The potential of circITFG2 as a therapeutic target in lung squamous cell carcinoma

Introduction

Lung cancer remains the leading cause of cancer-related death worldwide, accounting for approximately 18–20% of all cancer-related deaths (1). Among histological subtypes, lung squamous cell carcinoma (LUSC) constitutes 12–25% of all lung cancer cases (2). However, due to its considerable heterogeneity, lack of targeted drugs for classical driver mutations (3), and intrinsic resistance to conventional radiotherapy and chemotherapy (4), research progress in LUSC has lagged behind that of lung adenocarcinoma (LUAD). Although immune checkpoint inhibitors (ICIs) have revolutionized the treatment of advanced non-small cell lung cancer (NSCLC), approximately 60–70% of patients with LUSC eventually develop immune evasion or treatment resistance (5), resulting in a dismal 5-year survival rate of less than 18% (6). Therefore, there is an urgent need to elucidate the molecular mechanisms underlying LUSC progression and immune evasion to identify novel biomarkers and therapeutic targets.

Noncoding RNAs (ncRNAs) have opened new avenues in cancer research due to their critical regulatory roles in tumorigenesis, including lung cancer (7,8). Circular RNAs (circRNAs), a type of stable and specific ncRNA with closed-loop structures, show promise in facilitating early cancer diagnosis, prognosis, and targeted treatment due to their low immunogenicity (9). Recent studies have attested to their role in tumor growth, metastasis, and drug resistance through the modulation of key signaling pathways, acting as microRNA (miRNA) sponges, interacting with RNA-binding proteins, and possibly serving as protein translation templates (10-12). The use or targeting of naturally occurring circRNAs may represent a promising alternative to existing RNA-based therapies (13). In NSCLC, aberrantly expressed circRNAs have been implicated in epithelial-mesenchymal transition (EMT), chemotherapy resistance, and programmed death-ligand 1 (PD-L1)-mediated immune evasion (14-16). Although certain circRNAs are known to act as oncogenes or tumor suppressors, the role of key circRNAs in LUSC and their interaction with the tumor immune microenvironment (TIME) needs to be further investigated.

Integral membrane protein 2A (ITM2A), a type II transmembrane protein encoded by the BRICHOS domain containing 2A (BRICD2A) gene (17-19) acts as a tumor suppressor in several cancers by activating mTOR-dependent autophagy and GATA-binding protein 3 (GATA3)-regulated T-cell maturation (20,21). Zhang et al. proposed that ncRNA-mediated competing endogenous RNA (ceRNA) networks might regulate ITM2A expression (22), although its upstream regulators and precise mechanisms in LUSC remain unknown.

In this study, we identified a circRNA (circITFG2) derived from the ITFG2 gene, which was significantly downregulated in LUSC clinical samples and negatively correlated with malignant progression. Mechanistic investigations revealed that circITFG2 overexpression sponged miR-526b-5p to promote ITM2A expression, thereby inhibiting malignant phenotypes and reshaping the TIME of LUSC by enhancing effector immune cell infiltration and cytotoxic effector molecule expression. We also preliminarily assessed the potential of the circITFG2-ITM2A axis to disrupt PD-L1 protein stability via the autophagolysosomal pathway, providing new insights into circRNA-mediated immune regulation in LUSC. We present this article in accordance with the ARRIVE and MDAR reporting checklists (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1632/rc).

Methods

Clinical samples and healthy peripheral blood

Tissue samples from adult patients (regardless of gender) with primary LUSC and adjacent normal tissues, confirmed at The Third Affiliated Hospital of Zunyi Medical University between November 2020 and the study’s conclusion, were collected along with clinical data. Fresh samples were rinsed with saline and frozen, while others were dehydrated and stored in paraffin. Clinical data were obtained from electronic records. circRNA expression profiles in three paired LUSC samples were analyzed with the Arraystar human circRNA microarray (Antitumor and Health reconstruction Biotechnology Co., Ltd., Shanghai, China). For the top 10 downregulated genes, screening was performed via the MiOncoCirc database—the first clinical tumor-derived circRNA repository. Research team members voluntarily provided peripheral blood samples, with informed consent. The study was approved by the Ethics Committee of The Third Affiliated Hospital of Zunyi Medical University (No. 2020-1-323) and informed consent was taken from all the patients. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Cell lines and animal experiments

The study used cell lines from Shanghai Cell Bank of Chinese Academy of Sciences (Shanghai Cell Bank, CAS), including BEAS-2B, NCI-H1703, HCC1588, HEK-293T, and LLC1, all authenticated by STR profiling and maintained within four passages. Human peripheral blood mononuclear cells (PBMCs) were isolated from healthy donors and activated. Animal experiments involved 5- to 6-week-old female BALB/c nude and C57BL/6 mice (n=5 per group, randomly divided into 2–4 groups) from Nantong Bestest Biotech Co., Ltd. (Nantong, China), which were housed under specific pathogen-free (SPF) conditions. Mice were euthanized when tumors reached predefined volumes. We conducted experiments at the Zhuhai Campus as students of Zunyi Medical University. Therefore, all ethical approvals for experiments conducted there were temporarily handled by the Fifth Affiliated Hospital (Zhuhai) of Zunyi Medical University, which is an official subsidiary of Zunyi Medical University. Animal experiments were performed under a project license (No. 2022ZH0025) granted by the Fifth Affiliated Hospital (Zhuhai) of Zunyi Medical University, in compliance with national guidelines for the care and use of animals. A protocol was prepared before the study without registration.

