The hepatocellular carcinoma cell lines Huh-7 endogenously expressing TP53_Y220C and SK-Hep-1, obtained from the cell bank of the Shanghai Institute for Biological Sciences, Chinese Academy of Sciences (Shanghai, China), were routinely screened for mycoplasma to ensure purity. Additionally, considering that CTLs are HLA-restricted in recognizing target cells for killing effects, a target cell line Huh7-A0201, stably engineered to express HLA-A*02:01, was constructed. The Huh-7 and Huh7-A0201 cell lines were cultured in Dulbecco’s Modified Eagle’s Medium, procured from BasalMedia (Shanghai, China). Meanwhile, SK-Hep-1 cells were cultured in Minimum Essential Medium, also sourced from BasalMedia. Both media were supplemented with 10% fetal bovine serum from PAN-Seratech GmbH (Aidenbach, Germany) and 1% penicillin-streptomycin from MP Biomedicals (California, USA). The cells were incubated at 37°C in a 5% CO2 atmosphere.

To further evaluate the killing effect of human-derived CTLs on hepatocellular carcinoma in vivo, immunocompromised NOD/SCID mice were selected as animal models with being inoculated human HCC cell lines subcutaneously, and peritumoral injection of CTLs for assessing its inhibitory effect on hepatocellular carcinoma. Male NOD/SCID mice, aged 4 to 5 weeks, were sourced from Guangdong Yaokang Biotechnology Co., Ltd. (Foshan, China). They were housed in a Specific Pathogen-Free facility at the Laboratory Animal Center of Fujian Medical University (Fuzhou, China), with protocols approved by the university’s Animal Care and Use Committee. The experimental procedures involving these mice strictly adhered to the established guidelines for animal welfare.

Monocytes were isolated from the peripheral blood of healthy individuals, followed by the separation of CD14⁺ monocytes using magnetic bead technology. These cells were cultured in a Lymphocyte Serum-free Medium supplied by Corning Incorporated. Cultivation occurred in six-well plates, supplemented with 100 ng/mL GM-CSF and 50 ng/mL IL-4, both sourced from MCE Tech. Cytokines were replenished on the third and fifth days, and the cells were harvested on the seventh day for experimental use. This research was approved by the Ethical Review Board of Fujian Cancer Hospital under the reference number K2023-455-01. Blood samples were obtained from consenting volunteers who had signed informed consent documents.

To obtain a kind of extracellular vesicle to deliver mRNA, chimeric genes encoding ClyA fused to L7Ae or LLO were constructed and inserted into the multiple cloning sites of the pACYCDuet-1 vector, respectively. Plasmids encoding the RNA-binding protein L7Ae and the lysosomal escape protein LLO were recombinantly produced and introduced into Escherichia coli BL21(DE3) for expression. Positive clones were selected on chloramphenicol plates (50 µg/mL) and cultured in an LB medium for mass amplification. Initially, the liquid portion of the bacterial culture was collected, and a 10-min centrifugation at 4,000 g was performed to remove the bacteria. This was followed by a secondary centrifugation at 24,000 g for 3 h to isolate the outer membrane vesicles (OMV-LL).(19).

The pcDNA3.1(+) plasmid was engineered to include the following sequences after the T7 promoter: alpha-globin 5’UTR, KOZAK sequence, SP sequence (MHC class I signaling peptide designed to guide the peptide through the endoplasmic reticulum into the lumen), TP53_Y220C sequence (GTGGTGCCCTGTGAGCCGCCTGAGGTT), intermediate linker (GGSGGGGGSGG), TP53_Y220C sequence, intermediate linker, FLAG tag, and alpha-globin 3’UTR. mRNA encoding neoantigens was synthesized via in vitro transcription using T7 polymerase. The process began with linearization of the plasmid DNA by the XbaI restriction enzyme, which then served as the template for the transcription reaction facilitated by a Transcription Kit from Thermo Fisher Scientific (MA, USA). Following transcription, the mRNA was purified using a Transcription Cleaning Kit from the same manufacturer, adhering to their provided protocol.

