Hybrid Nanobeads for Oral Indomethacin Delivery

The oral administration of the anti-inflammatory indomethacin (INDO) causes severe gastrointestinal side effects, which are intensified in chronic inflammatory conditions when a continuous treatment is mandatory. The development of hybrid delivery systems associates the benefits of different (nano) carriers in a single system, designed to improve the efficacy and/or minimize the toxicity of drugs. This work describes the preparation of hybrid nanobeads composed of nanostructured lipid carriers (NLC) loading INDO (2%; w/v) and chitosan, coated by xanthan. NLC formulations were monitored in a long-term stability study (25 °C). After one year, they showed suitable physicochemical properties (size < 250 nm, polydispersity < 0.2, zeta potential of −30 mV and spherical morphology) and an INDO encapsulation efficiency of 99%. The hybrid (lipid-biopolymers) nanobeads exhibited excellent compatibility between the biomaterials, as revealed by structural and thermodynamic properties, monodisperse size distribution, desirable in vitro water uptake and prolonged in vitro INDO release (26 h). The in vivo safety of hybrid nanobeads was confirmed by the chicken embryo (CE) toxicity test, considering the embryos viability, weights of CE and annexes and changes in the biochemical markers. The results point out a safe gastro-resistant pharmaceutical form for further efficacy assays.


Introduction
Indomethacin (INDO) is a non-steroidal anti-inflammatory (NSAID) agent widely used for pain, fever and inflammation control through oral administration [1]. However, severe side effects in the gastrointestinal system (e.g., nausea, indigestion, vomit, diarrhea and abdominal ache) are associated to the oral administration, due to the first-pass metabolism. These effects are exacerbated in chronic diseases, in which an extended treatment is required [2]. Considering the huge number of INDO prescriptions (>1 million/per year, only in the USA), new approaches to minimize INDO systemic toxicity and improve its efficacy are required [3].

Preparation of NLC Formulations Loading Indomethacin
NLC (as control formulation) and NLC-INDO (2%, w/v) were prepared by the ultrasonication emulsification method [17]. The composition and concentration of the excipients used in NLC are displayed in Table 1: myristyl myristate (MM) as the solid lipid, coconut oil (CO) as the liquid lipid and poloxamer as the surfactant. The total lipid (TL) concentrationthe sum of solid and liquid lipid concentrations-was kept in 10% (w/v). Briefly, the lipid phase (containing MM and CO), with or without INDO (2%), was heated 10 • C above the solid lipid melting point (~48 • C) under magnetic stirring [18]. The aqueous phase composed of P68 (2-5%) was heated at the same temperature and dropped into the lipid phase under high-speed homogenization (10,000 rpm) for 3 min in an Ultra-Turrax blender (IKA Werke, Staufen, Germany). The resultant microemulsion was tip sonicated (Vibra-Cell, Sonics & Materials Inc., Newtown, CT, USA) at 500 W and 20 kHz, in 30 s (on/off) cycles for 20 min and further cooled to 25 • C in an ice bath.

Structural Characterization
The structural characterization of NLC and NLC/INDO (2%) was performed by dynamic light scattering (DLS), in terms of particles size (nm) and polydispersity index (PDI). Zeta potential (mV) values were obtained by electrophoretic light scattering in the same Zetasizer Nano ZS90 (Malvern Instruments, Worcestershire, UK) equipment. The pH values of NLC were determined with a (Tecnal, Uberlândia, Brazil) pH meter. The results were expressed as means ± standard deviations (n = 3).

Long-Term Physicochemical Stability
A long-term stability study of NLC and NLC/INDO (2%), stored in Falcon tubes and maintained at room temperature for one year (25 • C), monitored the particle size (nm), PDI, Zeta potential (mV) and pH values, in triplicates. ANOVA/ Tukey tests were employed to elucidate intragroup statistically differences over time (α = 0.05).

Transmission Electron Microscopy
The morphological analyses of NLC and NLC/INDO (2%) samples were performed by transmission electron microscopy (TEM). The samples were diluted 50×, dispersed onto copper grids coated with a carbon film and dried at 25 • C. Uranyl acetate (2%) was added to provide contrast. After 24 h, micrographs of the samples were obtained using a JEOL 1200 EXII (Jeol, Peabody, MA, USA) microscope, at 60 kV.

