reported in our recent work [12]. T

reported in our recent work [12]. The complex nanoparticles shared a high similarity with SA-CS and NaCas in their characteristic peaks including amide I and amide II absorptions at 1647 and 1530 cm-1, while the shifts of wavenumbers at these peaks in the region of amide bonds clearly indicated the electrostatic interactions between two biopolymers [36]. Likely, the O–H and C–H stretchings were observed in the complex nanoparticles with significant shifts in their wavenumbers, which evidenced that hydrophobic interactions also played a role in the formation of such complex. Noteworthy, the nanoparticles only showed one band at 1727 cm-1, which may be attributed to the overlapping of carbonyl stretching from ester bond in SA-CS and unreacted aldehyde groups in Odex. 3.3. Encapsulation of ASTX Due to its poor water-solubility, it is necessary to select an appropriate organic solvent to solubilize ASTX for encapsulation. According to previous literature, DMSO appears to be a good solvent to aid the encapsulation of ASTX in a liposomal formulation [37]. Thus, in the current study, DMSO was adopted to prepare ASTX stock solution, facilitating encapsulation within the core of the complex nanoparticles. Prior to preparation of nanoparticles, ASTX stock solution was mixed with NaCas which has the high affinity to lipophilic bioactive compounds. The proposed scheme of nanoparticles preparation and ASTX encapsulation is presented in Fig. 3. Three batches of nanoparticles with different ASTX loading ratios were tested, i.e. 1%, 3%, and 5%. As depicted in Fig. 4A, compared to the empty nanoparticles, the particle size of ASTXloaded nanoparticles was in a range of 145-198 nm. This may be explained by the expansion of nanoparticles due to ASTX loading into the NaCas hydrophobic core. The same trend was observed from PDI results, indicating that the nanoparticles became more heterogeneous with increases in ASTX loading concentrations. Encapsulation of ASTX did not alter the surface charge of nanoparticles, as no significant change of zeta potential was detected among three loading ratios (Fig. 4B). Nevertheless, the amount of loaded ASTX exhibited negative impact on its encapsulation efficiency by the nanoparticles. With the increase of loaded ASTX from 1% to 5%, the encapsulation efficiency decreased from 81.5% to 65.3%, indicating that increasing ASTX loading gradually compromised the encapsulation capability of nanoparticles. In addition, the cytotoxicity of nanoparticles with high ASTX loading ratio may increase as more DMSO was introduced into the composition. Due to these important factors, the ASTX-loaded nanoparticles with 3% loading ratio was selected for subsequent study. The encapsulation efficiency of asprepared nanoparticles was 70.2%, which equaled to 72.1 μM (84 μg/mL) of ASTX in the SACS/ NaCas/Odex complex nanoparticles. The loading ratio of ASTX in current study is higher than that in DNA/chitosan colloidal system (65 μg/mL) in literature [25]. 3.4. Morphological observationThe complex nanoparticles were negatively stained with uranyl acetate and observed under TEM. As shown in Fig. 5A, the dimension of freshly prepared nanoparticles was in the range of 80-100 nm with a spherical shape and smooth surface, suggesting the monodispersity and wellcontrolled particle size. In order to further validate the stability of nanoparticles in simulated GI fluids, the morphological observation using TEM was carried out after the nanoparticles were incubated in SGF (pH 2), and SIF (pH 7) individually. Unlike the rough and eroded surface of nanoparticles cross-linked by glutaraldehyde/Odex observed in previous study [12], the particle size, smooth surface and monodispersity of as-prepared nanoparticles were maintained after the incubation in SGF (Fig. 5B). This result revealed better stabilization effect by SA-CS/Odex coating layer formed via Schiff base reaction than that of glutaraldehyde in harsh acidic environment. Besides, the swelling and expansion of particle size as well as heterogeneous distribution of nanoparticles were not observed after the incubation in SIF (Fig. 5C), indicating that