DCs (1×10⁵ cells per well) without any treatment were used as the blank control group. DCs (1×10⁵ cells/well) were seeded into 24-well plates 24 h before transfection. For transfection, 9 μL of OMV-LL and 1.5 μL of Lipofectamine™ 3000 were mixed with 1 μL of mRNA encoding EGFP (mEGFP, APExBIO Technology LLC, USA) in opti-MEM medium for 15 min; the mixture was then added to the cells in a 24-well plate. 24 hours later, fluorescence images of each culture well were captured under microscope to assess transfection efficiency, and quantitative analysis was performed using flow cytometry. For electroporation, 1×10⁶ cells were collected, resuspended in 100 μL of opti-MEM medium, and transferred into a 2 mm electroporation cuvette (Suzhou Etta Biotech, China). The cuvette was placed in the X-Porator H1 electroporator (Suzhou Etta Biotech, China) and subjected to electroporation under the following conditions: 200 V voltage, 3 pulses, 800 µs duration, and 612 ms interval. Afterward, 1×10⁵ cells were removed and seeded into 24-well plates. 24 hours later, images of each well were captured using a microscope to assess transfection efficiency, which was quantitatively analyzed through flow cytometry.

Proteins from lysed DCs were isolated and analyzed via Western blot. The primary antibodies used were Mouse anti-DDDDK-Tag mAb and Mouse anti-Human β-actin mAb, both diluted according to the recommendations provided by the supplier. These were followed by horseradish peroxidase-conjugated secondary antibodies. Detection was performed using the enhanced chemiluminescence (ECL) method, following Pierce’s protocol for the ECL kit. The antibodies were sourced from ABclonal Technology Co., Ltd. (Wuhan, China).

ELISA plates were prepared by coating them with either IFN-γ or TNF-α capture antibodies and incubating them overnight at 4°C. Subsequently, the plates were washed with a washing solution and blocked with a blocking solution for 4 h at room temperature. Samples were then added to the plates and incubated overnight at 4°C. After another washing step, the plates were incubated with secondary antibodies for 1 h at room temperature. This was followed by a 30-min incubation with streptavidin-HRP at room temperature. After a final wash, the plates were developed with TMB for 15 min at room temperature. The color development was stopped using a stop solution, and the absorbance was measured at 450 nm.

The Human IFN-γ ELISpot kit was procured from Dakewe Biotech (Shenzhen, China). Following the kit’s instructions, the plates pre-coated with an antibody specific to IFN-γ were rinsed with PBS and then incubated with RPMI 1640 medium supplemented with 10% FBS for 10 min at room temperature. After removing the RPMI 1640 medium, CTLs (5×10⁴ cells/well) and tumor cells (1×10⁴ cells/well) were added to the plates. The plates were then cultured at 37°C for 24 h. Post-washing with PBS, the plates were incubated with biotinylated IFN-γ antibody and streptavidin-HRP, each for 1 h at 37°C. The color development was performed using AEC at room temperature in the dark for 30 min. After washing, the plates were air-dried in the dark, and the spots were counted using an ELISpot reader (CTLS6).

Target cells (Huh7-A0201 or SK-Hep-1) were cultured with CTLs at different E/T ratios in 96-well plates at 37°C for 6h. The Cytotoxicity Assay kit (Promega, Madison, WI, USA) was used to measure LDH activity in the supernatant, indicating cell damage. The cytotoxic cell percentage was determined by the formula: (Cocultivation LDH release - Tumor spontaneous release - CTLs spontaneous release)/(Tumor maximum release - Tumor spontaneous release) × 100%, which evaluates the cytotoxic effect of CTLs on the target cells.