RNA extraction, genomic DNA (gDNA) extraction, and quantitative reverse-transcription polymerase chain reaction (RT-qPCR)

Total RNA was extracted with TRIzol (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA), and nuclear and cytoplasmic RNA were separated via the Cytoplasmic & Nuclear RNA Purification Kit (Wuhan EpiGentek Biotechnology Co., Ltd., Wuhan, China). RNA concentration and purity were assessed with the NanoDrop 2000 system (Thermo Fisher Scientific). Complementary DNA (cDNA) was synthesized with the Beyotime II First Strand cDNA Synthesis Kit (Beyotime, Shanghai, China), and genomic DNA (gDNA) was extracted with gDNA kit (Omega Bio-Tek, Norcross, GA, USA). For RNase R treatment, 2 µg of total RNA was incubated with or without RNase R (3 U/µg; Epicenter Technologies, Thane, India) for 20 minutes at 37 ℃, followed by purification with the RNeasy MinElute Cleanup Kit (Qiagen, Hilden, Germany). RT-qPCR was conducted with a miRNA qPCR kit (SYBR Green method; Shanghai Sangon Biotech, Shanghai, China), with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or U2 used as reference genes. The primer sequences are provided in Table S1.

Fluorescence in situ hybridization (FISH) and autophagosome and lysosome detection

Paraffin sections were prepared with products acquired from Tiangen Biotech (Beijing, China), and Cy3-labeled circITFG2 probes from Shanghai Gefan Biotechnology (Shanghai, China) were applied. For the tissue in the FISH assay, tissues were fixed, permeabilized, and hybridized with circITFG2 probes overnight at 37 ℃, which was followed by washing with saline-sodium citrate buffer. Nuclei were stained with 4’,6-diamidino-2-phenylindole (DAPI), and images were obtained with a fluorescence microscope (Olympus, Tokyo, Japan). After a 1-hour phosphate-buffered saline (PBS) treatment, cells on coverslips were stained with an autophagy detection kit (monodansylcadaverine, MDC) and Lyso-Tracker Red, both obtained from Beyotime Biotechnology (Suzhou, China). Fluorescence signals (green for MDC-labeled autophagic vacuoles and red for lysosomes) were observed under ultraviolet light.

Gene set enrichment analysis (GSEA)

Patients were grouped into high- and low-risk categories according to the risk score. GSEA was employed to assess differences in pathway enrichment between these groups. Reference gene sets were obtained from the Molecular Signatures Database (MSigDB) database. Pathways with an adjusted P value <0.05 were considered significantly enriched and ranked based on enrichment scores.

Western blot analysis

Proteins were extracted with radioimmunoprecipitation assay (RIPA) buffer (cat. no. P0013B; Beyotime), and their concentrations were quantified with a bicinchoninic acid protein assay kit (cat. no. P0010S; Beyotime). The proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently transferred onto polyvinylidene fluoride (PVDF) membranes. The membranes were blocked with a rapid blocking buffer and incubated overnight with the following primary antibodies: anti-PD-L1 (GeneTex, Irvine, CA, USA), anti-ITM2A (Wuhan Sanying Biotechnology, Wuhan, China), anti-LC3 I/II (Novus Biologicals, Centennial, CO, USA), anti-P62 (Novus Biologicals), anti-Lamp1 (Wanleibio, Shenyang, China), anti-perforin (Wanleibio), and anti-granzyme B (Wanleibio). Thereafter, the membranes were incubated at room temperature with horseradish peroxidase (HRP)-conjugated goat anti-mouse/rabbit immunoglobulin G (IgG) secondary antibodies. Protein bands were visualized by chemiluminescence and captured with the AlphaView analysis system (ProteinSimple, San Jose, CA, USA). Quantification of individual protein bands was performed via densitometry with ImageJ software (US National Institutes of Health, Bethesda, MD, USA).

Plasmid construction, transfection, and stable cell line generation

The circITFG2 overexpression vector (pLC5-CIRC) was a donated by Dr. Haitao Li (Zunyi Medical University). Short hairpin RNA (shRNA) vectors (pLKO.1-CMV-copGFP-PURO) and ITM2A-overexpression vectors (pCDH-CMV-MCS-EF1-RFP-T2A-Puro) were purchased from Tsingke Biotechnology (Guangzhou, China). miRNA mimics were obtained from GenePharma (Suzhou, China). Luciferase reporter plasmids (pmirGLO) were purchased from GENEWIZ (South Plainfield, NJ, USA). Lentiviral packaging plasmids (psPAX2 and pMD2.G) were obtained from System Biosciences (Palo Alto, CA, USA). All constructs were validated by sequencing.

Colony formation, apoptosis, migration, and invasion assays

For the colony formation assay, cells were plated in six-well plates at a density of 1×10³ cells per well and cultured for 14 days. Colonies were then fixed with 4% paraformaldehyde (PFA) and stained with 0.1% crystal violet solution. The terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) apoptosis assay was performed by fixing test cells with PFA followed by permeabilization. Cells were then incubated with TUNEL reaction mixture containing terminal deoxynucleotidyl transferase (TdT) and fluorescein-labeled dUTP to label DNA fragmentation ends in apoptotic cells. After washing, fluorescent signals were visualized under a fluorescence microscope, with apoptotic cells exhibiting specific fluorescence due to DNA breaks, enabling quantitative assessment of apoptosis. For Transwell migration and invasion assays, cells were seeded in serum-free medium into the upper chambers of Transwell inserts, either uncoated (migration) or Matrigel-coated (invasion). Following incubation, migrated or invaded cells were fixed with methanol and stained with crystal violet. Nonmigrated or noninvaded cells on the upper membrane surface were removed with a cotton swab, while migrated or invaded cells on the lower surface were counted under an optical microscope. These standardized protocols allow for the quantitative evaluation of cellular proliferation, apoptosis, and metastatic potential under experimental conditions.

RNA immunoprecipitation (RIP)

The RIP experiment was performed according to the protocol provided by BioAssay Biotechnology (Hayward, CA, USA). Briefly, the target cells were lysed, and the cell lysate was incubated with magnetic beads conjugated with either anti-argonaute 2 (AGO2) antibody or control anti-IgG antibody at 4 ℃ overnight under rotation. The beads were then collected and washed with RIP wash buffer, which was followed by digestion with proteinase K at 55 ℃ for 30 minutes. RNA was extracted with TRIzol reagent (Invitrogen), reverse-transcribed, and subsequently analyzed via gel electrophoresis.

