Tumor Cell-Intrinsic Circular RNA CircFNDC3B Attenuates CD8+ T Cells Infiltration in Non-Small Cell Lung Cancer

doi:10.21203/rs.3.rs-4663620/v1

Download PDF

Article

Tumor Cell-Intrinsic Circular RNA CircFNDC3B Attenuates CD8+ T Cells Infiltration in Non-Small Cell Lung Cancer

https://doi.org/10.21203/rs.3.rs-4663620/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Tumor-infiltrating CD8⁺ T cells are related with a satisfactory immunotherapy efficacy and a survival benefit. However, the mechanisms underlying differential infiltration levels are largely unknown, especially the role of circular RNAs (circRNAs) involving in CD8⁺ T cells infiltration remain poorly understood. Here, we identify 9912 distinct circRNAs from four paired normal and tumorous tissues of non-small cell lung cancer (NSCLC). We discover that circFNDC3B, derived from exons 2 and 3 of the FNDC3B gene, is significantly upregulated in NSCLC. TFⅡ-I interacting with STAT1 could function as an activator of transcription, facilitating CXCL10 and CXCL11 expression. However, circFNDC3B could bind to TFⅡ-I forming an RNA-protein complexus, thereby competitively inhibited the interaction between TFⅡ-I and STAT1, decreased the level of CXCL10 and CXCL11, and restricted infiltration of CD8⁺ T cells eventually. Meanwhile, the mouse circFndc3b ortholog is also inversely correlated with the infiltration of CD8⁺ T cells, attenuating the PD-1 blockade therapy’s effectiveness. These results demonstrate a pivotal role of tumor cell-intrinsic circRNA in immune infiltration and validate circFNDC3B as a potential target to develop a better effective immunotherapy strategy of NSCLC.

Biological sciences/Cancer/Lung cancer/Non-small-cell lung cancer

Biological sciences/Immunology/Immune evasion

non-small cell lung cancer

circFNDC3B

TFⅡ-I

STAT1

CD8 + T cells

immunotherapy

One of the most common causes of cancer-related incidence and mortality worldwide is lung cancer. About 85% of lung cancer cases are classified as non-small-cell lung cancer (NSCLC), mainly comprising of lung squamous cell carcinoma (LUSC) and lung adenocarcinoma (LUAD)¹. The 5-year survival rate for overall NSCLC is 26%, and only 8% for metastatic NSCLC^2,3. Immunotherapy has become the standard treatment for metastatic cancer, since the FDA approval of the first immune checkpoint inhibitor for NSCLC in 2015^4,5. However, only 27–46% of advanced NSCLC patients respond to immunotherapy⁶. Furthermore, 65% of the patients who were responded to immunotherapy, will progress within four years, eventually⁷. Further studies on the molecular and immunological mechanisms of NSCLC are required to develop novel therapeutic targets and enhanced the therapeutic effect of immunotherapy.

The CD8⁺ T cells are important defensive cells which can attack cancer cells in the tumor microenvironment (TME)^8,9. In NSCLC, high infiltration of CD8⁺ T cells predicts a better outcome overall, while the absence of CD8⁺ T cells in tumor environment makes a “immune-excluded” tumor environment and limits the effectiveness of immunotherapies^10,11. Moreover, chemokines, such as CXCL10 and CXCL11, are a family of small protein ligands that coordinate the activation and chemotaxis of CD8⁺ T cells^12,13. Although various transcriptional regulatory mechanisms are known to affect the expression of chemokines, current knowledge regarding the regulation of these chemokines through non-coding RNAs (ncRNAs) is limited.

As a class of endogenous ncRNAs, the circular RNAs (circRNAs) form a covalently closed circular structure with the ligation of 5’RNA and 3’RNA¹⁴. Growing evidence have revealed that circRNAs take part in the initiation and progression of cancer by binding proteins or/and sponging miRNA^15,16. Notably, some circRNAs could play a role in the modulation of immune response in TME^17,18. CircRNA-002178 derived from lung cancer cells could increase PD-1 expression in CD8⁺ T cells via exosomes¹⁹. CircFAT1 and circMET inhibit CD8⁺ T cells infiltration in head and neck squamous cell carcinoma and hepatocellular carcinoma, respectively^20,21. Exosome-derived circCCAR1 contributes to immunosuppression by facilitating CD8⁺T-cell dysfunction and anti-PD1 resistance in hepatocellular carcinoma²². Nevertheless, most circRNAs need to be further investigated to understand how they contribute to the tumor immunity and tumorigenesis in NSCLC.

The biochemical properties of the signal-induced multifunctional transcription factor TFⅡ-I indicate that it is involved in a variety of gene regulatory processes^23,24. TFⅡ-I consists of a nuclear localization sequence (NLS), a basic region (BR) DNA-binding domain, and multiple protein–protein interaction domains, including a leucine zipper (LZ) and six helix-loop-helix (HLH)-like domains, also referred to as I-repeats (R1–R6), which suggests that it may have the capacity for multiple protein-protein and protein-DNA interactions²⁵. However, the function of TFⅡ-I in NSCLC still remained unknown. Here, we characterized the expression profile of circRNAs in paired tumorous tissues and normal adjacent lung tissues from NSCLC patients. We identified that circFNDC3B (derived from exons 2 and 3 of the FNDC3B gene), was highly expressed in NSCLC tumorous tissues. Our study established that TFⅡ-I interacting with STAT1 could function as an activator of transcription, facilitating CXCL10 and CXCL11 expression. However, circFNDC3B competitively inhibited the binding of TFⅡ-I and STAT1, decreased the expression and secretion of CXCL10 and CXCL11, attenuating the intratumoral infiltration of CD8⁺T cells and response to PD-1 checkpoint blockade. Taken together, these findings suggested that circFNDC3B/TFⅡ-I complex may serve as a potential target for the treatment of NSCLC.

Tissues and cell lines

The tumorous tissues (T) and paired adjacent normal tissues (N) were collected from NSCLC patients who underwent surgery at Department of Thoracic Surgery in Union Hospital of Tongji Medical College of Huazhong University of Science and Technology from July 2014 to December 2018. Written informed consent was obtained from all patients. Ethical approval was confirmed by Institutional Review Board of Tongji Medical College of Huazhong University of Science and Technology. All NSCLC tissues were snap-frozen in liquid nitrogen and then stored at -80℃ until use. All the surgical samples were confirmed by the pathologists using H&E-stained sections. The clinical characteristics of patients were presented in Table S1. All of the patients were followed up on a regular basis, and overall survival (OS) time was determined from the date of surgery to the date of death or the date of the last follow-up visit for survivors.

Human NSCLC cell lines A549 (RRID:CVCL_0023), NCI-H1299 (RRID:CVCL_0060), NCI-H226 (RRID:CVCL_1544), SPC-A1 (RRID:CVCL_6955), HBE (human bronchial epithelial cell line 1, HBE1(RRID:CVCL_0287)), and LLC (murine Lewis lung cancer cell line, LL/2 (LLC1) (RRID:CVCL_4358)) were purchased from the American Type Culture Collection. H1299, H226, SPC-A1 and HBE cells were cultured in Roswell Park Memorial Institute 1640 (RPMI-1640) culture medium supplemented with 10% FBS. A549 cells were cultured in Ham's F-12K (Kaighn's) medium supplemented with 10% FBS (Gibco, Australia origin). LLC cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% FBS. All cells were cultured at 37 ℃ with 5% CO₂.

All cell lines have been authenticated in the past three years according to the manufacturer’s instructions. All experiments were performed with mycoplasma-free cells.

