br Fig A Hemolysis study of free mangostin
Fig. 5. (A) Hemolysis study of free α-mangostin, α-mangostin loaded FNPs, and blank FNPs in sheep red blood cells; and (B) cytotoxicity of free α-mangostin and α-mangostin loaded FNPs in Caco-2 and MCF-7 cells. Significant diﬀerences are noted between the free drug and formulas (*, p < 0.01); between PEI-FNP and other formulas (**, p < 0.01) (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.).
Fig. 6. Apoptosis assay of free α-mangostin, α-mangostin loaded FNPs, and blank FNPs in Caco-2 and MCF-7 cells.
(VITTDSDGNE) and (NINDFDED) located near the N-terminal of the fibroin heavy chains could bind specifically to the cell surface receptor integrin , which is overexpressed in many cancer Ruxolitinib (INCB018424) including colon and breast cancers . Thus, the strong binding could induce endocytosis process. Additionally, fibroin might adhere non-specifically to the cell membrane by electrostatic interactions between the nega-tively charged surfaces and the positively charged amine residues such as arginine and lysine . Overall, by incorporating α-mangostin in our crosslinked FNPs, we could not only decrease the drug side eﬀects by controlling the release rate, but also increase the drug eﬃcacy by the avoidance of cancer multi-drug resistance mechanisms of altered sur-face drug receptor and drug eﬄux pumps (i.e., P-glycoprotein). In Caco-2 cells, the free α-mangostin had an IC50 of 18.38 ± 0.93 μg/mL, which was in agreement with previous studies [40,41]. There was no significant diﬀerence in IC50 between EDClow-FNP and EDChigh-FNP. However, PEI-FNP possessed better cytotoxicity than both EDC-FNP formulations (p < 0.05). This phenomenon could be explained by PEI-FNP was endocytosed by Caco-2 cells much better than other formulations (unpublished data).
In MCF-7 cells, the IC50 of α-mangostin in the literature ranged from 1.6 μg/mL to 10.2 μg/mL [40–42]. Our result demonstrated an IC50 of 11.1 μg/mL. The reason behind this diversity could be related to the cell origin, passage, and the supply source of α-mangostin. All α-mangostin loaded FNPs demonstrated more than 2-fold lower IC50 than the free α-mangostin. In constrast to the Caco-2 cell line, no significant diﬀerence in IC50 between FNPs formulas was found on MCF-7 cells. This might be attributed to the diﬀerent uptake mechanisms and the degree of particle uptake between two cell lines.
3.10. DNA gel electrophoresis
The presence of DNA fragmentation ladder, which is observable under agarose gel electrophoresis, demonstrates the apoptotic cell death event after exposure to the test substances. The purpose of this experiment was to confirm the apoptotic eﬀect of α-mangostin in α-mangostin loaded FNPs. Clearly, the results were consistent with the cytotoxicity test in both cell lines. All blank particles showed no toxi-city, whereas the α-mangostin loaded FNPs demonstrated a greater DNA fragmentation ability than the free drug (Fig. 6).
In Caco-2 cells, the DNA patterns of EDClow-FNP and EDChigh-FNP were similar, however, the PEI-FNP pattern was diﬀerence than others, with lighter DNA fragments. This corresponded to the low IC50 of PEI-FNP, discussed in the in vitro cytotoxicity test section. The reason behind the absence of DNA ladder in 1-day incubation was due to fewer presented apoptotic cells, as apoptosis is a time-dependent process. Similarly, in MCF-7 cells, all formulations demonstrated clearer DNA apoptotic bands than the free drug α-mangostin, in agreement to the
low IC50 of the drug loaded FNPs. No significant diﬀerence between three formulations was noted, as they all possessed similar IC50.
In this study, we were successful in the development, character-ization, and in vitro toxicity evaluation of novel α-mangostin loaded crosslinked FNPs, using EDC or PEI, for cancer treatment. All formulas showed spherical particles with a mean size of 300 nm, a controllable zeta potential from −15 mV to +30 mV, and a higher drug EE (70%) and DL (up to 7%) than the non-crosslinked FNP. α-Mangostin was entrapped in molecular dispersion forms and diabetes mellitus reduced the crystallinity of the FNPs significantly. The crosslinked FNPs increased the drug so-lubility up to 3 times, sustained its release rate controllably to more than 3 days, and reduced its hematotoxicity by more than 90%. In the Caco-2 and MCF-7 cells, all formulas exhibited better cytotoxicity profiles than the free drug, while maintaining its apoptotic eﬀect. These particles were also physicochemically stable in intravenous diluent up to 24 h and in storage condition at 4 °C for at least 6 months. In con-clusion, the α-mangostin loaded crosslinked FNPs showed much po-tential for cancer chemotherapy. The versatility of our novel cross-linked FNPs are not only limited in cancer treatment, but also in other medical applications.