Screening of multidrug resistance genes and selecting siRNA target sequences in ESCC cells
The KYSE-510 cell line is an esophageal squamous cell carcinoma and is widely used in the study of the pathogenesis of ESCC. To screen for potential MDR genes, we established an Adriamycin-resistance cell line, named 510K, from KYSE510 cells. The cell viability results indicated that the 510K cells developed a stronger drug resistance than 510 cells (Fig. 1A, B). To discover the potential MDR mechanism of ESCC, we selected five extensively studied genes related to MDR as screening targets, glutathione S-transferase π, GST-π; P-glycoprotein, P-GP; Multidrug resistance-associated protein, MRP; MVP and BCL2 [31, 32]. The results of qPCR and Western blot showed that the expression levels of MVP gene and BCL2 gene in 510 k cells were significantly increased (Fig. 1C), which indicates that these two genes may play a vital role in the MDR of ESCC. Our findings are consistent with previous studies of these two genes in MDR of other tumors [33, 34]. According to MVP-mRNA and BCL2-mRNA gene sequences, we designed and verified siRNA that can effectively silence the MVP-mRNA (Fig. 1D, F) and BCL2-mRNA (Fig. 1E, G). In the subsequent experiments, we choose MVP-siRNA1 and BCL2-siRNA2 to silence MVP-mRNA and BCL2-mRNA, respectively.
Preparation of CEAMB NPs
After identifying the two potentially critical genes for MDR in ESCC and selecting the RNAi sequence, we tried to eliminate ESCC resistance by suppressing their expression. To achieve a better anti-tumor effect, we prepared a novel type of CEAMB NPs with tumor targeting and pH-responsive protonation by self-assembly using CHCE, Adriamycin, MVP-siRNA, and BCL2-siRNA (Fig. 2A). The CEAMB NPs co-deliver MVP-siRNA, and BCL2-siRNA to silence MVP-mRNA and BCL2-mRNA simultaneously, inhibiting drug efflux and anti-apoptosis of tumor cells and enhancing the anti-tumor effect of Adriamycin (Fig. 2B).
In synthesizing CEAMB NPs, we first synthesized the carboxymethyl chitosan polymer modified with EGFR monoclonal antibody and histidine cholesteryl ester (CHCE) (Additional file 1: Fig. S1) and next synthesized the CEAMB NPs from CHCE, Adriamycin, MVP-siRNA, and BCL2-siRNA through self-assembly (Fig. 3A). As an amine-containing hydrophilic polysaccharide, carboxymethyl chitosan is an ideal material with excellent biocompatibility, biodegradability, and low immunogenicity . As a component of the cell membrane, the cholesterol itself is highly biocompatible and may enhance the cellular uptake of the NPs . The imidazole group in the histidine located at the inner surface of the NPs renders the NPs with an ability to respond to pH changes and produce a proton sponge effect to facilitate cargo siRNA escape from the endosome . Studies by other researchers have found that EGFR is expressed at a higher level in ESCC compared with normal tissues, and its expression on the cell surface is a potential target molecule [38, 39]. Therefore, we coupled the designed nanoparticles with EGFR antibodies to achieve specific esophageal squamous cell carcinoma delivery.
Modification with the histidine-cholesteryl ester can form a hydrophobic domain, promoting the long-chain carboxymethyl chitosan polymer in the stretched state to curl and form encapsulated NPs [40, 41]. In addition, the lower critical micelle concentration (0.1 mg/mL) also indicated that the CHCE had a better ability to self-assemble into spheres and solid stability (Additional file 1: Fig. S2). In this experiment, we also tested the encapsulation efficiency of siRNA and Adriamycin. The results showed that the encapsulation efficiency of Adriamycin reached 95%, and the encapsulation efficiency of siRNA reached 85% (Additional file 1: Fig. S3A, B).
