Preparation, Physicochemical Properties, Invitro And In Vivo Release Evaluation of Chitosan Decorated Curcumin Loaded Niosome

Purpose: Based on several pharmacological activities of curcumin (Cn), it has been introduced as an ideal candidate for different neurological disorders. But poor solubility, rapid clearance and low stability have limited its clinical application. Development of curcumin loaded smart niosome for crossing blood brain barrier could be and interesting option to overcome these limitations. Object: The aim of this study is preparation and characterization of chitosan decorated curcumin loaded niosome(CH-CLN) and evaluation of invitro release and in vivo bioavailability and bio stability of curcumin in the brain tissue. Methods: Niosomal formulations were prepared by modied heating method by using 3-factor and 3 level mixture design. The formulations were characterized using DLS, zeta potential, Fourier transform infrared (FTIR) spectroscopy, high performance liquid chromatography (HPLC) and transmission electron microscopy(TEM). The entrapment eciency and invitro release were also evaluated. Wistar rats were subjected to intraperitoneal injection (i. p.) of formulations, and curcumin 15 minutes before perfusion. Cn concentration in different parts of the central nervous system, liver, and plasma was analyzed for in vivo analysis. Results: Encapsulation eciency was obtained 75% for optimized formulation and presented sustain release, followed by the Hixon-Crowell model. The particle size was about 100nm with polydispersity index (PDI) of 0.2. Chitosan decorated niosomal formulation increased Cn concentration in central nervous system. Conclusion: We can conclude the chitosan decorated curcumin loaded niosome improved bioavailability of curcumin in brain tissue and cold be a promising tool for crossing blood-brain barrier

Although Cn is claimed to be promising in numerous therapeutic interventions, it has demonstrated a poor bioavailability due to its poor aqueous solubility and a low stability against alkaline pH conditions. More importantly, Cn exhibits an extensive rst-pass metabolism, which is known as the main reason for the limitation of the retention time in the body. Accordingly, a designed carrier system could considerably increase the range of its achievable pharmaceutical applications [6 , 7].
In the last two decades, different delivery systems containing Cn have been developed; for instance, cyclodextrin; polymeric microparticles; and lipid systems such as lipid nanoparticles, niosome, liposomes, nanostructured lipid carriers, and nanoemulsions [7][8][9][10][11]. The delivery system not only solves the mentioned Cn downside, but it also increases permeability via the BBB for brain drug delivery.
Noticeable attention has been drawn to niosomes stemming from their exceptional stability and exemplary property, as loaded by hydrophilic or hydrophobic molecules. Niosome can overcome the enhance the bioavailability of Cn via increasing its retention time in the blood stream and passing the BBB [12 , 13]. Moreover, there have been many efforts to improve the stability, and target delivery and pharmacokinetic behavior of niosomes. In this regard, different materials such as polyethylene glycol, hyaluronic acid, antibodies, and chitosan were utilized for the reformation of the niosme surface. Accordingly, chitosan[β-(1-4)-linked d-glucosamine (deacetylated unit) and N-acetyl-d-glucosamine (acetylated unit)] are the most frequently utilized polymers in the formulation of coated niosomes(Khalifa and Abdul Rasool 2017).
In the same context, Mythri et al. (2007) showed that retention time of Cn in rat serum extended by the use of Cn-phospholipid (Mythri et al. 2007 To the best of our knowledge, Cn, CLN, and CH-CLN distributions in the brain, plasma, and liver of Wistar rat have not been reported so far. Quantity of Cn in the target site is considered as an essential factor for clarifying the therapeutic effects of the drug in those organs. Hence, in this study, we presented the chitosan coated niosome as Cn carrier, to validate the hypothesis that CH-CLN could target the brain and increase the BBB permeability, which is distributed at different regions of brain, liver, and plasma.
Material And Methods

Mixture Design (M.D.)
In this study, with the aim of altering the formulations, M.D. with three-level, three-factor and seven experimental runs were adopted. Table 1 shows the independent and dependent variables as well as their levels. The obtained colloid was added to a preheated (5 min, 60°C) mixture of Cn (100 µM (Sharma et al. 2015b)), glycerol ( nal concentration of 3%, v/v), DCP (DCP: Surfactant, 0.1 molar ratio) and ethanol (1 cc), which were stirred ( at 60°C, 1000rpm) on a hotplate stirrer (IKA®C. MAG HS7 Safety Control, IKA, Malaysia) for 60 min. The reaction was performed under a nitrogen atmosphere in a handmade glass vessel introduced by Mozaffari. Subsequently, sonication of the sample was performed using a probe sonicator (Sonopuls HD-3100, BANDELIN electronic GmbH & Co. Germany) for 16 min (180 sec "pulse on" and 30 sec "pulse off"). After the preparation of the loaded niosome, they were kept for 30 min at room temperature.