Determination of Indomethacin Encapsulation Efficiency
INDO calibration curve by UV-vis at 210 nm was performed on five different concentrations in the range of 3.25-20 µg/mL (r 2 > 0.99). Each point represents the average of 9 measurements performed on 3 different days ( Figure S1 (Supplementary Materials)). INDO encapsulation efficiency (%EE) was quantified by the ultrafiltration-centrifugation method [9], using regenerated cellulose filters with a molecular exclusion pore size of 10 kDa (Millipore, Burlington/MA, EUA). The samples (triplicates) were diluted and centrifuged at 4000× g for 20 min. The total amount (INDO total ) and the non-encapsulated indomethacin (INDO free ) present in the filtered fraction were quantified by UV-vis (Waters, Agilent Technologies, Milford, MA, USA) at 210 nm [19]  de Janeiro cells bank, Federal University of Rio de Janeiro, Brazil. Briefly, the cells were seeded in 96-well culture plates at a density 10 4 cells/mL and incubated for 24 h at 37 • C under 5% CO 2 . The RPMI culture medium was then discarded and replaced with 100 µL of fresh medium containing increasing concentrations of INDO or NLC/INDO. After 24 h, the medium was removed, and the plate was washed with phosphate-buffered saline, at pH 7.4. After that, 100 µL of medium (without serum) containing 0.5 mg/mL of MTT supplemented each well, which was then incubated for 3 h at 37 • C. The MTT solution was then removed, and 100 µL of ethanol was added to solubilize the formed formazan crystals. The formazan absorbance was measured in a microplate reader, at 570 nm [20]. The obtained absorbance was normalized to % of control. The results were provided as means ± standard deviation (SD), n = 9. The statistical analyses were performed by a t test (α = 0.05).

Preparation of Biopolymers and Nanobeads Loading Indomethacin
Chitosan (CHT) was dispersed in 50 mL of 0.1% acetic acid, and xanthan (XAN) was dissolved in 50 mL of deionized water under magnetic stirring until complete homogenization, with a final concentration of 2% and 0.5% (w/v), respectively. Then, 2% hydroalcoholic INDO solution was added to the CHT solution and stirred for 2 h at room temperature (CHT/INDO). For the preparation of hybrid nanobeads, NLC loading 2% INDO was used to replace half of the acid solution for CHT solubilization, and the samples were stirred for  Table 2 displays the composition of the resultant beads and nanobeads loading INDO (2%).

FE-SEM Analyses of Beads and Nanobeads
The surface analysis of beads and nanobeads was carried out in a field emission scanning electron microscopy (FE-SEM) (FEI-NOVA NanoSEM 230, Sydney, Australia). XAN@CHT/NLC-INDO nanobeads were firstly frozen in liquid nitrogen and cross-sectioned with a scalpel. The samples were fixed on a carbon tape and subjected to a gold conducive coating on the surface.

Differential Scanning Calorimetry
Differential scanning calorimetry (DSC) analyses were carried out in a TA Q20 calorimeter (TA Instruments, New Castle, DE, USA) equipped with a cooling system of INDO, NLC excipients and nanobeads (XAN@CHT/NLC-INDO). The samples (5 mg) were introduced in aluminum pans and the thermograms registered from 25 to 180 • C, at 10 • C/min heating rate, under N 2 flow.

In Vitro Water Uptake of Beads and Nanobeads
The swelling properties of beads (CHT/INDO and XAN@CHT/INDO) and nanobeads (CHT/NLC-INDO and XAN@CHT/NLC-INDO) were determined. The samples (0.05 g) were placed in Petri dishes and immersed in 10 mM phosphate solution at pH 1.2 or phosphate buffer at pH 6.8, being regularly shaken, at 25 • C. At predetermined times, the swelled beads and nanobeads were carefully withdrawn, the excess water was removed and the samples were weighed on an analytical balance [21] in triplicate for 6 h. The water uptake (g H 2 O/g sample) was calculated from Equation (2): where W 2 and W 1 are the swelled and initial mass (0.05 g) of beads, respectively.