Flow cytometric analysis was conducted using a FACS Canto Ⅱ flow cytometer from BD Biosciences (CA, USA). The antibodies utilized in the flow cytometry analysis included anti-CD80 labeled with FITC, anti-CD86 labeled with PE, anti-CD137 labeled with PE, anti-CD28 labeled with PE, and anti-CD69 labeled with FITC, all of which were sourced from BD Pharmingen. Data analysis was carried out with FlowJo version 10 software.

Male NOD/SCID mice were subcutaneously inoculated with Huh7-A0201 cells (1×10⁶ cells/spot). On day 7 post-inoculation, CTLs (2×10⁶ cells/spot) were injected peri-tumorally. Tumor sizes were monitored every two days using calipers.

Tissues from the heart, liver, spleen, lungs, and kidneys were extracted from the mice and fixed in paraffin. These tissue sections were then subjected to dewaxing, rehydration, and staining using Mayer’s hematoxylin and eosin staining method.

Data are presented as the mean ± standard deviation, except where specified otherwise. Statistical analysis was conducted using GraphPad Prism version 8.0. Significance was determined using either an unpaired t-test or a one-way ANOVA. The levels of significance were denoted as follows: ns (no significant difference); *p<0.05; **p<0.01***p<0.001.

The dataset comprising 502 hepatocellular carcinoma (HCC) cases was sourced from the cBioPortal website and analyzed for mutation characteristics using the “Maftools” R package developed by Mayakonda (20). Mutation types included non-synonymous changes, premature stop codons, frameshift and non-frameshift deletions, and insertions, both with and without frameshift. Out of 502 samples, 458 mutations were detected. The ten genes exhibiting the highest mutation rates were identified as TP53 TTN, FAT1, CDKN2A, MUC16, CSMD3, PIK3CA, NOTCH1, SYNE1, and LRP1B, with respective mutation rates of 70%, 42%, 22%, 21%, 20%, 19%, 18%, 18%, 17%, and 17% ( Figure 1A ). Consequently, TP53 was selected for further study in anti-HCC immunotherapy due to TP53_Y220C, a relatively common mutation in HCC, serving as the focus of our study (21, 22). The HCC cell line Huh-7, harboring the endogenous TP53_Y220C mutation, was identified using the TP53 Database website. Huh7-A0201, transfected with HLA-A*02:01, was utilized as a specific target cell, while SK-Hep-1, also expressing HLA-A*02:01 but lacking the TP53_Y220C mutation, served as a non-specific target cell.

The OMV-LL presents the RNA-binding protein L7Ae and the lysosome-escaping protein LLO on its surface (19). The OMV-LL vector was linked to the box C/D sequence of mRNA by L7Ae and achieved lysosomal escape of antigen via LLO. Verification of OMV-LL dimensions through dynamic light scattering and electron microscopy revealed a consistent size of around 120 nm. ( Figures 1B, C ).

Monocyte-derived dendritic cells (MoDCs) were transfected with mEGFP using various delivery methods. The success of the transfection was evaluated by observing the expression of green fluorescent protein under a fluorescence microscope and quantifying the proportion of EGFP⁺DCs via flow cytometry. Compared with the blank control group, effective mRNA delivery was observed in the other three groups ( Figure 2A ). The transfection efficiency of the OMV-LL delivery system was approximately 3.75 ± 0.57%, that of the Lip3000 lipid nanoparticles was approximately 4.67 ± 0.43%, while the electroporation method achieved a significantly higher efficiency of approximately 89.47 ± 7.88%, markedly surpassing the performance of both OMV-LL and Lip3000 ( Figure 2B ).

MoDCs were transfected with mRNA encoding the mTP53_Y220C-Flag for further validation using the same delivery methods. Western blot analysis showed higher expression levels of the TP53_Y220C-Flag fusion protein in the electroporation group compared to the two groups ( Figure 2C ).

However, while investigating the impact of three different mRNA delivery methods on the maturation of DCs, it was found that all three delivery approaches significantly increased the expression levels of CD80 and CD86 on DC surfaces compared to the untreated control group. Notably, the OMV-LL group demonstrated significantly higher expression levels than the Lip3000 and electroporation groups ( Figure 3A ). These results implied that OMV-LL could induce a highly efficient immune response, although its efficiency in delivering mRNA is limited.