Preparation of Chitosan Coated CLN (CH-CLN).
The CH-CLNs were prepared according to Marianecci with some modi cations (Rinaldi et al. 2020). At rst, CH-CLNs were made by the addition of optimum formulation of CLNs into CH solution (1:2, v/v). The mixture was then stirred for 60 min at 25°C. Thereafter, the obtained mixture was adjusted to pH 4.5 (NaOH, 1M), and then sonication of the sample was performed using a probe sonicator for 5 min (180 sec "pulse on" and 30 sec "pulse off).

Cn Determination
The Cn measurement was performed at 40°C on a High-performance liquid chromatography (Waters, USA). The HPLC system contained a Waters 1525 binary pump, and 2489 Uv-Vis detectors at 420 nm. Also, a reversed-phase Inertsustian Swift C18 column (4.6 × 250 mm, particle Accordingly, both matrices contained 0.1% (v/v) Tween80.
To evaluate the kinetic behavior and release mechanism, the result derived from release studies was tted into different mathematical equations as demonstrated in Table 4.

Drug administration and accessing its bioavailability
The rats were randomly divided into the following three groups (4 rats in each group): (1)  After anesthetizing, the cardiac puncture was utilized to collect 400 µl blood with a heparinized syringe. Thereafter, the rats were perfused with normal saline to remove their brain and liver. Moreover, to evaluate crossing the BBB, the cerebral cortex, hippocampus, cerebellum, and striatum were collected.
Afterward, the samples were stored in microtubes at -80°C. Cn was extracted from plasma, and different regions of the brain and liver according to the protein precipitation technique proposed by Ravi (2018) with some modi cations (Dalvi et al. 2018). Brie y, 400 µL methanol was added to 90 µL Plasma. The mixture was then vortexed for 2 min. Subsequently, the sample was centrifuged at 14000g for 20 min at 4°C.
Following that, the supernatant was dried under N 2 gas at 40°C. Finally, 100 µL mobile phase was added for reconstitution, which was then analyzed by high-performance liquid chromatography (HPLC, Waters equipped with pump1525 binary pump along with UV2485and FLD2475 detector, USA).
Different regions of brain and liver tissues were homogenized using a probe sonicator for 1min (25son 15 s off) at 4°C in PBS (1:4 W/V, pH: 7.4). Next, the mixtures were centrifuged at 14000 g for 20 min at 4°C. Then, 100 µL of supernatant was moved to a new microtube, and 200 µL of methanol was added to it. After 2 minutes of vortexing, the samples followed the producer same as a plasma sample.

Statistical data analysis
In this study, all data were presented as mean ± standard deviation. The responses obtained from experiments were analyzed using the software Mini Table 18. Also, the possible mathematic models were analyzed using ANOVA one way.  Table 3 Summary of regression analysis for responses: Y1 (particle Size in nm), Y2 (encapsulation e ciency in %), Y3 (PDI), and Y4 (zeta potential) for niosome samples. The effect of experimental variables on particle size The particle sizes were within the range of 70-153 nm, which are shown in Table 2. The results indicate that the Chol concentration has signi cantly affected the niosome size, in agreement with previous ndings ,which showed that niosome particle size has linearly increased along with the Chol concentration (Goyena and Fallis 2019). The results also revealed that particle size has been obviously augmented by increasing the span60: tween60 ratio, which is in agreement with the reports published by Rajabzadeh (2017)  At a low Chol concentration, the shell is more exible and pliable to the consequences of ultrasound waves. Hence, by increasing the Chol amount, the shell rigidity and ultrasound power resistance increase; and niosome with larger particle size would be produced (Nasseri 2005). The