In Vitro Indomethacin Release Test
The beads and nanobeads (0.1 g) were immersed in 50 mL of the release medium in a thermostatic bath at 37 • C and under agitation (350 rpm). Considering the oral administration purpose, the samples were first maintained for 2 h in a 0.06 M HCl release medium (pH 1.2) to simulate conditions of the gastric tract. Then, they were kept at pH 6.8 (by adding 0.03 g NaOH and 0.40 g NaH 2 PO 4 ·H 2 O), to mimic the intestinal microenvironment, until the end of the test (26 h). At predetermined time intervals, 1 mL of each solution was withdrawn, and INDO was quantified by UV-vis spectrophotometer (λ = 210 nm) [19]. The measured samples were returned to the release medium to keep the volume constant. All the experiments were carried out in quintuplicate. Data were given as the mean ± standard deviation of INDO cumulative release percentages.
where Q = INDO amount released at the time t, k = rate constant and Q 0 = initial INDO release.
where m = INDO concentration released at the time t, b gives the release exponent and a is the time scale of release.

In Vivo Toxicity Assays through Chicken Embryo Model
The in vivo toxicity of the INDO solution as the positive control (PC), the 0.9% NaCl solution as the negative control (NC) and the XAN@CHT/INDO (20 mg INDO/egg) bead and XAN@CHT/NLC-INDO (20 mg INDO/egg) nanobead was evaluated in a chicken embryo (CE) model through different parameters-embryos viability, CE weight and annexes changes-and blood biochemical markers. Eggs with 10 days of incubation (EID) were used. Then, the chorioallantoic membrane (CAM) was lowered, and the samples were administered in CAM ( Figure S2) and followed for 7 days. After 24 h, the eventual CE deaths caused by the mechanical stress of CAM removal were excluded from the analysis.

Preparation of Eggs
The eggs of laying hens (Gallus gallus, Hy-Line W-36 lineage) were a gift of Hy-Line do Brazil (Uberlândia, Brazil). The eggs at 10 days of incubation (10 EID) were first carried out to a light ovoscopy, to confirm the successful embryonic development. Then, 48 eggs were weighted and incubated in an artificial incubator (Premium Ecológica ® ) at 37 • C and 58% relative humidity, being turned at a 2-hour interval up to subsequent analyses.

CE Viability Test
For the embryo mortality check, the eggshell was fragmented at 17 EID. The CE viability (%) was determined by the percentage of alive embryos with or without injuries, after 24 h of testing in comparison with the initial number of alive embryos (n = 12). The Chi-square test followed by binomial test between two proportions were used to elucidate intergroup significant differences (α = 0.05).

Changes in the Weights of CE and Annexes
At 17 EID, the CEs were weighed, and their annexes (CAM, egg yolk and amnion) were removed and also weighed (n = 12). The one-way ANOVA and Tukey post hoc tests were applied to elucidate intergroup statistically significant differences (p < 0.05). The weight of CE was adjusted based on the egg weight (EID = 10) according to the Equation (6): where aW is the CE weight adjusted to 50 g, final W is the CE weight at 17 EID and initial W is the weight of embryonic egg at 10 EID.

Biochemical Markers
At 17 EID, the umbilical vessel blood from the CE was collected for biochemical analyses and analyzed in an automatic biochemical analyzer (ChemWell ® 2910, Awareness Technology, Palm, FL, USA). The analytes-uric acid (UA), alkaline phosphatase (ALP), creatine kinase (CK), gamma glutamyl transferase (GGT), alanine aminotransferase (ALT) and aspartate aminotransferase (AST) (Bioclin ® , Minas Gerais, Brazil)-were quantified from the CE serum (n = 12). Data were expressed as median ± SEM. The one-way ANOVA/Tukey tests were performed for the analyses of intergroup differences of each marker.

Preparation of Nanostructured Lipid Carriers Loading Indomethacin
Ten different NLC formulations (Table 1) composed of coconut oil (CO), myristyl myristate (MM) and poloxamer (P68), with and without INDO (2%), were successfully prepared, showing a homogenous aspect and pale white color.