Identification and annotation of human NSCLC circRNAs

Total RNA from four pairs of NSCLC and adjacent normal lung tissues was isolated using the RNeasy Mini Kit (QIAGEN, Germany) according to the manufacturer’s instructions. The total RNA samples were treated with the RiboZero rRNA Removal Kit (Epicentre, WI, USA) for deleting rRNA. Next, RNA was digested with RNase R (Epicentre, Madison, WI, USA) to remove linear RNA and enrich for circular RNAs. The remaining RNAs were fragmented and synthesized cDNA with random primer. Then, the libraries were quality controlled and sequenced by Illumina NovaSeq 6000 (Illumina, San Diego, CA, USA). After that, circRNAs were detected and identified with the tools find_circ²⁶ and CIRI2²⁷. The DESeq R package (1.10.1) was used for differential expression analysis between NSCLC and normal lung tissues. The resulting P values were adjusted using the Benjamini and Hochberg’s approach for controlling the false discovery rate. The RNA-seq data was deposited in SRA database (Accession code: PRJNA863919).

RNA preparation, RNase R treatment and qRT-PCR

Total RNA was isolated from tumor tissue, normal tissue or cells by using RNeasy Mini Kit (QIAGEN, Germany). Nuclear and cytoplasmic RNA was extracted using nuclear and cytoplasmic RNA purification kit (Fisher scientific, AM1921). For RNase R treatment, one microgram of total RNA was incubated 15 min at 37°C with or without 3 U of RNase R (Epicentre Technologies, Madison, WI). Then, RNA was reverse-transcribed using the PrimeScript RT Master Mix (Takara, Dalian, China). SYBR Premix Ex Taq II (Takara) was used for real-time PCR. Particularly, the divergent primers annealing at the distal ends of circRNAs were used to determine the abundance of circRNAs. Genomic DNA (gDNA) was extracted using Genomic DNA Isolation Kit (Sangon Biotech, Shanghai, China). The cDNA and gDNA PCR products were observed using 2% agarose gel electrophoresis. The Amplification procedure was performed with the StepOne Plus Real-Time PCR system (Applied Biosystems, Foster City, CA). The Ct thresholds were determined by software. The primers are demonstrated in Table S2.

Actinomycin D treatment and RNA stability assay for RNA lifetime

Cells were placed into six-well plates for actinomycin D (Sigma-Aldrich, USA) experiments. Up to 60% confluency after 24 h, cells were treated with 5 µg/ml Actinomycin D or DMSO (Sigma-Aldrich, USA) and collected at indicated time points. The turnover rate and half-life of RNA was estimated according to a previously published paper²⁸.

RNA Fluorescent in situ hybridization (FISH)

Cy3-labelled circFNDC3B and circFndc3b probes (Table S2) were synthesized by TSINGKE (Wuhan, China). FISH was performed as described with minor modifications²⁹. The cancer cells were fixed and permeabilized, followed by hybridizing with the probes at 37 ℃ overnight in a dark. The coverslips were washed three times 2 × SSC (Solarbio, Beijing, China) for 10 min, before they were sealed with parafilm containing DAPI (Roche Diagnostic, Mannheim, Germany). The images were obtained with a confocal laser scanner (LSM 780, Carl Zeiss).

Cell counting Kit-8 (CCK-8) experiment

The proliferation of cells was measured by CCK-8 kit (Dojindo, Japan) following the manufacturer’s instructions. The optical density at 450 nm was measured using an automatic microplate reader (Synergy4; BioTek, Winooski, VT, USA).

Transwell matrigel invasion assay

The matrigel invasion assay was performed using the transwell precoated matrigel chamber following the manufacturer’s protocol (BD Science, Bedford, MA, USA). The homogeneous single cell suspensions (1×105 cells/well) were added to the upper chambers and incubated for 24 h. The invasion rates were quantified by counting the invaded cells at least three random fields.

Apoptosis assay

The cell apoptosis assay was detected according to the manufacturer’s instructions of FITC Annexin V Apoptosis Detection Kit I (BD Biosciences). The data was then analyzed using the FCS Express 5 software (De Novo Software, Los Angeles, CA).

Cell cycle assay

Cells were fixed in 75% ice-cold ethanol overnight at 4°C. After washing the fixed cells with PBS, propidium iodide (PI) buffer (BD Pharmingen, USA) was applied for staining. Cell cycle analysis was performed by flow cytometer. The data were analyzed using ModFit LT 5.0.

RNA sequencing

Total RNA was isolated from circFNDC3B-overexpressed A549 and H226 cells and the corresponding control cells by using RNeasy Mini Kit (QIAGEN, Germany). Library preparations were conducted using NEBNext® Ultra™ RNA Library Prep Kit for Illumina® (New England Biolabs). The gene expression profiles were determined by Illumina high-throughput RNA sequencing (Novogene, Beijing, China). Significant differentially expressed transcripts were screened by p < 0.05 and log2(fold change) ≥ 1.

Enzyme linked immunosorbent assay (ELISA)

The expression levels of CXCL10 and CXCL11 in the cell culture supernatants were determined with the commercially available ELISA kits from (Biolegend and Raybiotech) following the instructions provided by the manufacturer.

RNA pull-down assays

Biotin-labeled oligonucleotide probes targeting junction site of circRNA (sense) and control (antisense) probes (Table S2) were synthesized by TSINGKE (Wuhan, China). For RNA pull-down assay, 1 × 10⁷ cells were washed in ice-cold phosphate-buffered saline, lysed in 500 µl Co-IP buffer (20mM Tris-HCL, pH 7.5, 150mM NaCl, 1mM EDTA, 0.5% NP-40, and complete protease inhibitors cocktail and RNase inhibitors), and then incubated with 3 µg biotin-labelled circFNDC3B (sense) and control (antisense) probes for 2 h at room temperature. A total of 50 µl washed Streptavidin C1 magnetic beads (Invitrogen) were added to each binding reaction and incubated at room temperature for another hour. After washing with the Co-IP buffer for five times, the proteins bounding to the beads were eluted, and analyzed by mass spectrometry or western blotting.

Silver staining and mass spectrometry analysis

Silver staining was performed using the PAGE Gel Silver Staining Kit (Solarbio, Beijing, China) based on the protocol described. Retrieved proteins were detected by mass spectrometry analysis at Novogene Company (Beijing, China). Afterwards, protein identification and quantification were implemented with Proteome Discoverer software (version 1.4; Thermo Fisher Scientific, USA).

Western blot and Co-immunoprecipitation (Co-IP)

For western blot, whole cell lysates were prepared using RIPA buffer containing protease inhibitors (Beyotime). The total protein lysates from cells were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and transferred onto polyvinylidene difluoride membranes. After blocking with 5% non-fat milk, membranes were incubated with primary antibody overnight at 4°C, followed by an incubation with HRP conjugated secondary antibodies 1 h at room temperature. Immunoreactive bands were detected by using the enhanced chemiluminescence (E412–01, Vazyme, Nanjing, China).

For co-immunoprecipitation, cells were lysed in Co-IP buffer containing protease inhibitor cocktail for 40 min on ice. Cell lysates were incubated with the corresponding antibodies adsorbed to protein A/G Agarose (Thermo Fisher Scientific, #20421) for 4 h at 4°C. Afterwards, the cell lysates were washed three times in Co-IP buffer and eluted at 95°C for 10 min.

Antibodies used included primary antibodies against TFⅡ-I (Santa Cruz Biotechnology, sc-136330, 1:500), AGO2 (Thermo Fisher Scientific, MA5-23515, 1:500), STAT1 (Abcam, ab239360, 1:1000), pSTAT1 (Abcam, ab109461, 1:1000), Flag (Abcam, ab205606, 1:1000), GAPDH (Proteintech, 60004–1-Ig, 1:50000), Histone-H3 (Proteintech, 17168-1-AP, 1:1000). HRP-conjugated secondary goat anti-mouse (Proteintech, SA00001–1, 1:2000), HRP-conjugated secondary goat anti-rabbit (Proteintech, SA00001–2, 1:2000), HRP-mouse anti-rabbit IgG heavy chain specific (Proteintech, SA00001-7H, 1:5000) and HRP-mouse anti-rabbit IgG light chain specific (Proteintech, SA00001-7L, 1:4000) antibodies were used as second antibodies.