Characteristic of CEAMB NPs
The CEAMB NPs had a spherical structure and lower size distributions (Fig. 3B). The size distributions of CEAMB NPs in different pH environments (90.26, 100.2, and 129.1 nm) indicated that the particle size of CEAMB NPs had little change in different pH environments (Fig. 3C). The PDI also fully indicated that CEAMB NPs had better stability in different pH (Fig. 3C). Furthermore, the Zeta charges reversal of CEAMB NPs at different pH (-8.6 mv, 17 mv, and 30 mv) indicated that CEAMB NPs have sensitive pH-responsive protonation ability (Fig. 3D). Next, we measured the serum stability of CEAMB NPs in serum, the average size distribution at 0, 6, and 12 h were about 82, 218, and 450 nm, respectively, which indicated that the CEAMB NPs could swell and become larger in volume over time. However, the morphology images showed that almost all CEAMB NPs still maintained the spherical contour, indicating that CEAMB NPs had good stability and could protect siRNA from nuclease degradation within 12 h (Fig. 3E–G). We also tested the stability of CEAMB NPs in serum by gel blocking experiment. The results showed that as the time of CEAMB NPs in the serum increased, more siRNA was released, and the siRNA was almost wholly released at the 12th hour (Additional file 1: Fig. S5). According to the above results, the total dissociation time of CEAMB NPs in serum was about 12 h, which indicated that the CEAMB NPs had good stability and could protect siRNA from nuclease degradation within 12 h. Compared with siRNA and chemotherapeutics alone, the CEAMB NPs can effectively prolong the circulation time of siRNA and chemotherapeutics in vivo. The above experimental results proved that the CEAMB NPs had better physical and chemical properties: better stability, appropriate particle size, and sensitive pH-responsive protonation ability.
These excellent physical and chemical characteristics could endow CEAMB NPs with higher delivery efficiency. The better stability can protect the integrity of the biological structure of siRNA and prevent the dissociation and release of siRNA, causing severe side effects during the delivery process . Appropriate particle size can reduce the clearance efficiency of the kidney to the NPs and improve the ability of NPs to pass through the vascular endothelial space and dense extracellular matrix . Sensitive pH-responsive protonation can make the surface charge of the NPs with the change of environmental pH. In the blood circulation system (Ph7.4), the surface of the NPs has a low negative charge, which can prevent the protein opsonin from adhering to the NPs and prevent the NPs from being quickly cleared by the mononuclear macrophage system . In the acidic environment of tumor tissue (pH 6.5), the surface charge of the NPs is transformed into a positive charge, which promotes the cellular adhesion, uptake, and lysosomal escape of the NPs .
Cellular adhesion and uptake of CEAMB NPs
Studies have shown that the siRNA must first form an RNA complex in the cytoplasm before silencing specific target genes . Therefore, the NPs must have higher cellular adhesion and uptake efficiency to successfully deliver siRNA into the cytoplasm . In our design, the CHC/Adriamycin/MVP-siRNA/BCL2-siRNA NPs (CAMB NPs) were modified with EGFR monoclonal antibody, which not only endowed NPs with the ability to specifically recognize tumor cells but also triggered receptor-mediated endocytosis to promote cellular uptake efficiency of NPs . The adhesion efficiency showed that compared to pure siRNA (6.5 ± 5.3%), pure Adriamycin (6.7 ± 5.5%), and CAMB NPs (siRNA: 40.1 ± 6.7%, Adriamycin: 37.8 ± 6.5%), the CEAMB NPs (siRNA: 90.1 ± 7.6%, Adriamycin: 92.5 ± 6.9%) had the best cellular adhesion ability (Fig. 4A, B).