Long-Term Stability of NLC
The shelf life of the NLC formulations with or whitout 2% INDO were assessed through size (nm), polydispersity index (PDI), Zeta potential (mV), pH and visual inspection monitoring for 1 year at 25 • C ( Figure 1). Visually, all the obtained colloid-like formulations were homogenous and white, with no phase separation during one year. As determined by DLS, the particle sizes of NLC formulations were smaller than 300 nm ( Figure 1A). It is noteworthy that the F5 and F5-INDO formulations showed no statistically significant differences over time (p > 0.05). All systems showed desirable particle sizes after 1 year of storage. According to Figure 1B, all the prepared formulations exhibited PDI values lower than 0.2, throughout the stability monitoring. Zeta potential values of formulations were in the range of −20 to −40 mV for the control (NLC) and NLC/INDO ( Figure 1C) formulations, respectively. F5 and F5-INDO formulations showed no statistically significant differences (p > 0.05) in terms of Zeta values over 365 days. Finally, the pH values ranged from 3.0 to 3.5 for all samples ( Figure 1D) and remained constant (p > 0.05) over the time.

Indomethacin Encapsulation Efficiency by NLC
INDO encapsulation efficiency by NLC was determined by the ultrafiltrationcentrifugation method. Table 3 shows excellent (%EE very close to 100%) in all formulations tested. Table 3. Encapsulation efficiency from different NLC formulations loading indomethacin (2%), n = 3. Considering the absence of significant changes in the long-term stability and the %EE results, F5-INDO was chosen to be evaluated in further tests. Figure 2 shows the spherical morphology F5/INDO and F5 (as control) formulations, with well-defined boundaries. The particle sizes calculated from the micrographs using the ImageJ software agreed with those quantified by DLS (<300 nm).

F3
- Considering the absence of significant changes in the long-term stability and the %EE results, F5-INDO was chosen to be evaluated in further tests. Figure 2 shows the spherical morphology F5/INDO and F5 (as control) formulations, with well-defined boundaries. The particle sizes calculated from the micrographs using the ImageJ software agreed with those quantified by DLS (<300 nm).

Cell Viability Tests
The in vitro viability test (

Cell Viability Tests
The in vitro viability test (Figure 3  Considering all these results, NLC composed of 10% TL (MM = 80% and CO = 20%), 5% P68 and 2% INDO were selected as the best formulation, called from now on NLC-INDO, the lipid excipient of hybrid lipid-biopolymer nanobeads.

Preparation of Biopolymers and Hybrid Nanobeads Loading Indomethacin
The biopolymer beads (CHT/INDO and XAN@CHT/INDO) and hybrid nanobeads (CHT/NLC-INDO and XAN@CHT/NLC-INDO) were successfully prepared. Digital photos of all the samples (data not shown) were taken with at least 150 beads that were spatially dispersed. Then, the frequency of particle sizes from the images were estimated by ImageJ software, as exemplified in Figure 4, for XAN@CHT/NLC-INDO sample. Table 4 reveals that the particle sizes of beads were lower than 0.4 cm, with a monodisperse distribution (PDI < 0.130) for all the prepared samples.     Figure 5B) images showed expected profiles of the flake-like chitosan structure [21] and rough surface of biopolymer-lipid assembling [27], respectively. The top-of-view of CHT/NLC-INDO ( Figure 5C) shows the spherical and uniform morphology of the nanobead. Finally, the cross-section of XAN@CHT/NLC-INDO ( Figure 5D) nanobead proves the successful coating of XAN, with a well-delimited external layer with around 192 µ m of thickness. There was no evidence of dispersed drug in the biopolymer matrices in any of the images.

