Cells and tissues

The lung adenocarcinoma cell lines of human, including Calu-3, H358, H23 and H2122, were obtained from Center for Molecular and Cellular Sciences of Shanghai. The Lewis lung cancer cell line derived from mice, LLC, was obtained from FengHuiShengWu in Hunan, China. These cell lines were cultured in RPMI-1640 or DMEM medium which were purchased from VivaCell in Shanghai, China. Additionally, the culture media were supplemented with 1% penicillin-streptomycin solution from Beyotime in Shanghai and 10% fetal bovine serum from VivaCell. All the cells mentioned above were cultured in a controlled environment at 37℃ with an atmosphere containing 5% CO2. From 2013 to 2023, a total of fifty-five lung adenocarcinoma tissues and twenty normal lung epithelial tissues were collected from individuals who were undergoing lung cancer resection at our hospital (the 920th Hospital of the Joint Logistics Support Force of PLA). The selection criteria for inclusion in the study required a confirmed diagnosis of lung adenocarcinoma through pathology examination and a clear determination of the KRAS status through genetic testing. The entire process of sample collection was conducted in accordance with the ethical guidelines set forth by the ethics committee of our hospital.

Transfection

The lentivirus carrying ZNF24 gene knockdown constructs were transfected into the cells. The lentivirus and plasmid were acquired from Genechem, a biotechnology company located in Shanghai, China. The cells were initially cultured in a 6-well plate at a density of 2 × 104 cells. Then, the lentivirus carrying the gene knockdown constructs (at a multiplicity of infection (MOI) of 10 and a volume of 2 mL) was added to the cells. After 6–8 h, the RPMI 1640 medium containing 10% Fetal Bovine Serum was replaced, and the cells were further cultivated for 48 h. To select for cells that underwent successful gene knockdown, the researchers used RPMI 1640 medium containing 2 mg/mL purinomycin for a period of 3–5 days. This allowed for the elimination of cells that did not exhibit stable ZNF24 expression decrease. Additionally, the researchers were also interested in studying the overexpression of KRASG12C and SLC7A5 genes. To achieve this, the over-SLC7A5 or over-KRASG12C plasmids were transfected into the cells using the Lipofectamine 3000 kit, following the instructions provided by the manufacturer.

Screening and validation of ZNF24 inhibitors

Molecular docking is a virtual screening of the compound database based on various scoring functions, and the strongest compounds that can bind to target proteins can be predicted. Accelrys Discovery Studio 3.5 [32], Pymol [33, 34] and Autodock [32] were used for docking. The crystal structure of ZNF24 was mainly screened from the protein database (www.pdb.org), and human protein and prioritizing the protein crystal structure with higher resolution were selected and prioritized from the database. Then, the Pymol software was used to optimize the crystal structure of ZNF24 protein, remove water molecules and ligand molecules and add hydrogenation and charges to obtain a ZNF24 protein structure that is close to the activity in the organism. In this study, the LibDock tool of Discovery Studio (DS) was used to specify the number of conformations and docked them through ‘’High Quality’’. LibDock was a high-throughput docking algorithm used to locate the formation of protein active site based on polar interaction sites (hot spots) in ligands generated by catalysts [32]. The conformational space of ligands was recognized by AutoDock using the Lamarckian Genetic Algorithm (LGA). Both receptors and ligands were used in MOL2 format. Virtual screening of small molecules was completed in this study from small molecule databases (https://www.apexbt.com/search/screening-library?cat=778).