Additionally, the expression levels of CD69, a surface indicating cellular activation, along with the co-stimulatory molecules CD28 and CD137, were assessed via flow cytometry. The results showed a substantial increase in the expression of CD69, CD28, and CD137 on the surfaces of CTLs in both the OMV-LL and electroporation groups compared to the control group ( Figure 3B ). This suggests that introducing mTP53_Y220C into DCs using OMV-LL and electroporation methods more effectively promotes T-cell activation, thereby enhancing the anti-tumor effects of the subsequent CTL responses.

To evaluate the anti-hepatocellular carcinoma effects of CTLs induced by three different mRNA delivery methods, Huh7-A0201 and SK-Hep-1 cell lines were selected as target cells. LDH assays revealed that the cytotoxic activity of CTLs from the OMV-LL and electroporation groups against the target cells Huh7-A0201 at E:T ratios of 20:1 and 10:1 was notably higher than that of the Lip3000 group, and the OMV-LL group demonstrated the same cytotoxicity level as the electroporation group ( Figure 4A ). Meanwhile, no significant difference in the cytotoxicity against SK-Hep-1 cells was observed across all groups at the same E:T ratios ( Figure 4A ), suggesting that the CTLs’ cytotoxicity induced by mTP53_Y220C is antigen-specific.

Additionally, the secretion of pro-inflammatory cytokines IFN-γ and TNF-α was assessed by ELISA in each group of CTLs co-cultured with Huh7-A0201 at an E:T ratio of 10:1. The results indicated that the production of IFN-γ and TNF-α was significantly enhanced in both the OMV-LL and electroporation groups compared to the Lip3000 group. The OMV-LL group exhibited less levels of IFN-γ and TNF-α secretion than the electroporation group ( Figure 4B ). To further confirm the secretion of IFN-γ at the individual cell level, the ELISPOT method was used to identify IFN-γ⁺ CTLs. The results showed that both the OMV-LL and electroporation groups had a significantly greater number of IFN-γ⁺ CTLs compared to the Lip3000 group. The OMV-LL group also had a lower count of IFN-γ⁺ CTLs than the electroporation group ( Figure 4C ).

In the in vivo study, a subcutaneous tumor model using Huh7-A0201 cells was established to assess the anti-tumor effects of CTLs activated by DCs transfected with Lip3000/mTP53_Y220C, OMV-LL/mTP53_Y220C, and electroporation/mTP53_Y220C, respectively ( Figure 5A ). Compared to the control group, the average tumor growth curves in the OMV-LL, electroporation, and Lip3000 groups were slower, and the average tumor volumes showed a substantial decrease ( Figures 5B, C ). Analysis of tumor tissues revealed that infiltration of CD8⁺ T-cells was considerably higher in the OMV-LL and electroporation groups compared to the Control and Lip3000 groups ( Figure 5D ). These findings suggest that the anti-tumor effect of CTLs induced by DCs loaded with mTP53_Y220C was significantly more effective than that of CTLs not loaded with mTP53_Y220C, and the ability of CTLs to infiltrate tumor tissue was markedly greater in the OMV-LL and electroporation groups compared to the Lip3000 group.

Eleven days after initiating CTL therapy, blood samples were collected from the mice to measure serum concentrations of blood urea nitrogen (BUN), creatinine (CRE), alanine transaminase (ALT), and aspartate transaminase (AST). The results indicated no notable differences in the BUN, CRE, ALT, and AST levels between the treated and control groups, suggesting that the CTL injections did not burden the mice’s liver and kidneys ( Figure 5E ). The main organs were also isolated and subjected to H&E staining, revealing no obvious pathological changes ( Figure 5F ). These results confirm that CTL therapy, administered via in vivo injection, exhibits good biosafety.