DSC Analyses
The DSC thermodynamic profiles of the excipients and nanobeads are shown in   (Figure 7). It was performed at two different pH levels, simulating the gastric (pH 1.2) and intestinal (pH 6.8) microenvironments [23].
In acid medium, CHT/INDO and XAN@CHT/INDO beads swelled to 0.16 and 0.026 g, respectively, and disrupted after 15 min of experiment. On the other hand, the hybrid CHT/NLC-INDO and XAN@CHT/NLC-INDO nanobeads did not disrupt, and swelled to 0.21 and 0.042 g, respectively, after 360 min. The water uptake in the simulated intestinal system (pH 6.8) evidenced that all beads were visually intact after 6 h (see the digital photo of all beads and nanobeads, before and after the experiment, in Figure 7). Moreover, the final water uptake of CHT/INDO, XAN@CHT/INDO, NLC-INDO and XAN@CHT/NLC-INDO was 0.035, 0.033, 0.027 and 0.022 g, respectively.

In Vitro Indomethacin Release Test
The in vitro INDO release test was conducted for 26 h at 37 °C (Figure 8). In the first 2 h, the experiment simulated the gastric system (pH 1.2). Then, the pH of the external

In Vitro Indomethacin Release Test
The in vitro INDO release test was conducted for 26 h at 37 • C (Figure 8). In the first 2 h, the experiment simulated the gastric system (pH 1.2). Then, the pH of the external medium was changed to 6.8, mimicking the intestinal system [21] until the end of the analysis. Mathematical modeling of the kinetics was carried out by KinetD software [22]. Different models were tested and considering the highest coefficient of determination (R 2 ; Table 5), the beads (CHT/INDO and XAN@CHT/INDO) were best fitted by the zero-order model, while CHT/NLC-INDO and XAN@CHT/NLC-INDO hybrid nanobeads were best fitted by the Baker-Lonsdale and Weibull models, respectively.

In Vivo Toxicity Assay through Chicken Embryo Model
The alternative in vivo toxicity assay through CE model [31] was used to evaluate the toxicity of INDO (in solution) in comparison with the coated bead (XAN@CHT/INDO) and nanobead (XAN@CHT/NLC-INDO). Figure 9 shows the CE viability (%) after treatment with the different formulations. Since all CE treated with INDO (as positive control) died during the experiment, the viability percentages of other groups were statistically different (p < 0.05). On the other hand, there was no significant differences (p > 0.05) in the viability of CE treated with beads and nanobeads, in comparison to the negative control, NC ( Figure 9A). The percentages of the CE viability for beads and nanobeads were 60% and 80%, respectively. Among the CE survivors, approximately 30% treated with beads and 40% with nanobeads induced injury to the embryos ( Figure 9B). The weights of CE and annexes treated with NC, beads and nanobeads containing INDO (20 mg/egg) are given in Figure 10. In terms of CE weight, it was noticed values of 24.53, 23.31 and 21.71 g when treated with NC, beads and nanobeads, respectively. On the other hand, the weights of the annexes were 13.05, 12.82 and 12.68 g for NC, beads and nanobeads, respectively. There were no statistically significant differences among the CE (p > 0.05) and annexes (p > 0.05) weights in comparison with NC.  Table 6 displays the values of all analytes of CE treated with NC, beads and nanobeads. Among all the enzyme activities, only the ALP values were statistically significant different (higher) to the NC (p < 0.05). For all the other markers, no statistically significant differences were observed. Table 6. Biochemical markers indexes in the CE serum treated with different groups. Data are provided as median ± SEM, n = 12. One-way ANOVA/Tukey tests were performed for the analyses of intergroup differences among each marker. a,b = (p < 0.05).