RNA Immunoprecipitation (RIP)

RIP experiment was carried out with the Magna RIP RNA-Binding Protein Immunoprecipitation Kit (Millipore, Bedford, MA) according to the manufacturer’s instructions. Cell lysates were incubated with dynabeads coated with AGO2 (Thermo Fisher Scientific, MA5-23515, 8µg), TFⅡ-I (Santa Cruz Biotechnology, sc-136330, 1µg), Flag (Abcam, ab125243, 1µg) or Mouse IgG (Santa Cruz Biotechnology, sc-2025, 1µg) antibody at 4°C overnight, respectively. The co-precipitated RNAs were abstracted using RNeasy MinElute Cleanup Kit (Qiagen), and detected by qPCR.

Immunocytofluorescence

NSCLC cells were grown on the coverslips and fixed with 4% paraformaldehyde. After permeabilizing with 0.1% TritonX-100, cells were washed with PBS, blocked with 5% BSA and incubated with TFⅡ-I (Santa Cruz Biotechnology, sc-136330, 1:500) antibody overnight. Subsequently, cells were washed with PBS again, following by incubating with fluorophore-labeled secondary antibody to mouse IgG (Abcam, ab150113, 1:500). Finally, parafilm containing DAPI was applied for the sealing of cells. Fluorescence images were obtained with a confocal laser scanner (LSM 780, Carl Zeiss).

Vector construction and cell transfection

To establish circFNDC3B, circFndc3b, TFⅡ-I and STAT1 over-expression plasmids, circFNDC3B, circFndc3b, TFⅡ-I and STAT1 cDNA were synthesized by TSINGKE (Wuhan, China). Then they were cloned into pcDNA3.1(+) CircRNA Mini Vector (addgene #60648) and p3XFLAG-CMV-10 vector (Sigma-Aldrich), respectively. The truncations of TFⅡ-I were constructed as region containing 102-399aa (3×Flag_102–399), 400-681aa (3×Flag_400–681), 682-972aa (3×Flag_682–972), 102-681aa (3×Flag_102–681) and 400-972aa (3×Flag_400–972). The Flag-tagged TFⅡ-I truncations were subcloned into p3XFLAG-CMV-10 vector. Oligonucleotides encoding short hairpin RNAs (shRNAs, Table S2) specific for circFNDC3B and STAT1 were cloned into pLKO.1-puro (Sigma-Aldrich).

Transfection was performed using Lipofectamine 2000 (Life Technologies) according to the manufacturer’s instructions. Stable cell lines were screened by administration of neomycin or puromycin (Invitrogen). Scramble shRNA and empty vector were used as controls.

Chromatin immunoprecipitation and qPCR (ChIP-qPCR) assays

The procedure for ChIP experiment was performed with the EZ-Magna ChIP A/G Chromatin Immunoprecipitation kit (Sigma-Aldrich). The DNA fragments were resuspended in PBS and incubated with anti-STAT1 (Abcam, ab239360, 5µg) or Rabbit IgG (Abcam, ab172730, 5µg) at 4°C overnight. The precipitated DNA-chromatin products were determined by qPCR. The data were demonstrated as fold enrichment (normalized to IgG). The primer sequences utilized in ChIP-qPCR were showed in Table S2.

Dual-luciferase reporter assay

A549 cells were co-transfected with a reporter constructs for CXCL10 or CXCL11 promoter (-1000 to 300 bp) in pGL3-Basic vector, and a Renilla luciferase reporter vectors pRL-TK as the internal control. The firefly and Renilla luciferase activities were quantified using Dual-Luciferase® Reporter Assay System (Promega, USA) according to the manufacturer’s protocol, after transfection for 24 h.

Purification of human CD8⁺ T cells

Ficoll-Hypaque gradient centrifugation was utilized for the isolation of mononuclear cells from healthy volunteers’ peripheral blood. Purified CD8⁺ T cells were isolated with the human CD8⁺ T cells isolation kit (Miltenyi Biotec, Teterow, Germany) according to the manufacturer’s protocol. The purity of CD8⁺ T cells was > 90%, as confirmed by flow cytometry. Purified CD8⁺ T cells were cultured in RPMI-1640 medium with a 10% FBS concentration. The medium contained recombinant human IL-2 (Abcam, ab119439, 10 ng/ml), soluble anti-human CD28 mAbs (Biolegend, 302937, 1 µg/ml) and plate-bound anti-human CD3 (Biolegend, 317301, 1 µg/ml) mAb. The cells were then cultured for three consecutive days for the subsequent experiments.

Chemotaxis assay

Preactivated CD8⁺ T cells were resuspended in RPMI 1640 medium with 0.5% FBS, and placed in the upper chamber of Transwell tissue culture inserts (5.0-µm pore diameter) (Corning, New York, USA). Supernatant from A549 cells were added in the lower chamber. RPMI-1640 medium alone as a negative control. When indicated, anti-CXCL11 (R&D Systems, MAB672, 1µg/ml), anti-CXCL10 (R&D Systems, MAB266, 1µg/ml), recombinant human CXCL11 (R&D Systems, 672-IT, 50ng/ml), recombinant human CXCL10 (R&D Systems, 266-IP, 50ng/ml) were added to the lower chamber. The chambers were incubated at 37°C, 5% CO₂ for 4 h. Transmigrated cells in the lower chambers were collected and stained with anti-human CD8 mAbs (Biolegend, 980908). After staining, precision count beads (Biolegend, 424902) were added, and the numbers of transmigrated CD8⁺ T cells were determined using flow cytometry (FACS Canto II; BD Bioscience). The chemotactic index was calculated as the ratio of the number of CD8⁺ T cells that migrated to corresponding wells divided by the number of CD8⁺ T cells that migrated to RPMI-1640 medium alone.

In vivo tumorigenesis assays

The NOD-SCID mice or C57BL/6 female mice (4 weeks old) were obtained from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). The mice were bred on a chow diet in a 12-h light/12-h dark environment at 25°C in the Tongji Medical College Animal Care Facility under high health status conditions. All animal studies were approved by the Animal Care and Utilization Committee of Tongji Medical College, Huazhong University of Science and Technology.

For NOD-SCID mice experiments, circFNDC3B-overexpressing and control A549 cells (2 × 10⁶ cells) were subcutaneously injected in the right and left flank of the same NOD-SCID mice, respectively (n = 6). For some experiments, purified human CD8⁺ T cells (1 × 10⁷ per mice) were injected via the tail vein of NOD-SCID mice (n = 6), beginning a week before inoculation of tumor cells, weekly. After four weeks, mice were euthanatized, the tumor were dissected. The volumes of the dissected tumors were measured using calipers and calculated with the following formula: a² × b × 0.5 (a indicates the smallest diameter and b represents the diameter perpendicular to a).

For C57BL/6 mice experiments, circFndc3b-overexpressing and control LLC cells (5 × 10⁵ cells) were subcutaneously injected in the right flank of C57BL/6 mice (n = 6 for each group). After two weeks, mice were euthanatized, the tumor were dissected. The volumes of the dissected tumors were determined as mentioned above.

For anti-PD1 antibody treatment, circFndc3b-overexpressing and control LLC cells (5 × 10⁵ cells) were subcutaneously injected in the right flank of C57BL/6 mice (n = 6 for each group) on day 0. Anti-mouse PD-1 monoclonal antibody (Biolegend, 135202) or its isotype control (Biolegend, 400501) were injected into C57BL/6 mice through the tail vein with at 100 µg per dose every four days for four times, beginning on day 6. The growth of the tumor was determined with calipers every 3 days. The volumes of tumor were measured as mentioned above. When tumors reached a maximum of 2000mm³, the mice were euthanized.

CD8⁺ T cells infiltrating analysis by flow cytometry and immunohistofluorescence

Dissected tumors were treated with DNase I (Sigma, 50 µg/ml) and collagenase type IV (Gibco, 200 U/ml) for 20 minutes at 37°C. To obtain a single cell suspension, cells were passed through a 70-µm filter to remove clumps, diluted in medium. Then cells were surface-stained with anti-mouse CD45 mAbs (Biolegend, 147708), anti-mouse CD3 mAbs (Biolegend, 100206) and anti-mouse CD8 mAbs (Biolegend, 126606) for flow cytometry.