Next, we detected the cellular uptake of CEAMB NPs; the results showed that with the increase of the concentrations, the uptake efficiency of siRNA and Adriamycin also increased, and the increased efficiency of both siRNA and Adriamycin were consistent (Fig. 4C). The fluorescence images also showed that siRNA and Adriamycin were spatially consistent in tumor cells, which indicated that the NPs had good stability in cellular uptake and could ultimately deliver siRNA and Adriamycin into cells through cell membranes (Fig. 4D). Then we observed the fate of CEAMB NPs co-delivering siRNA and Adriamycin in cells. The fluorescence images showed that with the increase of time, the more siRNA and Adriamycin uptake by tumor cells (Fig. 5A). The colocalization rate of Adriamycin and siRNA in cells showed that Adriamycin and siRNA were tightly combined in the early stage (1 h: 90 ± 5%, 2 h: 85 ± 4%), and then gradually separated over time (3 h: 60 ± 5%, 4 h: 42 ± 4%) (Fig. 5B). The above results indicate that the CEAMB NPs had good stability and could effectively deliver siRNA and Adriamycin into the cell and release siRNA and Adriamycin into the cytoplasm.
Lysosomal escape of CEAMB NPs
As an important biological barrier in the process of NPs delivery, lysosomes play an important role in limiting the delivery efficiency of NPs . Therefore, the lysosomal escape ability of NPs determines its ultimate delivery efficiency. To enhance the lysosomal escape ability, we designed and endowed the CEAMB NPs with histidine cholesteryl ester, which could induce the protonation sponge effect of lysosomes, thus causing instability of lysosomal membrane or rupture of lysosomes . The colocalization rate of Adriamycin and lysosomes (0.5 h: 45.6%, 1 h: 73.7%, 2 h: 92.5%, 3 h: 68.5%, 4 h: 53.3% and 5 h: 45.6%) (Fig. 6A, D), siRNA and lysosomes (0.5 h: 42. 5%%, 1 h: 7.6%, 2 h: 91.6%, 3 h: 65.6%, 4 h: 50.6% and 5 h: 39.5%) (Fig. 6B, E) showed that the CEAMB NPs entered in lysosomes at early and reached the maximum at about 2 h, then gradually separated from lysosomes and the more CEAMB NPs separated from lysosomes over time. The results could powerfully prove that the CEAMB NPs had better lysosomal escape ability. In addition, we also tested the colocalization rate of Adriamycin and siRNA. The results were consistent with the intracellular delivery results of CEAMB NPs (Fig. 6C, F). Summarizing the above results, the CEAMB NPs had better cellular uptake and lysosomal escape ability and could release siRNA and Adriamycin into the cytoplasm.
CEAMB NPs used to effectively inhibit the expression of siRNA target genes
Because the MVP gene and BCL2 gene respectively increase the drug efflux and anti-apoptosis of tumor cells, our experimental results also show that these two genes are highly expressed in Adriamycin-resistant cells. Therefore, we adopted a double sensitization strategy of co-loading NPs with MVP-siRNA and BCL2-siRNA to simultaneously silence the MVP gene and the BCL2 gene and eliminate the drug resistance of 510 K cells. The qPCR results showed that the CAM NPs and CAB NPs could only silence MVP-mRNA and BCL2-mRNA, respectively, while the CEAMB NPs could silence MVP-mRNA and BCL2-mRNA simultaneously (Fig. 7A, B). Lowered protein expression indicated by results of western blot analysis further proved that CEAMB NPs could silence MVP-mRNA and BCL2-mRNA simultaneously (Fig. 7C, D). In addition, the above results indicated that the co-loading of Adriamycin did not affect multiple siRNAs to silence the target gene.
Next, we detected the anti-proliferative ability of CEAMB NPs. The 510K cells treated with the CEAM NPs and CEAB NPs had lower activity than those treated with the CEA NPs, while the cells treated with the CEAMB NPs had the lowest activity (Fig. 7E). The results indicated that silencing MVP-mRNA and BCL2-mRNA simultaneously could more effectively eliminate multidrug resistance and promote the anti-tumor effect of Adriamycin. The cytotoxicity results further proved that the CEAMB NPs had the most effective anti-tumor ability (Additional file 1: Fig. S7A–C). Researchers have demonstrated that Adriamycin resulted in cell cycle arrest in G0/G1 phase , so the cell cycles were tested to detect the anti-tumor ability of CEAMB NPs. The results showed that the CEAMB NPs had the highest proportion of cells in the G0-G1 phase (Fig. 7F, G). Also, our results showed that the CEAM NPs and the CEAB NPs could induce more tumor cells into apoptosis (44.2%, 48.4%) than the CEA NPs (30.1%), while the CEAMB NPs could induce the most tumor cells to undergo apoptosis (66.2%), these findings are in line with previous research on the function of Adriamycin (Fig. 7H) . The results indicated that the CEAMB NPs might eliminate drug efflux and anti-apoptotic ability of MDR 510K cells to improve the anti-tumor effect of Adriamycin, leading the most tumor cells to arrest in the G0/G1 phase and into apoptosis.