Discussion
The oral route is the major choice of drug administration [32]. After absorption, the actives are subjected to hepatic first-pass metabolism, which can result in low efficacy and side effects, mainly to the gastrointestinal system [33]. That is the case of INDO, a NSAID largely prescribed for the management of chronic pain and inflammatory conditions. Side effects are even more accentuated in prolonged therapies, decreasing the patient compliance [2].
The current work was divided in two parts: the development and characterization of nanostructured lipid carriers (NLC-INDO), followed by the preparation and evaluation of beads (XAN@CHT/INDO) and nanobeads (XAN@CHT/NLC-INDO).
Different NLC formulations (10) exhibited excellent in vitro structural properties, reinforcing such systems as promising DDS. INDO encapsulation by NLC caused slight changes in the biophysical parameters, preserving the morphology, size homogeneity and nanoparticles steric repulsion over time, as required for a long-term stability. In fact, such formulations showed lower pH values than control (without drug), given to the INDO acid character [24] and high amount of INDO upload by NLC/INDO (close to 100%), as expected for hydrophobic NSAID [6,34].
F5-INDO formulation confirmed the decrease in the INDO cytotoxicity in comparison with free INDO treatment. Unfortunately, the oral administration of colloid systems is contraindicated, given their poor gastro-resistance. It is worth mentioning that the prepared NLC-INDO can be further explored as DDS aiming other administration routes, such as parenteral, intranasal and/or transcorneal, for the sustained delivery of INDO. Therefore, F5-INDO was selected as the lipid component (called as NLC/INDO) in the nanobeads development.
Lipid-polymer assembling is a remarkable combination of organic matrices resulting in hybrid DDS with extraordinary properties and complex supramolecular arrangements [35][36][37]. In here, the first step in the development of the lipid-biopolymer nanobead consisted in the CHT/NLC-INDO preparation. Such a system was intended to combine the advantages of INDO nanoencapsulation by NLC with the deeply explored mucoadhesion property of cationic CHT [13]. The challenge of the second step was to protect the resultant hybrid nanobead against the acidic stomach pH, once both excipients-CHT and NLC/INDO-were not gastro-resistant (a requirement to avoid fast absorption and first-pass metabolism). In this sense, the enteric coating of beads, microspheres, tablets and colloids to provide gastro-resistance to the systems is not a novelty. Different coatings based on pectin, alginate, carboxymethylcellulose, eudragit, gelatin, gellan and xanthan gums, among others, avoided drug release in simulated stomach condition successfully [10,21,38,39]. In here, XAN, which is a rheological improver [15], coated CHT/NLC- From the structural point of view, the particle size distribution of the nanobeads (Figure 4) did not increase in relation to pure biopolymer forms (Table 3). In addition, FE-SEM images successfully confirmed the XAN coating procedure, with a well-delimited and homogenous external layer ( Figure 5D), acting as physical barrier to protect CHT and NLC-INDO to the degradation. Moreover, calorimetric analyses suggested good compatibility between the lipid-biopolymer blend and INDO, since the thermodynamic properties of XAN@CHT/NLC-INDO are very similar to that of NLC-INDO, the major component of nanobeads [15,40]. Still, no degradation peaks were observed in the thermograms, ensuring the thermal stability of the system up to 180 • C (Figure 6), far above the physio/pathological temperature range (32.5-41.0 • C).
The hybridization and coating procedures were responsible to modulate the in vitro performance of nanobeads, regarding water uptake ( Figure 6 Despite that XAN@CHT/NLC-INDO prolonged INDO release up to 26 h, the desirable burst release effect (~30%) was also noticed in the first two hours of test. This is due to the amount of INDO directly incorporated in the CHT matrix (to reach the drug final concentration of 2%), allowing for its faster release. In general, a burst effect followed by a prolonged release profile are essential for DDS aiming pain and inflammation management [42]. Specifically, XAN exerted a physical protection to the nanobead (XAN@CHT/NLC-INDO) and bead (XAN@CHT/INDO) that decreased CHT and NLC/INDO solubilization in acid in comparison with their respective non-coated forms. Still, XAN@CHT/NLC-INDO, which combined the hybridization and coating processes, exhibited the lowest water uptake and a higher INDO sustained release profile among all the samples. This inverse relation between swelling and release profiles is expected for hybrid polymer-based forms, once a higher resistance to water adsorption is correlated with a more sustained profile of entrapped drugs [21,23]. The prolonged INDO delivery was explained by the two physical barriers that INDO had to overcome, the lipid structural matrix and the swelled biopolymers chains of XAN@CHT/NLC-INDO [15,27]. In fact, the differences of XAN@CHT/INDO (as bead) and XAN@CHT/NLC-INDO (as nanobead) swelling capacity were also observed in the vivo assays, where the hybrid nanobead was intact after 7 days of CE administration, while the beads were totally solubilized at this time ( Figure S2).
It is also interesting to notice that the mathematical modeling indicated that the novel supramolecular arrangement of nanobeads was also responsible for different INDO kinetics mechanisms. Both biopolymer beads were best fitted by the zero-order model, as expected, considering their total INDO release up to 2-3 h of the test. On the other hand, CHT/NLC-INDO was best fitted by the Baker and Lonsdale model, which describes a sustained diffusion of drug from a spherical matrix [43]. Besides, XAN@CHT/NLC-INDO was best fitted by Weibull model, indicating a biphasic profile from complexes matrices, given by the release exponent values (b < 1). The nanobead combined a burst effect with prolonged release INDO profile. In fact, the Weibull model is widely employed to explain the complexes lipid-polymer DDS kinetics mechanisms [15,40,41,44].
Finally, the safety of DDS is another mandatory requisite, especially in case of nanomaterials that exhibit novel kinetics and supramolecular arrangements, to detect possible risk factors to health [45]. In this sense, CE is an alternative toxicity model that is very promising to evaluate DDS, once different parameters (microscopical, microscopical and biochemical) are evaluated in a single model [31]. In here, the CE viability test showed that INDO solution (PC) killed all embryos after 24 h, which was not observed for bead and nanobead ( Figure 9A) treatments. Moreover, the indexes of surviving embryos of NC did not show statistically significant differences (p > 0.05) compared to the groups treated with beads or nanobeads. In relation to the changes in the CE and annexes weight, significant differences among the groups and NC were not detected (p > 0.05).
Furthermore, the analysis of biochemical parameters provides essential information for the evaluation of animal global health status. All the metabolites of CE treated with bead and nanobead did not significantly change in comparison with NC (p > 0.05), except for the ALP values of CE treated with beads (p < 0.05). In fact, there is some evidence that INDO can induce an increase in ALP in rats, indicating its ability to induce toxicity to the liver [46]. It is worth mentioning that the CE treated with beads, where INDO was only dispersed in CHT matrix, resulted in a liver injury, given by ALP increase and macroscopic changes in liver. Interestingly, the treatment with nanobeads showed less ability to damage the CE liver than beads. This is probably due to the different lipid-biopolymer barriers and the nanoencapsulation that decreased INDO toxicity in CE together with the biopolymer protective effect. Such evidence suggests that the nanobead was the least toxic treatment for CE, not causing significant deaths, macroscopic damages or biochemical markers alterations to CE.
These results strongly suggested that XAN@CHT/NLC-INDO was effective to prolong the INDO release profile, exhibited gastro-resistant properties and decreased the drug toxicity, as claimed. Such a system can be a promising candidate as an INDO oral delivery system, and should be further tested in efficacy assays.