Immunohistofluorescence was performed on formalin-fixed tumor tissues. Five-micrometer paraffin-embedded sections were deparaffinized and rehydrated in graded alcohol series. Staining for human CD8 (Abcam, ab237710, 1/100) or mice CD8 (Abcam, ab217344, 1:500) was carried out with non-labeled primary antibodies and fluorophore-labeled secondary antibodies (Thermo Fisher Scientific, A-11008, 1:1000).

Then the slices were counterstained with DAPI in mounting medium. The images were obtained with a confocal laser scanning microscope (LSM 780, Carl Zeiss) and analyzed using ImageJ software.

Statistical analysis

All analyses of the data in this study were performed using SPSS 26.0 software, and the results were presented as mean ± standard deviation. Two tailed Student’s t-test was carried out between two groups and one-way ANOVA was applied in multiple groups. Overall survival was estimated by the Kaplan-Meier method and compared by the log-rank test. P < 0.05 was considered statistically significant.

Profiling of circRNAs in paired normal and tumorous tissues of NSCLC

To study the circRNAs’ expression patterns in NSCLC, we performed circRNA sequencing on ribosomal RNA- deprived total RNA from four pairs of human normal and tumorous tissues of NSCLC. The omnipresent circRNAs can be found in various genomic regions (Fig. 1A). In total, 9 912 distinct circRNAs had been found in the samples, and 9696 of them had at least four back-splicer reads (Fig. 1B). Most circRNAs composed of protein coding exons, while smaller portions aligned to intergenic regions and introns of the genome (Fig. S1A). Exonic circRNAs had a median length around 300 nucleotides, and most of them had a length which was < a thousand nucleotides (Fig. S1B). We also discovered that one gene was able to produce several circRNAs (Fig. S1C). Moreover, most (82.3%) of identified exonic circRNAs consisted of 2–4 exons (Fig. S1D). Compared to the databases previously published obtained from circBase³⁰, we discovered that there were 3340 novel circRNAs and 6 572 overlapped circRNAs identified in this study. (Fig. S1E). Additionally, we found that 793 circRNAs were significantly dysregulated in NSCLC tissues, of which 678 circRNAs were downregulated and 115 circRNAs were upregulated (Fig. 1C and S1F). Therefore, our findings indicate that circRNAs, a large family of RNA, may participate in NSCLC progression.

Characterization of circFNDC3B in NSCLC

Among the dysregulated circRNAs, we noticed that circFNDC3B, which came from exons 2 and 3 of the FNDC3B gene, was the most significantly upregulated circRNA in NSCLC compared to normal adjacent lung tissues (Fig. 1C and S1F, Table S3 and S4, read count 175.61 vs 55.94, log₂fold change = 1.6181, p = 0.0000000111). It was confirmed further with a larger NSCLC patient cohort (Fig. 1D). The high level of circFNDC3B was also identified in human lung cancer cells A549, H1299, H226 and SPC-A1, compared to human bronchial epithelial cells (HBE) (Fig. S1A). Furthermore, Kaplan-Meier curves demonstrated that high expression of circFNDC3B were related with a poorer overall survival (OS) (Fig. 1E). Utilized divergent primers, the head-to-tail splice junction of circFNDC3B was magnified and identified with Sanger sequencing (Fig. 1F). The divergent and convergent primers was utilized to amplify the head-to-tail splice site of circFNDC3B, along with gDNA (genomic DNA) and cDNA as templates. GAPDH was utilized as a control. As expected, a PCR product was amplified with the divergent primers of circFNDC3B but not GAPDH (Fig. 1G and S2B). In addition, circFNDC3B transcripts were more stable upon treatment with RNase R exonuclease or actinomycin D, as compared with FNDC3B mRNA (Fig. 1H, I and S2C), indicating that circFNDC3B harbored a circular RNA structure. Subsequently, the cytoplasmic/nuclear fractionation and FISH detection confirmed that circFNDC3B mainly distributed in the nucleus of NSCLC cells (Fig. 1J, K and S2D). In brief, these findings display that circFNDC3B, a bona fide circRNA, is significantly upregulated and is suggestive of functionality in NSCLC.

CircFNDC3B has no effect on the invasion, apoptosis, cell cycle and proliferation of NSCLC cells

The veracity of the overexpression and knockdown systems for circFNDC3B expression was evaluated to explore the role of circFNDC3B in the progression of NSCLC. Enforced expression of circFNDC3B did not affect the expression of mFNDC3B (Fig. S3A). Meanwhile, sh-circFNDC3B#2 could target circFNDC3B, but it cannot specifically target mFNDC3B (Fig. S3B). Unexpectedly, circFNDC3B expression did not influence the invasion, apoptosis, cell cycle, and cell viability of NSCLC cells in vitro (Fig. S3C-I). We further sought to examine the effect of circFNDC3B in vivo performing xenograft studies in NOD-SCID mice. Consistent with the in vitro findings, tumor size was not affected by circFNDC3B (Fig. S3J, K).

CircFNDC3B downregulates the level of CXCL10 and CXCL11 in NSCLC cells

Considering the lack of lymphocytes in NOD-SCID mice, we wondered if circFNDC3B could drive tumor immune microenvironment shifts. To verify our assumptions, the purified CD8⁺ T cells from healthy volunteers’ peripheral blood, were activated with recombinant human IL-2, soluble anti-human CD28 mAbs and plate-bound anti-human CD3 mAb for three days. Then the preactivated CD8⁺T cells were administered intravenously into tumor-bearing NOD-SCID mice. Surprisingly, the tumor growth in NOD-SCID mice was accelerated by circFNDC3B when the mice were injected with CD8⁺ T cells (Fig. 2A). Thus, our results reveal that tumor cell-intrinsic circFNDC3B could accelerate the development of NSCLC by affecting CD8⁺ T cells rather than by directly influencing the biological behavior of tumor cells.

To gain a deep understanding of how circFNDC3B reshaped the tumor's immune microenvironment, we performed a transcriptomic analysis. In circFNDC3B-overexpressed A549 cells, there were 255 downregulated genes and 156 upregulated genes, compared with the corresponding control cells. In circFNDC3B-overexpressed H226 cells, there were 161 downregulated genes and 101 upregulated genes, compared with the corresponding control cells (Fig. 2B-D, log₂ fold change < 1, p < 0.05). Venn diagram showed that 17 downregulated and 3 upregulated overlapping genes were both significantly dysregulated in both A549 cells and H226 cells, after they were overexpressed with circFNDC3B (Fig. 2C, D).

We noted that CXCL10 and CXCL11, two important chemokines for CD8⁺ T cells recruitment¹³, were simultaneously downregulated after ectopic circFNDC3B expression (Fig. 2E, Table S5,S6, CXCL10 read count in circFNDC3B-overexpressed A549 cells vs control A549 cells: 4.94 vs 33.38, log₂fold change = -2.72, p = 0.000372; CXCL11 read count in circFNDC3B-overexpressed A549 cells vs control A549 cells: 1.98 vs 17.20, log₂fold change = -3.04, p = 0.003515; CXCL10 read count in circFNDC3B-overexpressed H226 cells vs control H226 cells: 233.16 vs 763.73, log₂fold change= -1.71, p = 0.000081; CXCL11 read count in circFNDC3B-overexpressed H226 cells vs control H226 cells: 370.19 vs 1128.49, log₂fold change= -1.61, p = 0.000175). We further confirmed transcriptomic analysis via qRT-PCR detection and ELISA. Consistently, enforced expression of circFNDC3B significantly reduced the expression and production of CXCL10 and CXCL11, while silencing circFNDC3B affected CXCL10 and CXCL11 oppositely (Fig. 2F, G). Meanwhile, the chemotaxis of CD8⁺ T cells towards the supernatant of NSCLC cells was impaired upon circFNDC3B overexpression, and was enhanced after circFNDC3B knockdown (Fig. 2H). Collectively, tumor cell-intrinsic circFNDC3B reduces the chemotaxis of CD8⁺ T cells towards tumor by downregulating the level of CXCL10 and CXCL11, implying a physiological role for circFNDC3B in immune infiltration.

