santos et al. [58] and López et al. [59], HPβCD can increase biological membrane fluidity by disruption (removal of cholesterol molecules) and this can increase drug permeation across the membrane.
In general, the permeation of curcumin was higher across the excised olfactory epithelium than across the respiratory epithelium.This is in agreement with previously published work by Fransén et al. [60].They proposed that it was possibly due to transport along the olfactory nerves, the gaps between the nerves and other diffusion mechanisms.Furthermore, Du et al. [61] explained that the differences in the morphology and thickness of the epithelial tissue types from the different nasal regions may contribute to the difference in their permeability.
Similar to the permeation results, the curcumin PVP VA64 ASD produced a significantly lower entrapment percentage of curcumin within the olfactory tissue than the curcumin RM, and as previously discussed, it could likely be due to the curcumin molecules being released at a low rate and extent from the PVP VA64 polymer matrix [50].The higher percentage curcumin entrapped within the respective nasal tissues after administration of the curcumin-HPMC ASD could be due to the higher concentration gradient combined with mucoadhesion [48,54] and potential membrane disturbance [17,54], can be caused by the charged HPMC polymer chains, which might have been responsible for the increase in the percentage of curcumin delivered into the membrane.Due to the highly lipophilic properties of the curcumin molecules, they remained inside the membrane and did not partition into the aqueous environment on the basolateral side.The significant membrane entrapment of curcumin from the curcumin-HPβCD complex is ascribed to the increased concentration gradient across the membrane providing the driving force to increase the diffusion of curcumin molecules into the membranes.However, due to the highly lipophilic nature of curcumin, it did not partition out into the aqueous environment on the basolateral side [56].

Conclusion
The curcumin-cyclodextrin (HPβCD) inclusion complex elevated the solubility and nasal epithelial membrane permeation of curcumin substantially across the respiratory epithelial tissue and statistically significantly across the olfactory epithelial tissue.The HPβCD inclusion complex therefore showed potential as a formulation approach to enhance intranasal delivery of curcumin into the systemic circulation (across respiratory epithelial tissue) and into the brain (olfactory epithelial tissue).Enhanced solubility and intranasal delivery of curcumin by means of HPβCD inclusion complex formation means that lower doses can be administered to reach effective concentrations at the site of action.This has the additional advantage of decreasing side effects.Both the ASD formulations (HPMC and PVP VA64) increased the solubility of curcumin noticeably.However, the HPMC ASD only slightly influenced the curcumin permeation as compared to that of curcumin raw material, while the PVP VA64 ASD decreased the permeation significantly probably due to slow release of the curcumin molecules.The ASDs formulated for curcumin with the selected polymers (HPMC and PVP VA64) therefore did not show potential as a formulation approach to enhance intranasal delivery of curcumin.

Fig. 3
Fig. 3 XRPD diffractogram a of curcumin raw material (Cur RM), HPβCD raw material (CD RM), a physical mixture of curcumin and HPβCD (Cur-CD PM), the prepared curcumin-HPβCD inclusion complex (Cur-CD Complex), diffractogram b ASD prepared with curcumin and HPMC (Cur HPMC ASD), relevant raw materials and physical mixture, and, diffragtogram c with PVP VA64 (Cur PVP VA64 ASD), relevant raw materials and physical mixture

Fig. 4 Fig. 5 Fig. 6
Fig. 4 DSC thermograms of curcumin raw material (Cur RM) including that of a HPβCD raw material, physical mixture (PM) as well as inclusion complex (Cur-CD complex), b HPMC raw material, physical mixture (PM) and ASD, c) PVP VA64 raw material, physical mixture (PM) and ASD

Fig. 8 Fig. 9
Fig.8 Apparent permeability (P app ) values of curcumin raw material (CURCUMIN RM), curcumin-HPβCD complex (CD COMPLEX) and the ASD formulation with curcumin and HPMC (HPMC ASD) as well as curcumin and PVP VA64 (PVP VA64 ASD), respectively, across excised sheep nasal mucosa as well as the dialysis membrane.*The curcumin-HPβCD complex group showed a statistically significant (p < 0.05) difference when compared to the curcumin raw material (control) across the olfactory nasal tissue.# The curcumin-PVP VA64 ASD also resulted in a statistically significant (p < 0.01) difference when compared to the control across both tissue types

Table 1
Validation results for the ultraviolet (UV) spectroscopic and fluorescence spectrophotometric analytical methods