Luciferase reporter assay

HEK293T cells were plated in 96-well plates for 24 hours and then co-transfected with miRNA mimics and a luciferase reporter vector containing the circITFG2 3’UTR. After 48 hours, dual-luciferase activity (firefly/Renilla) was measured with a microplate reader (Promega, Madison, WI, USA). The results were normalized to the control group and expressed as a percentage of the relative luciferase activity.

T cell-mediated cytotoxicity assay

Human PBMCs were cultured in RPMI-1640 supplemented with human interleukin-2 (IL-2) and anti-CD3/CD28 beads (Novoprotein Scientific, Beijing, China) to activate CD8⁺ T cells for 1 week. The activated PBMCs were then cocultured with target stable cells (at a 3:1 ratio) in a Transwell six-well plate for 48 hours, which was followed by protein extraction from the lower chamber cells for downstream analysis.

Immunohistochemistry

Cells on coverslips were fixed with 4% PFA, treated with 3% H₂O₂ to block endogenous peroxidase, and blocked with 10% horse serum. Sections were incubated with anti-PD-L1 primary antibody (GeneTex), washed, and then incubated with a biotinylated secondary antibody, followed by HRP-conjugated streptavidin. After a 5-minute DAPI reaction, slides were counterstained with hematoxylin and imaged under a light microscope.

Statistical analysis

All experiments were repeated at least three biological replicates. Data are presented as mean ± standard error and analyzed using t-test or ANOVA, with sample inclusion/exclusion strictly following pre-defined protocol criteria. Pearson correlation analysis was performed to assess the associations among circITFG2, miR-526b-5p, and ITM2A expression. Statistical significance was set at a P value <0.05 Statistical analyses were conducted with SPSS 18 (IBM Corp., Armonk, NY, USA).

Results

Expression of circITFG2 was downregulated in LUSC and correlated with aggressive clinical characteristics

circRNA expression profiles in three paired LUSC samples revealed 223 upregulated and 173 downregulated differentially expressed circRNAs (Figure 1A). From the top 10 downregulated circRNAs, hsa_circ_0000374 (derived from exons 5–9 of ITFG2 on chromosome 12) was selected (Figure 1B). RT-qPCR demonstrated that circITFG2 expression was significantly lower in LUSC cell lines than in normal bronchial epithelial cells (Figure 1C). Its circular structure was confirmed through back-splicing junction validation (Sanger sequencing; Figure 1D) and RNase R resistance (Figure 1E), while subcellular localization analysis demonstrated predominant cytoplasmic distribution (Figure 1F,1G). These results collectively identified circITFG2 as a bona fide cytoplasmic circRNA. Validation in 25 paired LUSC tissues confirmed consistent downregulation of circITFG2 (Figure S1A), with RNA expression levels significantly correlating with clinical characteristics (Figure S1B-S1H).

Figure 1 Characterization of circITFG2 in LUSC cells. (A,B) Scatter plot and cluster heatmap for evaluating differential circRNA expression between LUSC tumor tissues and adjacent normal tissues (|FC| >1.5 and P<0.05). (C) Expression levels of circITFG2 in LUSC cell lines and normal bronchial epithelial cell (BEAS-2B). (D) Schematic diagram of the genomic splicing pattern of circITFG2. The splicing junction was validated via Sanger sequencing. (E) RT-qPCR analysis of circITFG2 and its linear counterpart abundance in HCC1588 and H1703 cells after RNase R treatment. The levels of circITFG2 and its linear counterpart were normalized to values measured in mock-treated samples. (F) Cellular RNA nuclear-cytoplasmic fractionation assay. circITFG2 was predominantly localized in the cytoplasm of LUSC cells. (G) RNA FISH of circITFG2. Nuclei were stained with DAPI. Scale bar =20 µm. Data are presented as the mean ± SEM (C,E,F). **, P<0.01; ***, P<0.001. DAPI, 4',6-diamidino-2-phenylindole; FC, fold change; FISH, fluorescence in situ hybridization; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; LUSC, lung squamous cell carcinoma; N, normal tissue; RT-qPCR, reverse transcription quantitative polymerase chain reaction; SEM, standard error of the mean; T, tumor tissue.

CircITFG2 significantly inhibited the proliferation, migratory, and invasion of LUSC both in vitro and in vivo

To investigate the functional role of circITFG2 in LUSC, we generated stable cell lines overexpressing circITFG2. RT-qPCR confirmed its efficient overexpression without affecting host gene ITFG2 expression (Figure 2A). Functional assays revealed that circITFG2 overexpression significantly suppressed colony formation in H1703 cells (Figure 2B) and impaired migratory and invasive capacities in Transwell assays (Figure 2C). Additionally, TUNEL assays indicated enhanced apoptosis upon circITFG2 overexpression (Figure 2D). These findings—demonstrating circITFG2’s antiproliferative, antimigratory, and proapoptotic effects—were further corroborated in HCC1588 cells (Figure S2).

Figure 2 Overexpression of circITFG2 suppressed clonal proliferation, migration, and invasion while inducing apoptosis in H1703 cells. (A) RT-qPCR analysis of circITFG2 and its linear counterpart expression. Relative expression levels were normalized to the values of the vector group. (B) Colony formation assay. Cells were stained with crystal violet. (C) Transwell migration and invasion assays. Cells were stained with crystal violet staining. Scale bar =20 µm. (D) TUNEL apoptosis assay, with the red fluorescence indicating apoptotic cells. Scale bar =20 µm. Data are presented as the mean ± SEM (A-D). **, P<0.01. DAPI, 4',6-diamidino-2-phenylindole; RT-qPCR, reverse transcription quantitative polymerase chain reaction; SEM, standard error of the mean; TUNEL, terminal deoxynucleotidyl transferase dUTP nick end labeling.