Next, to demonstrate the affinity between small molecule compounds and target protein ZNF24 through Surface Plasmon Resonance (SPR). Firstly, the ZNF24 gene was inserted into the prokaryotic expression plasmid pET-28a (+) (Qingke, China). The expression plasmids of ZNF24 were transformed into Escherichia coli BL21 (DE3) and screened with 50 µg/ml kanamycin. After PCR and agarose gel electrophoresis identification, a single positive colony was inoculated into 100 mL of LB medium and grown at 37 °C. The protein was expressed inducibly with Auto-induction Medium for 6 h at 37 °C and overnight at 30 ℃, 200 rpm. E. coli BL21 (DE3) was collected by centrifugation at 12,000 rpm for 20 min and ultrasonicated. The supernatant contained soluble protein, and the precipitate contained inclusion body protein. The inclusion body protein was collected by bacterial sonication in a bacterial lysis buffer. The dissolved inclusion body proteins were purified with the HisPur Ni-NTA Purification Kit (88,229, Thermo, Germany). 8% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and WB were used to detect the expression of target proteins. The ZNF24 protein was diluted to 50 µ g/mL with a fixed reagent (10 mM sodium acetate, pH 4.0). Firstly, the surface of the CM5 chip is coated with 400 mM 1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) and 100 mM N-Hydroxy succinimide (NHS) at a flow rate of 10 µ L/min for 420 s. Secondly, the ZNF24 protein (50 µg/mL) was added to the experimental channel (Fc2) for flow rate of 10 µ L/min with a fixed amount of approximately 10,000 RU. Finally, the chip was loaded with 1 M ethanolamine at a rate of 10 µ L/min was blocked for 420 s. The reference channel (Fc1) was not injected with ZNF24 protein and other operations were the same as the experimental channel. Daptomycin was diluted 20 times with a dilution reagent to reduce the DMSO content from 100 to 5%. Daptomycin was injected into the experimental and reference channels at a rate of 30 µ L/min with appropriate time for binding and dissociation.

Bioinformatics analysis

The association between the mutation status of the KRAS gene and the infiltration of CD8+ T cells was examined using the TIMER database (https://cistrome.shinyapps.io/timer/) [35]. Specifically, it included a total of 164 cases of lung adenocarcinoma tissues with KRAS mutation and 351 cases of lung adenocarcinoma tissues with wild-type KRAS. In addition, the relationship between the expression of PD-L1 and the infiltration of CD8+ T cells was also predicted using the TIMER database.

Immunohistochemical (IHC)

Initially, the sections were incubated at a relatively high temperature of 68℃ for a duration of 2 h. Following this, the sections underwent dewaxing using xylene, anhydrous ethanol, a gradient ethanol series, and distilled water. To create a suitable environment for further analysis, the sections were then boiled in an acid buffer with a pH of 6.0 under high temperature and pressure for a duration of 2 min. After ensuring optimal conditions for analysis, the endogenous peroxidase activity in the tissue sections was blocked using a 0.3% hydrogen peroxide solution. To specifically evaluate the expression levels of PD-L1, ZNF24, and SLC7A5 proteins, the tissue sections were first incubated with a normal goat serum containing 5% concentration for 25–30 min. Subsequently, the ZNF24 antibody (11219-1-AP, Proteintech, USA), PD-L1 antibody (ab213524, Abcam, England), and SLC7A5 antibody (28670-1-AP, Proteintech, USA) were diluted at a ratio of 1:200 and allowed to incubate overnight at a temperature of 4℃. Additionally, we conducted immunohistochemistry tests of animal tumor tissues using antibodies specifically targeting CD4 antibody (ab183685, Abcam, England), CD8 antibody (ab217344, Abcam, England), CD34 antibody (L08JU02, ZEN-BIOSCIENCE, China), Ki-67 antibody (ab279653, Abcam, England), PD-1 antibody (ab214421, Abcam, England) and PD-L1 antibody (ab233482, Abcam, England). To detect the presence of the primary antibodies, the sections were further processed by incubating them with HRPlabeled secondary antibody (PV-6000, ZSGB-BIO, China) at a temperature of 37 °C for 1 h. Following this incubation step, the sections were stained with DAB (3,3’-diaminobenzidine) for a duration of 30 to 60 s, allowing the visualization of the target proteins.

RNA extraction and real-time fluorescent quantitative PCR assay (RT-qPCR)

Total RNA was isolated from cells according to the instructions of the RNA extraction kit (LS1040, Promega, Shanghai, China). The quality and concentration of total RNA were checked by using a NanoDrop 2000 spectrophotometer. Next, cDNA was synthesized using a fixed one-step RT-PCR kit (A6120, Promega, Shanghai, China) according to the manufacturer’s instructions. RT‐qPCR experiments were completed using the SYBR Green SuperMix system (Tsingke Bio technology, TSE201, Beijing, China) as follows: 94 ℃ for 10 min, followed by 40 cycles at 94 ℃ for 15 s and 60 ℃ for 1 min. GAPDH was used as a reference gene. The changes in gene level were estimated by the 2−ΔΔCT method. Finally, the primer sequences used for RT-qPCR were visually presented and listed in Table 1.