CircFNDC3B interacts with TFⅡ-I protein in NSCLC cells

As miRNA sponge was one of the pivotal functions of circRNAs, we explored if circFNDC3B could bind to miRNAs, even though only a small amount of circFNDC3B localized in the cytoplasm. Reciprocal immunoprecipitation (RIP) assay showed that circRNA CDR1as, but not circFNDC3B and U6, could bind to AGO2(Fig. 3A). Our findings ruled out the possibility of circFNDC3B sequestering homologous miRNAs, considering that AGO2 was one of the important constituent elements of the RNA-induced silencing complex (RISC)³¹.

We conducted an RNA pulldown assay to probe into the biological function of circFNDC3B in the nucleus. The biotin-labeled probes targeting the circFNDC3B back-spliced sequence were utilized to screen the potential proteins interacting with circFNDC3B. A specific band was detected on silver-stained SDS-PAGE gel (Fig. 3B). Following the mass spectrometry (MS) and analysis pipeline (Table S7 and Fig. S4A, B), TFⅡ-I protein was identified as circFNDC3B-binding protein (Fig. 3C), while no significant MS peaks of AGO2 protein were detected, further suggesting that circFNDC3B did not act as miRNA sponge. Indeed, TFⅡ-I was selectively apprehended by sense but not antisense probes targeting circFNDC3B (Fig. 3D). Using purified TFⅡ-I protein, we validated the specific interaction between circFNDC3B and TFⅡ-I with RIP assay (Fig. 3E). Using immunofluorescence staining and FISH assay, the colocalization of circFNDC3B and TFⅡ-I in NSCLC cells was confirmed further (Fig. 3F). To portray the structural core elements in the interactions between circFNDC3B and TFⅡ-I, the TFⅡ-I domains were subdivided as shown in the deletion mapping (Fig. 3G). Our observations demonstrated that the aa102-399 domain of TFⅡ-I contributed to its association with circFNDC3B, whereas deletion of other domains of TFⅡ-I had almost no effect on its interaction with circFNDC3B (Fig. 3H, I). Moreover, we noticed that high-expression of GTF2I (gene symbol of TFⅡ-I protein) patients had a longer survival and GTF2I was significantly downregulated in NSCLC tissues compared to normal adjacent lung tissues (Fig. S4C, D). Thus, these findings propose that circFNDC3B/TFⅡ-I utilized the aa102-399 domain of TFⅡ-I to form an RNA-protein complex in the nucleus of NSCLC cells.

CircFNDC3B competitively inhibited the binding between TFⅡ-I and STAT1 in NSCLC cells

Previous reports have showed that TFⅡ-I could form a protein-protein complex with STAT1 in murine fibroblasts in vivo³², independently validated by our findings in human NSCLC cells (Fig. 4A). Alternatively, the overexpression of TFⅡ-I did not change the overall expression of STAT1, but enhanced the phosphorylation of STAT1(Fig. 4B). Analyses of available ChIP-seq data demonstrated that STAT1 peaks were abundant on the CXCL10 and CXCL11 promoters (Fig. S5A, B). Supporting the results, ChIP-qPCR experiments revealed that STAT1 specifically occupied on the promoters of CXCL10 and CXCL11 (Fig. 4C, D). Strikingly, we found a high co-binding localization between TFⅡ-I and STAT1 occupancies on the promoters of CXCL10 and CXCL11 (Fig. S5C, D). To further elucidate the influence of TFⅡ-I and STAT1 in the transcriptional activity of the CXCL10 and CXCL11 promoters, we conducted dual-luciferase reporter experiment. We demonstrated that the overexpression of TFⅡ-I could augment the luciferase activity of CXCL10 and CXCL11 promoters, whereas the knockdown of STAT1 could reverse the effect of enhanced TFⅡ-I expression (Fig. 4E and S5E, F). Consistently, enforced TFⅡ-I expression increased the mRNA level of CXCL10 and CXCL11, while silencing STAT1 was able to rescue the phenomenon (Fig. S5G). Collectively, TFⅡ-I interacting with STAT1 could function as an activator of transcription, facilitating CXCL10 and CXCL11 expression.

Considering that circFNDC3B formed a stable RNA-protein complex with TFⅡ-I, we wondered whether circFNDC3B/TFⅡ-I complex might affect the stability of TFⅡ-I /STAT1 complex. We observed that circFNDC3B overexpression weakened the interaction of TFⅡ-I and STAT1(Fig. 4F). Of note, while total amount of STAT1 was not altered, STAT1 phosphorylation was significantly reduced upon circFNDC3B overexpression, and the decreased STAT1 phosphorylation was restored after enforced TFⅡ-I expression (Fig. 4G). Then we investigated whether circFNDC3B coordinated with TFⅡ-I and STAT1 to modulate the transcription of CXCL10 and CXCL11. It was found that the decreased transcriptional activity and expression of CXCL10 and CXCL11 upon overexpression of circFNDC3B was reversed after enforced TFⅡ-I expression (Fig. 4H; Fig. S5H). In addition, STAT1 overexpression rescued circFNDC3B-mediated the reduction of transcriptional activity and expression of CXCL10 and CXCL11 (Fig. 4I and S5I). Taken together, these findings illustrate that circFNDC3B competes with STAT1 for binding to TFⅡ-I, thereby inhibiting the transcriptional activity of CXCL10 and CXCL11.

Tumor cell-intrinsic circFNDC3B restricts infiltration of CD8⁺ T cells

Previous studies verified that CXCL10 and CXCL11 lead CD8⁺ T cells migrating towards tumor microenvironment³³. Based on Tumor Immune Estimation Resource (TIMER) database, CD8⁺ T cell immune infiltration is positively related with the expression of CXCL10 and CXCL11, both in LUAD and LUSC (Fig. 5A). Remarkably, the CD8⁺ T cells’ infiltration was upregulated in NSCLC tissues with low level of circFNDC3B, as compared to NSCLC tissues with high level of circFNDC3B (Fig. 5B, C). In vitro chemotaxis assays demonstrated that overexpression of circFNDC3B weakened the chemotaxis of CD8⁺ T cells towards the supernatant of A549 cells, which would be restored by the addition of rh-CXCL10 and rh-CXCL11 (Fig. 5D). Conversely, knockdown of circFNDC3B strengthened the chemoattractant activity of CD8⁺ T cells, which would be reversed by anti-CXCL10 or anti-CXCL11 (Fig. 5E). Furthermore, both of the overexpression TFⅡ-I and STAT1 could reverse the inhibition of circFNDC3B on chemotaxis of CD8⁺ T cells (Fig. 5F, G). Hence, we propose that tumor cell-intrinsic circFNDC3B attenuates infiltration of CD8⁺ T cells towards NSCLC microenvironment by inhibiting CXCL10 and CXCL11.

IFN-γ facilitates the secretion of CXCL10 and CXCL11 in NSCLC cells

The main source of secreted IFN-γ was the activated CD8⁺ T cells³⁴. The biological function of IFN-γ is educed by cell-surface receptor (Interferon gamma receptor, IFNGR), leading to activated STAT1 pathway and antitumor immunity^34,35. Consistent with the prior study³³, we confirmed that IFN-γ upregulated the level of CXCL10 and CXCL11 (Fig. S6A, B). Dual-luciferase reporter experiments displayed that IFN-γ facilitated the transcriptional activity of CXCL10 and CXCL11 promoters, which could be reversed when STAT1 was silenced (Fig. S6C). Moreover, IFN-γ reversed the inhibition effect of circFNDC3B on CXCL10 and CXCL11(Fig. S6D). Meanwhile, the silence of circFNDC3B increased the level of CXCL10 and CXCL11, which could be augmented with the addition of IFN-γ (Fig. S6E). Thus, we speculated that cancer cells with low circFNDC3B increased the secretion of CXCL10 and CXCL11, facilitating CD8⁺ T cells infiltration. Additionally, IFN-γ produced by CD8⁺ T cells promoted cancer cells to secrete CXCL10 and CXCL11. Therefore, our results hint that IFN-γ/CXCL10-CXCL11 positive feedback loop may amplify the anti-tumor immune response in NSCLC with low circFNDC3B, whereas high circFNDC3B terminates the loop.