To evaluate the in vivo function of circITFG2, H1703 cells stably overexpressing circITFG2 (oe-circITFG2) were subcutaneously injected into nude mice. Compared to controls, mice with oe-circITFG2 exhibited markedly suppressed tumor growth, as evidenced by reduced tumor size and weight (Figure 3A,3B). Consistent with this, immunohistochemical staining demonstrated a significant decrease in Ki-67-positive proliferative cells in oe-circITFG2 tumors (Figure 3C), further indicating its tumor-suppressive role.

Figure 3 Overexpression of circITFG2 inhibited tumor growth and proliferation of H1703 cells in vivo. BALB/C nude mice were subcutaneously injected with H1703 cells stably transfected with either the control vector or the circITFG2-overexpressing construct (n=5 per group). (A) Representative images of subcutaneous xenograft tumors on day 26. The black circle indicates the location and size of the tumor. (B) Excised subcutaneous tumors, along with tumor volume and weight measurements. (C) Representative IHC staining and quantitative analysis. Nuclei brown staining indicated Ki-67-positive cells. Compared to the oe-circITFG2 group, Vector tissue cells exhibited intense Ki-67 staining, indicated by strong nuclei brown staining. Scale bar =50 µm. The inset shows a 10× magnified view of the original image. Data are presented as the mean ± SEM (B,C). ***, P<0.001. IHC, immunohistochemistry; oe-circITFG2, overexpression-circITFG2; SEM, standard error of the mean.

Overexpression of circITFG2 targeted ITM2A and affected the tumor microenvironment of LUSC

RNA sequencing of HCC1588 cells overexpressing circITFG2 identified ITM2A as a key downstream target (Figure 4A; Figure S3A-S3E). Pathway enrichment analysis revealed significant association with immune-related processes, particularly lymphocyte activation and positive regulation of immune response (Figure 4B). Immunohistochemical analysis demonstrated an inverse correlation between ITM2A and PD-L1 expression in H1703 cells (Figure 4C), while bioinformatics analysis linked ITM2A to enhanced CD8⁺ T-cell infiltration (Figure S3F). Cross-referencing with immune databases identified 22 circITFG2-regulated immune-stimulatory genes (Figure 4D), which was supported by in vitro validation of circITFG2’s effect on CD8⁺ T-cell cytotoxicity (Figure 4E). To functionally validate these findings, we established LLC stable cell lines with four genetic modifications: oe-circITFG2, vector control, oe-circITFG2 + sh-ITM2A, and sh-ITM2A alone. Following subcutaneous injection into C57BL/6 mice (n=5/group), the oe-circITFG2 treatment showed maximal tumor suppression, while sh-ITM2A alone resulted in the largest tumors (Figure 4F,4G). Protein analysis revealed a positive correlation between circITFG2 expression and ITM2A, granzyme B, and perforin levels, whereas PD-L1 expression showed an inverse relationship (Figure 4H).

Figure 4 circITFG2 regulated PD-L1 expression by interacting with ITM2A in LUSC. (A) Volcano plot of differentially expressed genes (|log₂FC| >1, FDR <0.05) from RNA-seq analysis in circITFG2-overexpressing H1703 cells. (B) GO enrichment analysis of ITM2A-related DEGs revealed significant associations with immune regulatory pathways. (C) Immunohistochemistry confirmed that ITM2A overexpression attenuated PD-L1 membrane expression in LUSC cells. The positive reading indicates brown staining of the tumor cell membrane. (D) Intersection analysis of ImmPort, InnateDB, and RNA-seq data identified 22 stimulatory immune-related genes. (E) Western blotting demonstrated that circITFG2 overexpression, particularly when combined with ITM2A knockdown, reduced PD-L1 expression while enhancing cytotoxic effector molecules in CD8⁺ T-cell cocultures. (F,G) circITFG2 overexpression alone or with ITM2A silencing significantly inhibited H1703 tumor growth and (H) downregulated PD-L1 and cytotoxic effector proteins in vivo. ***, P<0.001. DEG, differentially expressed gene; FC, fold change; FDR, false discovery rate; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GO, Gene Ontology; IHC, immunohistochemistry; LUSC, lung squamous cell carcinoma; NC, the blank control group; PBMC, peripheral blood mononuclear cell; PD-L1, programmed death-ligand 1.

Overexpression of circITFG2 upregulated ITM2A expression by “sponging” hsa-miR-526b-5p and influenced certain immune indicators of LUSC

Bioinformatic analysis via the circRNAdb database indicated that circITFG2 lacks peptide-coding potential. Venn diagram analysis (Figure S4A) identified two shared miRNA binding sites between circITFG2 and ITM2A: hsa-miR-526b-5p and hsa-miR-583. Overexpression of circITFG2 significantly suppressed hsa-miR-526b-5p (Figure 5A), implicating its role in circITFG2-dependent ITM2A regulation. To validate this finding, we transfected cells with a hsa-miR-526b-5p mimic plasmid, which robustly increased its expression (Figure S4B). Bioinformatics analysis indicated that hsa-miR-526b-5p is markedly upregulated in LUSC (Figure S4C) and negatively correlates with ITM2A expression (Figure S4D). Elevated hsa-miR-526b-5p levels were also linked to poorer overall survival in patients with LUSC (Figure S4E). Given AGO2’s critical role in miRNA-mediated silencing, RIP assays in H1703 cells transfected with hsa-miR-526b-5p mimics revealed significant circITFG2 enrichment (Figure 5B), confirming direct interaction. Dual-luciferase assays with wild type (WT) and mutant (Mut) circITFG2 3’UTR constructs further validated the specificity of miR-526b-5p binding. Transfection with hsa-miR-526b-5p mimics reduced luciferase activity in the circITFG2 (WT) group. (Figure 5C), an effect reversed by circITFG2 overexpression (Figure S4F). In cocultures with activated CD8⁺ T cells, hsa-miR-526b-5p transfection reversed the effects of circITFG2 modulation on granzyme B and perforin expression (Figure 5D). Collectively, these results demonstrate that hsa-miR-526b-5p mediates circITFG2’s regulation of ITM2A and modulates antitumor immune responses.