Effect of variables on EE
The EE of CLNs was between 33% (NHM810) and 90% (NHM840). in order to nd the relationship between the independent factors and the EE of niosomes, Minitab 18 was used.
As reported by Nadzir (2018), the EE is correlated with the particle size, since the entrapment of Cn increased the vesicle diameter (Nadzir 2018).
The effects of variable concentrations on EE are shown in Table 2 and Fig. 1(d, e, f). The encapsulation e ciency was linearly increased by increasing the Chol content, due to the membrane stability brought by Chol (Hayashi et al. 2011). This data is in agreement with the results previously reported by Basiri (Basiri et al. 2017a) and Rinald (Kassem et al. 2017).
Another important critical parameter affecting the EE is surfactant characteristics such as alkyl chain length and surfactant content. It was indicated that the surfactant ability to form the vesicle mostly depends on the balance between hydrophobic and hydrophilic portions and surfactant structure (Noronha et al. 2013 et al. 2019). Accordingly, the high PDI value in the case of NHM840 could be attributed to the agglomeration problems. Figure 1(g, h, i) re ects the variations in the size distribution dependency to the Chol content. The results indicate that the higher value of the PDI belonged to the sample, which had a higher Chol content.

The effect of variables on Zeta potential
Zeta potential values of different samples from − 23 to -30mV in Fig. 1 show that reduction of the HLB value means an rise in hydrophobicity, and then zeta potential increases. Also the surfactant type might affect the zeta potential (Sadeghi Ghadi et al. 2019). 3.1.6. Optimization by M.D.
The MD de nes the relationship between independent and dependent variables. In addition, the multiple response optimizations were carried out to predict three set levels of experimental conditions by the minimum particle size and PDI, maximum EE, and optimum zeta potential. In this study, the maximum EE was predicted to be obtained at a combined level of 0.85 (g), 0.219 (g), and 0.0953 (g) of Span 60, Tween 60, and Chol, respectively (Fig. 2). EE has linearly increased by increasing the HLB; and PDI, size, and zeta potential decreased by increasing the HLB value. These ndings are in the relevant agreement with the previous studies (Basiri et al. 2017b). Also, it was found that in the range of set levels of the main preparation conditions, all seven combinations showed acceptable particle sizes (70-153 nm).

Fourier Transform Infrared (FT-IR) spectroscopy
FT-IR spectra were performed in order to consider the functional groups and their possible interactions in the prepared niosome structure. As shown in Fig. 3, FT-IR spectrum of Chol indicates the characteristic peak of hydroxyl stretching at 3400 cm − 1 , and aliphatic C-H stretching in the region between 2800 and 2990cm − 1 . The characteristic peaks located at 2918 and 2851 cm − 1 in Tween 60 spectrum (Fig. 3, c) were shown to be related to asymmetric and symmetric vibrations of methylene (− CH2) groups, respectively.
The formation of niosomes was proven with two characteristic peaks at 1737 and 1106 cm − 1 (Fig. 3, a,  In CH-CLNs, the absorption peak at 1415cm − 1 in CH-CLNs can be attributed to the chitosan OH bending (Fig. 3. g). Moreover, the C--O--H deformation peak of CLN at 1106 and 1045cm − 1 were shifted to 1111 and 1051 cm − 1 in CH-CLN, respectively, due to the formation of the hydrogen bonds between chitosan and niosome. 3.4. CLN and CH-CLN Sizes, morphologies and zeta potential DLS, TEM, and AFM were used to consider their sizes and morphologies. The TEM and AFM images of the optimum CLN and CH-CLN are shown in Fig. 4.
A large particle size in the AFM images can be attributed to the vesicle aggregation that occurred during the air-drying production of the sample preparation. To evaluate the AFM validity result, CH-CLN and CLN were supplementary investigated using TEM. A relatively spherical shaped morphology for the vesicles was demonstrated by the TEM picture as well, exhibiting a unilamellar membrane for the prepared CLN and CH-CLN. Notably, no aggregates were observed for the prepared samples in TEM (Zou et al. 2014).
The vesicle diameters determined from DLS, AFM, and TEM are illustrated in Table 5. Besides, particle sizes observed from AFM and TEM images were within the range of 40-100 nm. These results are in agreement with the DLS results, which are more precise because the number of the analyzed particle by AFM and TEM were smaller than DLS (Rinaldi et al. 2018).
The particle size of CH-CLN was more prominent than CLN, suggesting the formation of an extra chitosan layer on the CLN surface (Guo et al. 2003).
It is clear from Fig. 4