Conclusions
This work described the preparation of elegant hybrid nanobeads composed of nanostructured lipid carriers loading indomethacin (NLC/INDO; 2%) and chitosan (CHT; 2%), coated by xanthan (XAN; 0.5%). Such systems presented excellent structural and thermodynamic properties, in vitro swelling properties and a prolonged drug release profile up to 26 h. The synergism between the nanolipid and biopolymer excipients allowed to prevent the burst release effect in the first 2 h of experiments, simulating the gastric medium, followed by a prolonged release in the pH 6.8, mimicking the intestinal medium. The XAN coating of nanobeads (XAN@CHT/NLC-INDO) acted as a gastro-resistant excipient protecting NLC/INDO and CHT from an abrupt swelling and consequent total drug release in the first hours of experiments. The in vivo toxicity assay in chicken embryos (CE) confirmed the nanobeads safety, which combined the advantages of the hybridization and coating procedures, decreasing the intrinsic toxicity of INDO in CE in all the analyzed parameters. Overall, the lipid-biopolymer nanobead proposed here is an excellent candidate for the delivery of INDO, as well as other anti-inflammatories through the oral route.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/pharmaceutics14030583/s1, Figure S1: Indomethacin calibration curve (detection at 210 nm) performed on five different concentrations in the range of 3.25-20 µg/mL. Correlation coefficient was >0.99. Each point represents the average of nine measurements performed in three different days; Figure S2: Digital photos of the chicken embryos treated with beads (above) and nanobeads (below) in different times of incubation.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.