CircFndc3b ortholog accelerates tumor progression and weakens the effect of anti-PD1 therapy in vivo

Strikingly, human circFNDC3B and mice circFndc3b ortholog were highly conserved (Fig. S7A), indicating that mice circFndc3b might have similar function with human circFNDC3B. FISH detection demonstrated that circFndc3b ortholog was predominantly localized in the murine Lewis lung carcinoma (LLC) cells’ nucleus (Fig. 6A). The interaction between circFndc3b and TFⅡ-I in LLC cells was validated through RNA pulldown assay with biotin-labled probes targeting the circFndc3b back-spliced sequence (Fig. 6B) and RIP analysis (Fig. 6C). We further confirmed that the overexpression of circFndc3b did not influence the expression of STAT1, but enhanced the phosphorylation of STAT1 (Fig. S7B). As expected, TFⅡ-I could form protein-protein complex with STAT1 in LLC cells in vivo (Fig. 6D). Enforced expression of circFndc3b ortholog hindered the binding between of TFⅡ-I and STAT1 in LLC cells, indicating that circFndc3b/TFⅡ-I complex disrupted the interaction of TFⅡ-I and STAT1(Fig. S7C). Furthermore, the overexpression of TFⅡ-I reversed the reduced phosphorylation of STAT1 caused by the overexpression of circFndc3b (Fig. S7D). Notably, circFndc3b significantly suppressed the level of CXCL10 and CXCL11 in LLC cells (Fig. 6E).

To examine the significance of circFndc3b ortholog in tumorigenesis, subcutaneous administration of LLC cells stably containing either circFndc3b or a control vector was performed on C57BL/6 mice. Compared to the control group, LLC cells overexpressing circFndc3b recipient mice showed a more prominent tumor progression (Fig. 6F). Consistent with these biological effects, immunofluorescence staining and flow cytometry of CD8 confirmed that the overexpression of circFndc3b reduced the infiltration of CD8⁺ T cells in LLC tumor tissues (Fig. 6G, H). Due to the clinical trial results highlighting the potential of immunotherapy for patients with NSCLC³⁶, we tested if circFndc3b ortholog could impede anti-PD1 immunotherapy. Indeed, enforced expression of circFndc3b dramatically decreased the effect of anti-PD1 therapy on tumor regression (Fig. 6I). In brief, circFndc3b ortholog inhibited CD8⁺ T cells recruitment towards tumor microenvironment, accelerated tumor progression and weakened the immunotherapy effect.

Limited infiltration of CD8⁺ T cells in tumors skews the immune responses toward a suppressive phenotype^9,37. Understanding the mechanisms by which the tumors hamper CD8⁺ T cell infiltration is fundamental to find out potential effective therapeutic drugs to enhance the cancer immunotherapy’s efficacy in the clinic. In this study, we identified tumor cell-intrinsic circFNDC3B as a negative modulator of CD8⁺ T cell infiltration in NSCLC. CircFNDC3B could competitively inhibit the binding of TFⅡ-I and STAT1 via circFNDC3B/TFⅡ-I complex, thereby repressing STAT1-mediated chemokines CXCL10 and CXCL11 transcription and causing a reduction in tumor-infiltrating CD8⁺ T cell. In turn, decreased IFN-γ secretion contributed by reduced CD8⁺ T cells infiltration restricted the activation of STAT1 in NSCLC cells, further impairing expression and secretion of CXCL10 and CXCL11 (Fig. S8). Additionally, mouse NSCLC cells expressed an ortholog of this circRNA, and we demonstrated that the mouse circFndc3b ortholog was also inversely correlated with CD8⁺ T cell infiltration, leading to insensitive to PD-1 blockade therapy. These results revealed circFNDC3B/TFⅡ-I complex as attractive cancer immunotherapy targets, and an important link between tumor cell-intrinsic circRNA and molding of tumor-infiltrating immune cells in NSCLC.

Previous studies had established that circRNAs are not cellular byproducts, but rather discrete functional entities¹⁴. To gain a comprehensive expression of circRNAs in NSCLC, we conducted circRNAs profiling of four pairs of tumor and adjacent normal lung tissues. Consequently, this study provides a portrait of circRNAs in NSCLC and expands our knowledge of the complexity of the NSCLC transcriptome, presenting an immense potential for future exploration. Furthermore, this study displayed that circFNDC3B was significantly upregulated in NSCLC and associated with a poor prognosis. Previous studies demonstrated the multiple function of circFNDC3B in different tumor and other nonneoplastic diseases, including gastric cancer, esophageal cancer, colorectal cancer, renal carcinoma, oral squamous cell carcinoma and bladder cancer³⁸. circRNAs may function as miRNA sponges by sequestering homologous miRNAs¹⁵. Specifically, circ0001361, which is identical to circFNDC3B, sequesters miR-491-5p, promoting bladder cancer cell metastasis and invasion through MMP9³⁹. However, our results provided no evidence that circFNDC3B derived from NSCLC affected invasion in vitro. Moreover, we found that circFNDC3B failed to function as a miRNA sponge in NSCLC. On the basis of anti-AGO2 RIP and mass spectrometry following circFNDC3B pulldown, the rationale for our findings was that AGO2, a component of RNA-induced silencing complex (RISC), did not interact with circFNDC3B. Rather than being a miRNA sponge, circFNDC3B bound physically to TFⅡ-I in NSCLC, corroborating the function of binding proteins of circRNAs¹⁴. Notably, mouse circFndc3b ortholog interacts with FUS to modulate the expression of VEGF in mouse heart³¹, while human circFNDC3B and FUS were unable to form a stable interaction in our model. We speculate that circFNDC3B may play its biological roles in tissue- or cell-specific context. Our study reveals high expression of circFNDC3B as a valuable feature of NSCLC progression and uncovers pathways regulated by circFNDC3B, comprehensive studies based on independent experiments and larger patient cohorts are required to further fully illuminate the function and clinical implications of circFNDC3B.

The main structure of TFⅡ-I comprises of a nuclear localization sequence (NLS), a basic region (BR) DNA-binding domain, and multiple protein–protein interaction domains, including a leucine zipper (LZ) and six helix-loop-helix (HLH)-like domains, also referred to as I-repeats (R1–R6), which suggests that it may have the capacity for multiple protein-protein and protein-DNA interaction^23,25. NLS and BR were located between R1 and R2. The peptide domain aa102-399 of TFⅡ-I contained domain between R1 and R2, which make this domain could bind with DNA and protein. The peptide domain aa400-681 and aa682-972, contained R3-R4 and R5-R6, respectively, which make them has a role of binding protein and could not bind with DNA. Our result indicated that the peptide domain aa102-399 of TFⅡ-I is indispensable for interaction with circFNDC3B despite there is none obvious RNA-binding domain. The detail mechanism underlying the interaction requires further investigation. Our observations provide novel functional insights into TFⅡ-I domains and their implication in RNA binding activity. Given potential HLH-like domain characterizing protein-protein interaction module, TFⅡ-I is considered as an unusual transcription factor which could simultaneously act both as a basal factor and as a repressor or activator of transcription through recruiting either co-repressors or co-activators²⁵. In addition, the genome-wide association analysis reveals the dual function of TFⅡ-I as transcriptional repressor or activator due to the fact that there was no explicit correlation between TFⅡ-I binding sites in the genome and genes expression^40,41. In agreement with prior reports³², we showed that TFⅡ-I could bind to STAT1 in NSCLC, collaborating to promote formation of gene-specific activator complexes and facilitate CXCL10 and CXCL11 transcription. Our study further uncovered circFNDC3B as a critical partner that competitively inhibited the interaction of TFⅡ-I and STAT1, highlighting a new regulatory mechanism of TFⅡ-I. Further studies are needed to investigate the genome-wide target genes of TFⅡ-I in NSCLC and the precise signal pathway by which TFⅡ-I modulates transcriptional partners association and dissociation.