Cn Release Study
In this study, the optimum samples were used to study the release behavior in the SGF and SIF. The cumulative release pro le of the niosomes is shown in Fig. 5 (I). Accordingly, the results reveale that CLN and CH-CLN released 50% and 35% of Cn within 8h, respectively. Moreover, both of the CLN and CH-CLN exhibited a sustained release pro le. zero-order, rst-order, Higuchi, and Hixon-Crowell models. (Table 4). Figure 5(II) shows the different model releases of CLNs and CH-CLN.
Comparing the four equations reported in Table 4 and Fig. 5(II), it can be said that the release of Cn followed the Hixson-Crowell model, due to the higher correlation coe cient (R 2 ). Furthermore, the model of Hixcon-Crowell was based on the corrosion mechanism, which explains that Cn release is a corrosion -controlled model (Xu et al. 2019). Overall, it can be reported that the chitosan, which coated the niosome surface, affected the way the drug was released, because it has an interaction with release media such as salt and biological uid, leading to a better control of drug release. 3.6 Distributions of Cn, CLN, and CH-CLN in organs after i.p. injection

HPLC analysis and method validation
The HPLC calibration curve for plasma, and different regions of the brain and liver showed an excellent linearity (R 2 > 0.995) over the ranges of 0.4, 0.8, 1.2, 1.6, and 2µg/ml. Also, the LOD (limit of detection) and LOQ (limit of quanti cation) Cn were obtained as 0.0186 and 0.0565 µg/ml, respectively. Cn inter assay and intra assay precision, accuracy, and precision value in the brain, plasma, and liver were 15%. Altogether, these results show the validity of the method in this study.

In vivo study
As shown in Fig. 6, the concentrations of Cn have signi cantly increased in plasma, liver, and brain regions when loaded in niosomes nanoparticles, compared to free Cn, which indicate the improved stability and delivery of CLN and CH-CLN.
It is noteworthy that, there were some crucial differences in Cn existence in the brain, liver, and plasma. Cn, in its natural form, was not detected in the above-mentioned organs. Also, a difference between Cn and nanocarrier can be attributed to the resistance to liver metabolism (Tsai et al. 2011), (Raza et al. 2017) and BBB nature.
The BBB is a protective reticulum around the brain that only allows the penetration of low molecular weight and/or small molecules with a high lipid solubility. Thus, passing BBB seems di cult for curcumin. In nanocarrier, the appropriate size and existence of tween 60 in niosome structure could act as an apolipoprotein E anchor in the bloodstream, which can consequently improve passing BBB(Hombach and Bernkop-Schnürch 2009). The vesicles can also interact with the LDL receptor and pass the BBB(Goyena and Fallis 2019).
The Cn distribution and concentration in rat organs were increased by modifying the niosome surface. The results demonstrate that Cn concentration was signi cantly higher in different regions of the central nervous system of rats that received CH-CLN compared to CLN: 16 (0.019; 0.001 mg/Kg) in the striatum, 16 (0.019; 0.001 mg/Kg) in the cerebellum, and 3.5 (0. 320; 0.009 mg/Kg) in the cerebral cortex. Accordingly, it suggests that positively charged chitosan-coated niosome can cooperate with all the negatively charged elements in a biological milieu, which consequently increases CH-CLN permeation ability [50,51].
Cn has a rapid degradation under physiological condition, so its encapsulation in niosome and chitosan-coated niosome may hinder its degradation. Also, in the case of CH-CLN, Cn concentration in rat plasma has reduced more gradually compared to the CLN form, due to the protective effect of chitosan layer (

Conclusion
In the present study, CLN and CH-CLN were successfully synthesized using a modi ed heating method. All the niosomes were < 200 nm in diameter. Also, M.D. was used to optimize the niosome components to achieve the minimum vesicle size and a high EE. Then, chitosan was utilized to modify the optimal niosome surface.
Free Cn, CLN, and CH-CLN were i. p. administrated on Wistar rat and the presence of Cn in the brain, plasma, and liver was investigated utilizing HPLC. The concentration of Cn in brain and plasma, which was administered with CH-CLN, was higher than those that were treated with CLN. Also, CH-CLN increased the BBB permeability and bloodstream stability of Cn in comparison with CLN. The results con rm the e cacy of chitosan on the improvement of niosome, as the brain target delivery candidate.   The concentrations of CLN and CH-CLN in a different region of the central nervous system, liver, and plasma after 15 min administration (i.p. 50 mg/kg mean± SEM, n=4, *p<0.05 and **p<0.001).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.