Figure 5 Overexpression of circITFG2 sponged hsa-miR-526b-5p to regulate PD-L1 and cytotoxic effector molecules. (A) RT-qPCR analysis of two candidate miRNAs shared between oe-circITFG2 and vector in LUSC cell lines. (B) RIP assay confirmed the direct binding of miR-526b-5p to AGO2, indicating its role in miRNA-mediated silencing. (C) Dual-luciferase reporter assays comparing wild type circITFG2 (circITFG2-WT) and its 3'UTR mutant (circITFG2-MUT) revealed significantly reduced luciferase activity in miR-526b-5p mimic-transfected cells, confirming that miR-526b-5p directly targets circITFG2. (D) Western blot demonstrated that miR-526b-5p mimics counteracted circITFG2 overexpression-induced changes in PD-L1 and cytotoxic effector molecules (granzyme B and perforin), further supporting their regulatory interplay in immune modulation. *, P<0.05; **, P<0.01. AGO2, argonaute 2; IgG, immunoglobulin G; IP, immunoprecipitation; LUSC, lung squamous cell carcinoma; MUT, mutant; NC, the blank control group; PBMC, peripheral blood mononuclear cell; PD-L1, programmed death-ligand 1; qRT-PCR, quantitative real-time polymerase chain reaction; WT, wild type.

Overexpression of circITFG2 disrupted the stability of PD-L1 protein through the autophagic flux mediated by ITM2A

Previous studies have demonstrated that PD-L1 facilitates tumor immune evasion and enhances aggressiveness, serving as a clinically used predictive biomarker for antitumor immune response efficacy. Additionally, quantitative analysis revealed that circITFG2 overexpression significantly reduced PD-L1 levels via ITM2A upregulation (Figure 4E). We will further investigate how the oe-circITFG2/ITM2A axis regulates PD-L1 protein stability. Molecular docking simulations (Figure 6A) and immunofluorescence colocalization studies (Figure 6B) demonstrated direct ITM2A-PD-L1 interaction at the cytoplasmic membrane, establishing a structural framework for their functional relationship. To delineate the degradation pathway, we performed cycloheximide (CHX) chase assays in H1703 cells, which indicated that ITM2A overexpression markedly accelerated PD-L1 turnover (Figure 6C). Pharmacological inhibition experiments confirmed lysosomal degradation as the predominant mechanism, with BafA1 treatment effectively rescuing PD-L1 levels (Figure 6D)—consistent with established autophagy-mediated PD-L1 regulation. Further investigation demonstrated that circITFG2 upregulated key autophagy markers (LC3II/I, P62, and LAMP1), effects attenuated by ITM2A knockdown (Figure 6E). Bafilomycin A1 (BafA1) treatment revealed dynamic autophagic flux alterations, with ITM2A knockdown antagonizing oe-circITFG2 induced autophagy activation (Figure 6F). Ultrastructural analysis by transmission electron microscopy provided direct visual evidence of increased autophagosome formation following ITM2A overexpression (Figure 6G). Together, these results preliminarily suggest that circITFG2 overexpression promotes PD-L1 degradation through ITM2A-mediated autophagy activation, establishing a mechanistic connection between the circITFG2-ITM2A axis and autophagic PD-L1 turnover.

Figure 6 ITM2A overexpression promoted PD-L1 degradation via the autophagy pathway. (A) Molecular docking prediction of PD-L1-ITM2A interaction. The X-ray crystal structure of PD-L1 (4Z18) was obtained from the Protein Data Bank, while the ITM2A structure was predicted via AlphaFold. Protein-protein interaction analysis was performed via PyMOL, with ITM2A depicted as a deep blue model and PD-L1 as a cyan model. Potential binding sites are shown as stick models in their respective colors, with detailed views highlighting key interaction residues. (B) Representative immunofluorescence images demonstrating colocalization of ITM2A (red) and PD-L1 (green) at the plasma membrane in H1703 cells. Scale bar =10 µm. (C) Western blot analysis of PD-L1 protein levels in control and ITM2A-overexpressing H1703 cells treated with CHX for 0, 4, and 8 hours. (D) Western blot showing PD-L1 expression in ITM2A-overexpressing H1703 cells pretreated with CHX followed by DMSO, MG132 (proteasome inhibitor), or BafA1 (lysosome inhibitor). Western blot analysis of autophagy-related proteins (LC3II/I, P62, and LAMP1) in cells overexpressing circITFG2 alone or combined with ITM2A knockdown, either untreated (E) or treated with BafA1 (F). (G) Confocal microscopy images showing autophagosomes (red puncta) and autolysosomes (green puncta) in H1703 cells with ITM2A overexpression as compared to controls. Cells were stained with MDC and LysoTracker Red. Scale bar =10 µm. BafA1, bafilomycin A1; CHX, cycloheximide; DAPI, 4',6-diamidino-2-phenylindole; DMSO, dimethyl sulfoxide; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; MDC, monodansylcadaverine; PD-L1, programmed death-ligand 1.

Discussion

LUSC is among the most aggressive subtypes of NSCLC. A comprehensive understanding of its molecular regulatory network is essential for the development of more effective therapeutic strategies. circRNAs play a pivotal role in the progression of lung cancer, with their aberrant expression being closely linked to tumor heterogeneity, microenvironment remodeling, and treatment resistance. This necessitates systematic exploration. Research into circRNAs have provided significant breakthroughs in the identification of novel diagnostic markers and precise therapeutic targets, becoming a focal point of research. By integrating analyses of clinical samples, in vitro cellular experiments, and in vivo animal models, this study systematically elucidated the role of a newly discovered circRNA, circITFG2, identified through sequencing of LUSC tissue specimens. circITFG2 was implicated in the inhibition of cancer cell proliferation, migration, and invasion. Its tumor-suppressive effect was found to be primarily mediated through the overexpression of circITFG2, which acts as a molecular sponge for hsa-miR-526b-5p, thereby alleviating the miRNA-induced transcriptional repression of the target gene ITM2A, leading to direct inhibition of tumor proliferation. Concurrently, it enhances the therapeutic potential against LUSC.