Recently, chemotherapy plus PD-1/PD-L1 blockade immunotherapy has been treated as a standard therapeutic strategy for NSCLC^36,42. Interaction between PD-L1(expressed on cancer cells) and PD-1(expressed on effector T cells) induced inhibitory signal, led to effector T cell exhaustion and caused tumor immune escape. PD-1/PD-L1 blockade immunotherapy could prevent the inhibition of T-cell function, reactivated T-cell mediated antitumor immune response²⁴. However, less than 50% of the advanced NSCLC patients respond to immunotherapy, and most responders eventually develop progressive disease⁶. Further study is needed to improve the efficacy of immunotherapy. Tumor-infiltrating CD8⁺ T cells, as a key effector of anti-tumor immunity, are related with the immunotherapy efficacy and a survival benefit⁴³. In our study, we found that circFNDC3B/TFⅡ-I complex could prevent CD8⁺ T cell infiltration by inhibiting STAT1 activation in NSCLC, highlighting the importance of circRNAs in the modulation of tumor immunity. Indeed, recent studies suggest that circRNAs have the ability to recruit immune cells towards tumor microenvironment and regulate immune responses^18,21. By further in vivo functional studies, we demonstrated that circFNDC3B hindered the efficacy of PD-1 blockade immunotherapy. Thus, there is considerable interest in developing the next generation of circRNA-based immune therapies to enhance CD8⁺ T cell responses in NSCLC patients.

In summary, our work reveals a previously unrecognized mechanism by which tumor cell-intrinsic circRNA attenuates CD8⁺ T cells infiltration in NSCLC. We broaden the knowledge of circRNAs’ effect in NSCLC. The present study demonstrated that circFNDC3B/TFⅡ-I complex, as a potent suppressor of STAT1 activation, may have promising translational potential in the development of effective immunotherapies of NSCLC.

NSCLC non-small-cell lung cancer

LUSC lung squamous cell carcinoma

LUAD lung adenocarcinoma

TME tumor microenvironment

ncRNAs non-coding RNAs

circRNAs circular RNAs

HLH helix-loop-helix

TFⅡ-I transcription factor Ⅱ-I

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Author contributions

XW, XX, and ZW performed the design of all primers, western blot assay, circFNDC3B pull-down experiments, and data analysis; SX, WP, LY, YQ, LC, BY, and SZ accomplished most flow cytometry experiments, FISH assay, CCK8 assay, transwell migration matrigel invasion assay, apoptosis assay, cell cycle assay, and animal experiments; QX, HW, and JA implemented the data interpretation and analysis (e.g., statistical analysis, and bioinformatics analysis); XW, XX, ZW, SX, WP, LY, KJ and QZ fulfilled the manuscript's writing, review, and/or revision; XW, XX, ZW, KJ and QZ prepared material, technical, and administrative support. Having read and approved the final manuscript, all authors have given their approval.

Acknowledgements

Not applicable.

Funding sources and disclosure of conflicts of interest

This work was supported by the National Natural Science Foundation of China (grant numbers 82100111, 81973990, 81900096, 82170105, 81770090, 82200113) and Chen Xiao-ping Foundation for the Development of Science and Technology of Hubei Province CXPJJH122001-2203]. The funders did not take part in data analysis, manuscript writing and publication decision. The authors declare that there is no conflict of interest regarding the publication of this article.

Ethics approval and consent to participate

In the present study, the Institutional Review Board at Tongji Medical College of Huazhong University of Science and Technology has approved the research. All study participants provided informed consent. All authors consent to the publication of the manuscript.