Table 1 The primer sequences for qPCRProtein isolation and Western blotting

The total protein was isolated from cells using the protein extraction kit (Solarbio, R0010, Beijing, China). To determine the protein concentration, the BCA protein assay kit (Beyotime, P0010, Shanghai, China) was employed. For the experiment, primary antibodies specifically targeting ZNF24, SLC7A5, β-actin, and PD-L1 were obtained from Proteintech (11219-1-AP, USA, diluted at 1:1000), Santa Cruz (sc-374232, USA, diluted at 1:500), ZSGB-BIO (TA-09, China, diluted at 1:1000), and Abcam (ab82458, England, diluted at 1:1000) respectively. These primary antibodies were then incubated at 4℃ overnight. After the primary antibodies were incubated with the samples, the next step involved diluting the secondary antibodies. Specifically, either goat anti-mouse IgG-HRP or goat anti-rabbit IgG-HRP antibodies (ZSGB-BIO, Zhongshan, China) were mixed in a dilution ratio of 1:10000 and incubated at 37℃ for 1 h. ImageJ software was utilized for quantitative analysis.

Immunofluorescence

To prepare the cell monolayers, every slide was spread with 2 × 105 cells. These slides were then placed in an incubator at 37 ° C for a period of 4–6 h, allowing the cells to adhere and form a monolayer. Following this, the drugs were added according to pre-set groups and continued to cultivate for 24 h. Then, cells were fixed with 4% paraformaldehyde and permeabilized with 0.5% Triton X-100 at room temperature for 30 min. To further prepare the slides for subsequent analyses, they were percolated with a 0.5% Triton X-100 solution for a duration of 15 min. Following the permeabilization step, the slides were blocked using a 5% BSA solution. After blocking, a primary antibody specific to PD-L1 (ab213524, England) was incubated with the cells at a dilution of 1:100. The incubation was carried out overnight at 4 °C to ensure sufficient time for antibody binding. Following the overnight incubation, the slides were washed three times with PBS. To visualize the bound primary antibody, a goat anti-rabbit secondary antibody (ZSGB-BIO, Zhongshan, China) conjugated with a fluorescent marker was added to the slides. The secondary antibody was incubated with the slides for 1 h at 37 °C. Additionally, DAPI, a fluorescent dye that stains DNA, was added to the slides. The final step involved obtaining photographs of the slides using a fluorescence microscope.

Cell counting Kit-8 (CCK-8) assay

Cells (H2122, H2122-shZNF24, H358, H358-shZNF24) were evenly distributed in 96-well plates, with each well containing 4 × 103 cells. CD8+ T cells were added to the plates and allowed to co-culture with the target cells. The cells were subjected to different drug treatments based on their respective conditions. Following this, 10 µL of the CCK-8 solution (CA1210, Solarbio) was added to each well of the experimental plates. The plates were then incubated once again at a temperature of 37 °C, this time for a duration of 1 h. To measure the impact of the drugs and treatment on the cells, we utilized a microplate reader (Model 680, Bio-RAD). The reader was equipped with a 450 nm absorbance filter, which helped in reading and analyzing the values obtained from the experimental wells.

Isolation of CD8+ T cells

PBMCs were isolated with Human Peripheral Blood Lymphocyte Isolate (P8610, Solarbio, China). CD8 + T cells were then isolated from the PBMCs using the EasySep™ Direct Human CD8+ T cell Isolation kit (19663, StemCell Technologies, Canada). The separation methods were carried out according to the provided instructions. CD8+ T cells were cultured in RPMI-1640 medium and activated by Ultra-LEAF™ Purified anti-human T-Activator CD3/CD28 (317325/302933, BioLegend, USA) for 3 days. Then, pre-activated CD8+ T cells directly co-cultured with cancer cells.