This study evaluated the functional mechanisms of circITFG2 using a complementary experimental model system: demonstrating its cell-autonomous tumor suppression in immunodeficient mice and revealing its ability to synergistically modulate tumor immune microenvironment to suppress tumor growth in immunocompetent models. It is noteworthy that although these two mechanisms of action (cell autonomy and immune regulation) exhibit different phenotypic characteristics, they both depend on activation of the oe-circITFG2/ITM2A signaling axis, reflecting the unity of the molecular pathway. Given the biological necessity for immune-related phenotypes to be validated in the immune system, humanized mouse models can be used to further validate their clinical translational value in the future.

Tumor progression involves cancer cell growth, disrupted apoptosis, angiogenesis, migration, immune evasion, and treatment resistance, all of which are linked to clinical and pathological features (23,24). circRNAs are crucial in tumor pathology and immune regulation, offering potential therapeutic targets. Xu et al. found that lowering circRNA_0000392 expression inhibits colorectal cancer growth (25), while Liu et al. showed that targeting circCABIN1 increases glioblastoma sensitivity to temozolomide (26). This study confirmed that higher circITFG2 expression correlates with earlier LUSC T stages. circITFG2’s anticancer properties suggest that it could be a prognostic biomarker for LUSC. Furthermore, through in vitro and in vivo experiments, it has been confirmed that circITFG2 has anticancer functions, indicating that circITFG2 may become a potential biomarker for predicting the prognostic risk of patients with LUSC (27,28). circRNAs, characterized by their stable covalently closed circular structure, are abundant in plasma, saliva, and exosomes and may contribute to the prediction of cancer and other diseases. Although our study did not include the detection of circITFG2 in patient fluids, subsequent research will explore its potential utility as a prognostic biomarker for LUSC.

The circITFG2 identified in this study shares similarities with previously reported LUSC-associated circRNAs but also exhibits significant differences. Contrary to the pro-cancer circRNA (such as circTP63 and circTIMELESS) reported by Cheng et al. (29) and Zhang et al. (30), circITFG2 has a down-regulation trend in LUSC and exhibits the function of inhibiting tumor malignant proliferation and invasion after over-expression. This unique expression pattern and anti-cancer characteristics make it a more specific negative marker in the diagnosis of LUSC. At the mechanistic level, unlike the immunosuppressive HMGB2-circRNA found by Zhang et al. (31), overexpression of circITFG2 activates autophagy-mediated PD-L1 degradation via ITM2A, promoting immunopositive regulation. As a potential biomarker, circITFG2 down-regulation profile negatively correlates with tumor progression and may be more clinically distinguishable than upregulated circRNA alone, but its low abundance profile challenges detection sensitivity. In terms of therapeutic development, overexpression of circITFG2 in combination with immune checkpoint inhibitors may has potential synergistic effects, but the in vivo delivery efficiency and stability of circRNA are still key bottlenecks to be overcome in translational medicine.

Cytoplasmic circRNAs typically function as “sponges” for miRNAs, freeing the mRNAs they target and acting as ceRNAs (32). Our study found that circITFG2 is primarily located in the cytoplasm and plays a role in regulating ITM2A expression via hsa-miR-526b-5p, which is overexpressed in LUSC and linked to poor overall survival. This aligns with previous findings that hsa-miR-526b-5p promotes tumor progression in liver cancer and NSCLC (33).Since miRNAs are mainly derived from tumor, nerve, immune, and epithelial cells and are secreted extracellularly, they could serve as biomarkers for disease prognosis (34). Incorporating the miRNAs present in systemic circulation into disease monitoring protocols and use them as biomarkers for disease prognosis may have significant clinical implications. Additionally, this underscores the necessity for further investigation into the secretion and clinical relevance of hsa-miR-526b-5p in the body fluids of patients with LUSC.

The immune microenvironment characterized by excessive expression of immunoregulatory factors, diminished immune cell infiltration, dysregulated MHC molecule presentation, and various other immune evasion mechanisms, collectively drives tumor immune escape and presents substantial clinical therapeutic hurdles (35-38). Studies have shown that exogenous circRNAs can serve as vaccine adjuvants, enhancing the efficacy of vaccines and inducing antigen-specific T-cell activation (39), and ITM2A is also involved in the differentiation and maturation of T cells (21). Our study initially examined circITFG2 overexpression significantly upregulates ITM2A expression, consequently enhancing the activation of cytotoxic effector molecules in CD8⁺ T lymphocytes (40). Building upon the well-established role of PD-L1 overexpression in facilitating immune evasion and its clinical relevance in immunotherapy (41), along with the recognized involvement of autophagy in tumorigenesis and treatment responses (42), we made the intriguing observation that circITFG2 overexpression modulates autophagic flux to regulate PD-L1 protein stability and expression levels, thereby exerting synergistic antitumor effects. Although this mechanism is not exclusive to circITFG2-mediated tumor suppression, our findings nonetheless provide valuable novel insights into the immune evasion mechanisms in LUSC. Further investigations employing pharmacological and genetic approaches are warranted to fully elucidate the underlying molecular mechanisms.

This study was subject to several limitations that warrant careful consideration. First, while the expression of circITFG2 demonstrated a significant correlation with the characteristics of LUSC tumors, the absence of large-sample in vivo validation may restrict the generalizability of the findings. Second, the relatively small sample sizes in certain instances may undermine the robustness of the conclusions, particularly in the context of LUSC heterogeneity. It should be noted that our current research primarily focused on elucidating the functional value of circITFG2 in LUSC, while its potential biological roles in other lung cancer subtypes remain to be explored in future investigations. Furthermore, as circITFG2 is a newly identified gene currently absent from major genomics databases, the lack of clinical studies examining its relevance significantly constrains our comprehensive evaluation of its translational potential as a therapeutic target in LUSC, particularly regarding prognostic implications. These limitations underscore the necessity of expanding the patient cohort in validation studies to more precisely ascertain the clinical value of circITFG2.