Siegel, R. L., Miller, K. D., Wagle, N. S. & Jemal, A. Cancer statistics, 2023. CA Cancer J Clin 73, 17-48, doi:10.3322/caac.21763 (2023).
Thai, A. A., Solomon, B. J., Sequist, L. V., Gainor, J. F. & Heist, R. S. Lung cancer. Lancet (London, England) 398, 535-554, doi:10.1016/s0140-6736(21)00312-3 (2021).
Bai, Y. et al. Immunotherapy for advanced non-small cell lung cancer with negative programmed death-ligand 1 expression: a literature review. Translational lung cancer research 13, 398-422, doi:10.21037/tlcr-23-144 (2024).
Kazandjian, D. et al. FDA Approval Summary: Nivolumab for the Treatment of Metastatic Non-Small Cell Lung Cancer With Progression On or After Platinum-Based Chemotherapy. The oncologist 21, 634-642, doi:10.1634/theoncologist.2015-0507 (2016).
Ettinger, D. S. et al. NCCN Guidelines® Insights: Non-Small Cell Lung Cancer, Version 2.2023. Journal of the National Comprehensive Cancer Network : JNCCN 21, 340-350, doi:10.6004/jnccn.2023.0020 (2023).
Livanou, M. E., Nikolaidou, V., Skouras, V., Fiste, O. & Kotteas, E. Efficacy of NSCLC Rechallenge with Immune Checkpoint Inhibitors following Disease Progression or Relapse. Cancers 16, doi:10.3390/cancers16061196 (2024).
Antonia, S. J. et al. Durvalumab after Chemoradiotherapy in Stage III Non-Small-Cell Lung Cancer. The New England journal of medicine 377, 1919-1929, doi:10.1056/NEJMoa1709937 (2017).
van der Leun, A. M., Thommen, D. S. & Schumacher, T. N. CD8(+) T cell states in human cancer: insights from single-cell analysis. Nature reviews. Cancer 20, 218-232, doi:10.1038/s41568-019-0235-4 (2020).
St Paul, M. & Ohashi, P. S. The Roles of CD8(+) T Cell Subsets in Antitumor Immunity. Trends in cell biology 30, 695-704, doi:10.1016/j.tcb.2020.06.003 (2020).
Gueguen, P. et al. Contribution of resident and circulating precursors to tumor-infiltrating CD8(+) T cell populations in lung cancer. Sci Immunol 6, doi:10.1126/sciimmunol.abd5778 (2021).
Giles, J. R., Globig, A. M., Kaech, S. M. & Wherry, E. J. CD8(+) T cells in the cancer-immunity cycle. Immunity 56, 2231-2253, doi:10.1016/j.immuni.2023.09.005 (2023).
Wang, X. et al. The role of CXCR3 and its ligands in cancer. Frontiers in oncology 12, 1022688, doi:10.3389/fonc.2022.1022688 (2022).
Karin, N. CXCR3 Ligands in Cancer and Autoimmunity, Chemoattraction of Effector T Cells, and Beyond. Frontiers in immunology 11, 976, doi:10.3389/fimmu.2020.00976 (2020).
Liu, C. X. & Chen, L. L. Circular RNAs: Characterization, cellular roles, and applications. Cell 185, 2016-2034, doi:10.1016/j.cell.2022.04.021 (2022).
Kristensen, L. S., Jakobsen, T., Hager, H. & Kjems, J. The emerging roles of circRNAs in cancer and oncology. Nat Rev Clin Oncol 19, 188-206, doi:10.1038/s41571-021-00585-y (2022).
He, A. T., Liu, J., Li, F. & Yang, B. B. Targeting circular RNAs as a therapeutic approach: current strategies and challenges. Signal transduction and targeted therapy 6, 185, doi:10.1038/s41392-021-00569-5 (2021).
Zhang, Q. et al. Roles of circRNAs in the tumour microenvironment. Mol Cancer 19, 14, doi:10.1186/s12943-019-1125-9 (2020).
Li, B. et al. circNDUFB2 inhibits non-small cell lung cancer progression via destabilizing IGF2BPs and activating anti-tumor immunity. Nat Commun 12, 295, doi:10.1038/s41467-020-20527-z (2021).
Wang, J. et al. circRNA-002178 act as a ceRNA to promote PDL1/PD1 expression in lung adenocarcinoma. Cell Death Dis 11, 32, doi:10.1038/s41419-020-2230-9 (2020).
Huang, X. Y. et al. Circular RNA circMET drives immunosuppression and anti-PD1 therapy resistance in hepatocellular carcinoma via the miR-30-5p/snail/DPP4 axis. Mol Cancer 19, 92, doi:10.1186/s12943-020-01213-6 (2020).
Jia, L., Wang, Y. & Wang, C. Y. circFAT1 Promotes Cancer Stemness and Immune Evasion by Promoting STAT3 Activation. Adv Sci (Weinh) 8, 2003376, doi:10.1002/advs.202003376 (2021).
Hu, Z. et al. Exosome-derived circCCAR1 promotes CD8 + T-cell dysfunction and anti-PD1 resistance in hepatocellular carcinoma. Molecular cancer 22, 55, doi:10.1186/s12943-023-01759-1 (2023).
Roy, A. L. Pathophysiology of TFII-I: Old Guard Wearing New Hats. Trends Mol Med 23, 501-511, doi:10.1016/j.molmed.2017.04.002 (2017).
Roy, A. L. Role of the multifunctional transcription factor TFII-I in DNA damage repair. DNA repair 106, 103175, doi:10.1016/j.dnarep.2021.103175 (2021).
Linzer, N. et al. Regulation of RNA Polymerase II Transcription Initiation and Elongation by Transcription Factor TFII-I. Frontiers in molecular biosciences 8, 681550, doi:10.3389/fmolb.2021.681550 (2021).
Memczak, S. et al. Circular RNAs are a large class of animal RNAs with regulatory potency. Nature 495, 333-338, doi:10.1038/nature11928 (2013).
Gao, Y., Zhang, J. & Zhao, F. Circular RNA identification based on multiple seed matching. Brief Bioinform 19, 803-810, doi:10.1093/bib/bbx014 (2018).
Zhang, H. et al. CircLIFR synergizes with MSH2 to attenuate chemoresistance via MutSalpha/ATM-p73 axis in bladder cancer. Mol Cancer 20, 70, doi:10.1186/s12943-021-01360-4 (2021).
Zirkel, A. & Papantonis, A. Detecting Circular RNAs by RNA Fluorescence In Situ Hybridization. Methods Mol Biol 1724, 69-75, doi:10.1007/978-1-4939-7562-4_6 (2018).
Glazar, P., Papavasileiou, P. & Rajewsky, N. circBase: a database for circular RNAs. RNA 20, 1666-1670, doi:10.1261/rna.043687.113 (2014).
Garikipati, V. N. S. et al. Circular RNA CircFndc3b modulates cardiac repair after myocardial infarction via FUS/VEGF-A axis. Nat Commun 10, 4317, doi:10.1038/s41467-019-11777-7 (2019).
Kim, D. W., Cheriyath, V., Roy, A. L. & Cochran, B. H. TFII-I enhances activation of the c-fos promoter through interactions with upstream elements. Mol Cell Biol 18, 3310-3320, doi:10.1128/MCB.18.6.3310 (1998).
Tokunaga, R. et al. CXCL9, CXCL10, CXCL11/CXCR3 axis for immune activation - A target for novel cancer therapy. Cancer Treat Rev 63, 40-47, doi:10.1016/j.ctrv.2017.11.007 (2018).
Wang, W. et al. CD8(+) T cells regulate tumour ferroptosis during cancer immunotherapy. Nature 569, 270-274, doi:10.1038/s41586-019-1170-y (2019).
Spens, A. E., Sutliff, N. A., Bennett, S. R., Campbell, A. E. & Tapscott, S. J. Human DUX4 and mouse Dux interact with STAT1 and broadly inhibit interferon-stimulated gene induction. eLife 12, doi:10.7554/eLife.82057 (2023).
Reck, M., Remon, J. & Hellmann, M. D. First-Line Immunotherapy for Non-Small-Cell Lung Cancer. J Clin Oncol 40, 586-597, doi:10.1200/JCO.21.01497 (2022).
Anandappa, A. J., Wu, C. J. & Ott, P. A. Directing Traffic: How to Effectively Drive T Cells into Tumors. Cancer Discov 10, 185-197, doi:10.1158/2159-8290.CD-19-0790 (2020).
Sun, K. et al. Emerging landscape of circFNDC3B and its role in human malignancies. Frontiers in oncology 13, 1097956, doi:10.3389/fonc.2023.1097956 (2023).
Liu, F. et al. Hsa_circ_0001361 promotes bladder cancer invasion and metastasis through miR-491-5p/MMP9 axis. Oncogene 39, 1696-1709, doi:10.1038/s41388-019-1092-z (2020).
Fan, A. X. et al. Genomic and proteomic analysis of transcription factor TFII-I reveals insight into the response to cellular stress. Nucleic Acids Res 42, 7625-7641, doi:10.1093/nar/gku467 (2014).
Liang, S. et al. CBX4 Regulates Long-Form Thymic Stromal Lymphopoietin-mediated Airway Inflammation through SUMOylation in House Dust Mite-induced Asthma. American journal of respiratory cell and molecular biology 66, 648-660, doi:10.1165/rcmb.2021-0301OC (2022).
Chaft, J. E., Shyr, Y., Sepesi, B. & Forde, P. M. Preoperative and Postoperative Systemic Therapy for Operable Non-Small-Cell Lung Cancer. J Clin Oncol 40, 546-555, doi:10.1200/JCO.21.01589 (2022).
Jansen, C. S. et al. An intra-tumoral niche maintains and differentiates stem-like CD8 T cells. Nature 576, 465-470, doi:10.1038/s41586-019-1836-5 (2019).

There is NO Competing Interest.

Download PDF

Version 1

posted

You are reading this latest preprint version

Tumor Cell-Intrinsic Circular RNA CircFNDC3B Attenuates CD8+ T Cells Infiltration in Non-Small Cell Lung Cancer

Status:

Version 1

Abstract

Figures

Introduction

Material and methods

Tissues and cell lines

Identification and annotation of human NSCLC circRNAs

RNA preparation, RNase R treatment and qRT-PCR

Actinomycin D treatment and RNA stability assay for RNA lifetime

RNA Fluorescent in situ hybridization (FISH)

Cell counting Kit-8 (CCK-8) experiment

Transwell matrigel invasion assay

Apoptosis assay

Cell cycle assay

RNA sequencing

Enzyme linked immunosorbent assay (ELISA)

RNA pull-down assays

Silver staining and mass spectrometry analysis

Western blot and Co-immunoprecipitation (Co-IP)

RNA Immunoprecipitation (RIP)

Immunocytofluorescence

Vector construction and cell transfection

Chromatin immunoprecipitation and qPCR (ChIP-qPCR) assays

Dual-luciferase reporter assay

Purification of human CD8+ T cells

Chemotaxis assay

CD8+ T cells infiltrating analysis by flow cytometry and immunohistofluorescence

Statistical analysis

Results

Profiling of circRNAs in paired normal and tumorous tissues of NSCLC

Characterization of circFNDC3B in NSCLC

CircFNDC3B has no effect on the invasion, apoptosis, cell cycle and proliferation of NSCLC cells

CircFNDC3B downregulates the level of CXCL10 and CXCL11 in NSCLC cells

CircFNDC3B interacts with TFⅡ-I protein in NSCLC cells

CircFNDC3B competitively inhibited the binding between TFⅡ-I and STAT1 in NSCLC cells

Tumor cell-intrinsic circFNDC3B restricts infiltration of CD8+ T cells

IFN-γ facilitates the secretion of CXCL10 and CXCL11 in NSCLC cells

Discussion

Conclusion

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1

Purification of human CD8⁺ T cells

CD8⁺ T cells infiltrating analysis by flow cytometry and immunohistofluorescence

Tumor cell-intrinsic circFNDC3B restricts infiltration of CD8⁺ T cells