Co‑culture and lactate dehydrogenase (LDH) cytotoxicity assays

H2122 or H358 cells were seeded into 96-well plates at a concentration of 8 × 103 per well. After 24 h, activated CD8+ T cells were co-cultured with adherent NSCLC cells at a ratio of 3:1 or 5:1 or 10:1 for 48 h. After co-culture, CD8+ T cells were harvested with the application of the density separation method. The cytotoxicity of CD8+ T cells was detected using the LDH cytotoxicity detection kit (C0017, Beyotime, China). After selecting the optimal ratio of tumor cells and CD8+ T cells, the subsequent experiments were conducted by the above method.

Co‑culture and flow cytometry

CD8+ T cells of co-cultured supernatant were collected in a 15 ml centrifuge tube, centrifuged at 400 g for 5 min, washed twice with PBS, and resuspended with 100ul Cell Staining Buffer for cell precipitation. Following the standard procedures of flow cytometry, cells were stained with anti-CD3 (317306, Biolegend, USA), anti-PD-1 (379206, Biolegend, USA), anti-PD-L1 (329708, Biolegend, USA) or anti-CD8 (344722, Biolegend, USA). The results were obtained from the Fortessa platform (BD Biosciences) and analyzed using FlowJo software (BD Biosciences).

Co‑culture and enzyme‑linked immunosorbent assay (ELISA)

Tumor cells and CD8+ T cells were co-cultured at a ratio of 1:5 for 48 h. Cell culture medium was collected and centrifuged at 500 g 4 °C for 15 min. The top layer was collected, and the levels of interleukin-2 (IL-2, RXSWS106150H, QUANZHOURUIXIN, China), interferon-γ (IFN-γ, RXSWS106205H, QUANZHOURUIXIN, China) and Tumor Necrosis Factor-α (TNF-α, RXSWS8100010H, QUANZHOURUIXIN, China) were measured using a commercial ELISA kit following the manufacturer’s protocol. The OD value of each well was measured at 450 nm via a Microplate Reader (Thermo Fisher Scientific Inc., USA) and calculated at the linear portion of the curve.

Animal experiments

To investigate the effect of KRAS/ZNF24/SLC7A5/PD-L1 axis on the immune function of mice in vivo, we used C57BL/6 mice with normal immune function. Fifteen 6-week-old female C57BL/6 mice (SiBeiFu biotechnology company, Beijing, China) were randomly divided into three groups. These mice were kept under specific pathogen-free (SPF) conditions to ensure their health and wellbeing throughout the experiment. Mice experiments were approved by the Animal Ethics Committee of 920th Hospital of the Joint Logistics Support Force of PLA. To induce tumor growth, we subcutaneously injected cells (2 × 106 at each spot) into the armpits of the mice. Mice were challenged with PBS, DAPT or RGD-KGH-R1-ScFv in the intra-tumoral. Tumor size was measured by caliper every two days. At the end of the experiment, the method of spinal dislocation is used to euthanize mice. The tumor tissues were collected and prepared for further analysis by embedding them in paraffin. The changes of histology were observed by HE staining. Additionally, we conducted immunohistochemistry tests using antibodies specifically targeting CD4 antibody (ab183685, Abcam, England), CD8 antibody (ab217344, Abcam, England), CD34 antibody (L08JU02, ZEN-BIOSCIENCE, China), Ki-67 antibody (ab279653, Abcam, England), PD-1 antibody (ab214421, Abcam, England) and PD-L1 antibody (ab233482, Abcam, England). These antibodies were diluted according to the manufacturer’s instructions and were incubated with the tissue sections overnight at a temperature of 4 ℃. HRP‑labeled secondary antibody (PV-6000, ZSGB-BIO, China) was incubated at 37 °C for 1 h. Finally, the tissue sections were stained with DAB for 30–60 s to visualize the immune-related markers.

Statistical analysis

Data are shown as the mean ± standard deviation. All statistical analyses were performed using SPSS software version 22.0 and graphed with GraphPad Prism 8.0 software. Pearson’s chi-square test was used to analyze the IHC results. Unpaired Student’s t-test was used to compare the means of two groups of data. One-way ANOVA was used to compare multiple groups. The data are expressed as the mean ± SD of three independent experiments. p < 0.05 indicates statistical significance.