The targeting delivery of CEAMB NPs in vivo
To improve targeted delivery ability in vivo, the CAMB NPs were modified with EGFR monoclonal antibody. The fluorescence reflectance images showed that the CEAMB NPs had higher Adriamycin and siRNA accumulation concentrations in the tumor site (Fig. 8A). The organ and tumor images showed that the use of only Adriamycin and siRNA had a higher accumulation concentration in the liver and kidney and a lower concentration in tumor tissue. In contrast, both CAMB NPs and CEAMB NPs can be significantly enriched in the tumor sites of mice, and among them, CEAMB NPs have the highest degree of enrichment (Fig. 8B, C). Because EGFR is highly expressed in ESCC tumor cells, our design using EGFR antibody demonstrates that CEAMB NPs have a better tumor targeting ability as well as the ability to deliver Adriamycin and siRNA to the tumor effectively.
The anti-tumor effect of CEAMB NPs in vivo
The above experiments have proven that RNAi carried by our designed NPs can effectively enrich the tumor site, therefore inhibiting the expression of the target gene. The delivered Adriamycin can trigger the apoptosis of tumor cells. Next, we tried to perform functional verification in vivo. The 510K cell-xenograft animal model was established, and the first treatment of NPs was performed when the volume of tumors was about 100 mm3. The second treatment was then performed on the 18th day after the first treatment (Fig. 9A). The tumor growth curve showed the CEAMB NPs, CEA NPs, CEAM NPs, and CEAB NPs could effectively inhibit tumor growth, but only CEAMB NPs could decrease tumor volume (Fig. 9B). The tumor volume and tumor mass results at the end of treatment showed the same trends as the tumor growth curve (Fig. 9C, D). These results could be directly observed from the photographs of tumor-bearing mice and tumors excised from the mice (Additional file 1: Fig. S8A). In addition, the results of the weight of the mice at the end of the treatment showed no significant difference (Additional file 1: Fig. S8B).
Meanwhile, the histopathological analysis of tissue sections isolated from the mice displayed no significant pathological changes in the heart, liver, spleen, and lung in all treated groups, revealing the safety of CEAMB NPs (Additional file 1: Fig. S8C). The above results indicated that chemotherapeutic drugs combined with multiple drug resistance gene siRNAs had a better therapeutic effect than the use of a single-drug resistance gene siRNA or chemotherapeutic drugs alone. Finally, the synergistic anti-tumor mechanisms in vivo were investigated. The results of IHC showed that the CEAMB NPs could effectively reduce the expression of MVP and BCL2 proteins via silencing their transcript mRNA, respectively (Fig. 9E). An apoptosis marker protein, Caspase3, was tested to detect the anti-tumor ability of CEAMB NPs . The results showed the expression level of Caspase3 protein in the CEAMB NPs group was the highest, which indicated that CEAMB NPs could effectively induce tumor cell apoptosis in vivo.
Furthermore, apparent nuclear shrinkage, fragmentation, and absence in the hematoxylin and eosin-stained sections of tumor tissue proved that the CEAMB NPs had a better anti-tumor effect (Fig. 9E). The above results indicated that the CEAMB NPs could effectively inhibit drug efflux and anti-apoptosis of tumor cells from eliminating the MDR of tumors, thereby enhancing the anti-tumor effect of chemotherapeutics. Moreover, the dual sensitization strategy loading with multiple drug resistance gene siRNAs could effectively improve the drug’s anti-tumor effect.