Conclusions

Our investigation commenced with an examination of the recently identified circITFG2. Upon overexpression, circITFG2 enhanced the expression of ITM2A by sequestering hsa-miR-526b-5p. This mechanism resulted in the suppression of proliferation and migration in LUSC and facilitated the more dynamic remodeling of the immune microenvironment associated with LUSC. These findings offer novel insights into the pathological progression mechanisms of LUSC and contribute to the identification of therapeutic targets.

Acknowledgments

We sincerely thank Dr. Haitao Li for generously providing the plasmids. We are also grateful to the AME team for their valuable assistance in language editing and manuscript formatting.

Footnote

Reporting Checklist: The authors have completed the ARRIVE and MDAR reporting checklists. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1632/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1632/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1632/prf

Funding: This work was supported by the Guizhou Provincial Health and Family Planning Commission (No. gzwkj2023-283), the Technology and Science Bureau of Zunyi (No. 2024-13), Medical Research Union Found for High-quality Health Development of Guizhou Province (No. 2024GZYXKYJJXM0031) and Zhuhai City’s 2023 Industry-University-Research Cooperation and Basic and Applied Basic Research Projects (No. ZHKJ2024-01).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1632/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was approved by the Ethics Committee of The Third Affiliated Hospital of Zunyi Medical University (No. 2020-1-323) and informed consent was taken from all the patients. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Animal experiments were performed under a project license (No. 2022ZH0025) granted by the Fifth Affiliated Hospital (Zhuhai) of Zunyi Medical University, in compliance with national guidelines for the care and use of animals.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Siegel RL, Kratzer TB, Giaquinto AN, et al. Cancer statistics, 2025. CA Cancer J Clin 2025;75:10-45. [Crossref] [PubMed]
2. Zhang Y, Vaccarella S, Morgan E, et al. Global variations in lung cancer incidence by histological subtype in 2020: a population-based study. Lancet Oncol 2023;24:1206-18. [Crossref] [PubMed]
3. Lau SCM, Pan Y, Velcheti V, et al. Squamous cell lung cancer: Current landscape and future therapeutic options. Cancer Cell 2022;40:1279-93. [Crossref] [PubMed]
4. Gong M, Li Y, Ye X, et al. Loss-of-function mutations in KEAP1 drive lung cancer progression via KEAP1/NRF2 pathway activation. Cell Commun Signal 2020;18:98. [Crossref] [PubMed]
5. Yuan H, Liu J, Zhang J. The Current Landscape of Immune Checkpoint Blockade in Metastatic Lung Squamous Cell Carcinoma. Molecules 2021;26:1392. [Crossref] [PubMed]
6. Shi X, Dong A, Jia X, et al. Integrated analysis of single-cell and bulk RNA-sequencing identifies a signature based on T-cell marker genes to predict prognosis and therapeutic response in lung squamous cell carcinoma. Front Immunol 2022;13:992990. [Crossref] [PubMed]
7. Nemeth K, Bayraktar R, Ferracin M, et al. Non-coding RNAs in disease: from mechanisms to therapeutics. Nat Rev Genet 2024;25:211-32. [Crossref] [PubMed]
8. John A, Almulla N, Elboughdiri N, et al. Non-coding RNAs in Cancer: Mechanistic insights and therapeutic implications. Pathol Res Pract 2025;266:155745. [Crossref] [PubMed]
9. Guo Y, Huang Q, Heng Y, et al. Circular RNAs in cancer. MedComm (2020) 2025;6:e70079.
10. Hu G, Zhai S, Yu S, et al. Circular RNA circRHOBTB3 is downregulated in hepatocellular carcinoma and suppresses cell proliferation by inhibiting miR-18a maturation. Infect Agent Cancer 2021;16:48. [Crossref] [PubMed]
11. Ouyang X, He Z, Fang H, et al. A protein encoded by circular ZNF609 RNA induces acute kidney injury by activating the AKT/mTOR-autophagy pathway. Mol Ther 2023;31:1722-38. [Crossref] [PubMed]
12. Wang C, Ju C, Du D, et al. CircNF1 modulates the progression and immune evasion of esophageal squamous cell carcinoma through dual regulation of PD-L1. Cell Mol Biol Lett 2025;30:37. [Crossref] [PubMed]
13. Liu X, Zhang Y, Zhou S, et al. Circular RNA: An emerging frontier in RNA therapeutic targets, RNA therapeutics, and mRNA vaccines. J Control Release 2022;348:84-94. [Crossref] [PubMed]
14. Dong Y, Xu T, Zhong S, et al. Circ_0076305 regulates cisplatin resistance of non-small cell lung cancer via positively modulating STAT3 by sponging miR-296-5p. Life Sci 2019;239:116984. [Crossref] [PubMed]
15. Liu Z, Wang T, She Y, et al. N(6)-methyladenosine-modified circIGF2BP3 inhibits CD8(+) T-cell responses to facilitate tumor immune evasion by promoting the deubiquitination of PD-L1 in non-small cell lung cancer. Mol Cancer 2021;20:105. [Crossref] [PubMed]
16. Hong W, Zhang Y, Ding J, et al. circHIPK3 Acts as Competing Endogenous RNA and Promotes Non-Small-Cell Lung Cancer Progression through the miR-107/BDNF Signaling Pathway. Biomed Res Int 2020;2020:6075902. [Crossref] [PubMed]
17. Wang X, Chan S, Dai L, et al. Identification of novel T cell proliferation patterns, potential biomarkers and therapeutic drugs in colorectal cancer. J Cancer 2024;15:1234-54. [Crossref] [PubMed]
18. Zeng X, Yang D, Zhang J, et al. Integrating machine learning, bioinformatics and experimental verification to identify a novel prognostic marker associated with tumor immune microenvironment in head and neck squamous carcinoma. Front Immunol 2024;15:1501486. [Crossref] [PubMed]
19. Zhang R, Xu T, Xia Y, et al. ITM2A as a Tumor Suppressor and Its Correlation With PD-L1 in Breast Cancer. Front Oncol 2020;10:581733. [Crossref] [PubMed]
20. Zhou C, Wang M, Yang J, et al. Integral membrane protein 2A inhibits cell growth in human breast cancer via enhancing autophagy induction. Cell Commun Signal 2019;17:105. [Crossref] [PubMed]
21. Tai TS, Pai SY, Ho IC. Itm2a, a target gene of GATA-3, plays a minimal role in regulating the development and function of T cells. PLoS One 2014;9:e96535. [Crossref] [PubMed]
22. Zhang Z, Li J, He T, et al. The competitive endogenous RNA regulatory network reveals potential prognostic biomarkers for overall survival in hepatocellular carcinoma. Cancer Sci 2019;110:2905-23. [Crossref] [PubMed]
23. Zhang L, Yu D. Exosomes in cancer development, metastasis, and immunity. Biochim Biophys Acta Rev Cancer 2019;1871:455-68. [Crossref] [PubMed]
24. Lohan-Codeço M, Barambo-Wagner ML, Nasciutti LE, et al. Molecular mechanisms associated with chemoresistance in esophageal cancer. Cell Mol Life Sci 2022;79:116. [Crossref] [PubMed]
25. Xu H, Liu Y, Cheng P, et al. CircRNA_0000392 promotes colorectal cancer progression through the miR-193a-5p/PIK3R3/AKT axis. J Exp Clin Cancer Res 2020;39:283. [Crossref] [PubMed]
26. Liu X, Guo Q, Gao G, et al. Exosome-transmitted circCABIN1 promotes temozolomide resistance in glioblastoma via sustaining ErbB downstream signaling. J Nanobiotechnology 2023;21:45. [Crossref] [PubMed]
27. Yuan X, Mao Y, Ou S. Diagnostic accuracy of circulating exosomal circRNAs in malignances: A meta-analysis and systematic review. Medicine (Baltimore) 2023;102:e33872. [Crossref] [PubMed]
28. Wu X, Shi M, Lian Y, et al. Exosomal circRNAs as promising liquid biopsy biomarkers for glioma. Front Immunol 2023;14:1039084. [Crossref] [PubMed]
29. Cheng Z, Yu C, Cui S, et al. circTP63 functions as a ceRNA to promote lung squamous cell carcinoma progression by upregulating FOXM1. Nat Commun 2019;10:3200. [Crossref] [PubMed]
30. Zhang W, Shi J, Cheng C, et al. CircTIMELESS regulates the proliferation and invasion of lung squamous cell carcinoma cells via the miR-136-5p/ROCK1 axis. J Cell Physiol 2020;235:5962-71. [Crossref] [PubMed]
31. Zhang LX, Gao J, Long X, et al. The circular RNA circHMGB2 drives immunosuppression and anti-PD-1 resistance in lung adenocarcinomas and squamous cell carcinomas via the miR-181a-5p/CARM1 axis. Mol Cancer 2022;21:110. [Crossref] [PubMed]
32. Yi J, Li B, Yin X, et al. CircMYBL2 facilitates hepatocellular carcinoma progression by regulating E2F1 expression. Oncol Res 2024;32:1129-39. [Crossref] [PubMed]
33. Wu S, Tang T, Zhou H, et al. LINC01343 targets miR-526b-5p to facilitate the development of hepatocellular carcinoma by upregulating ROBO1. Sci Rep 2023;13:17282. [Crossref] [PubMed]
34. Dos Santos AF, Fazeli G, Xavier da Silva TN, et al. Ferroptosis: mechanisms and implications for cancer development and therapy response. Trends Cell Biol 2023;33:1062-76. [Crossref] [PubMed]
35. Nagano T, Tachihara M, Nishimura Y. Molecular Mechanisms and Targeted Therapies Including Immunotherapy for Non-Small Cell Lung Cancer. Curr Cancer Drug Targets 2019;19:595-630. [Crossref] [PubMed]
36. Memon D, Schoenfeld AJ, Ye D, et al. Clinical and molecular features of acquired resistance to immunotherapy in non-small cell lung cancer. Cancer Cell 2024;42:209-224.e9. [Crossref] [PubMed]
37. Wang X, Bai H, Zhang J, et al. Genetic Intratumor Heterogeneity Remodels the Immune Microenvironment and Induces Immune Evasion in Brain Metastasis of Lung Cancer. J Thorac Oncol 2024;19:252-72. [Crossref] [PubMed]
38. Wang DR, Wu XL, Sun YL. Therapeutic targets and biomarkers of tumor immunotherapy: response versus non-response. Signal Transduct Target Ther 2022;7:331. [Crossref] [PubMed]
39. Xia P, Gao X, Shao L, et al. Corrigendum to <Down-regulation of RAC2 by small RANSOM RNA restrains the progression of osteosarcoma by suppressing the Wnt signaling pathway> Int J Biol Macromol 2023;224:1600. [Int J. Biol. Macromol. 2019 Sep 15;137:1221-1231. doi: 10.1016/j.ijbiomac.2019.07.016].
40. Li Q, Xie D, Yao L, et al. Combining autophagy and immune characterizations to predict prognosis and therapeutic response in lung adenocarcinoma. Front Immunol 2022;13:944378. [Crossref] [PubMed]
41. Jiang T, Xia Y, Li Y, et al. TRIM29 promotes antitumor immunity through enhancing IGF2BP1 ubiquitination and subsequent PD-L1 downregulation in gastric cancer. Cancer Lett 2024;581:216510. [Crossref] [PubMed]
42. Liu X, Yin M, Dong J, et al. Tubeimoside-1 induces TFEB-dependent lysosomal degradation of PD-L1 and promotes antitumor immunity by targeting mTOR. Acta Pharm Sin B 2021;11:3134-49. [Crossref] [PubMed]