Skip to main content
Download PDF

Circular RNA-encoded oncogenic PIAS1 variant blocks immunogenic ferroptosis by modulating the balance between SUMOylation and phosphorylation of STAT1

Molecular Cancer volume 23, Article number: 207 (2024) Cite this article

Abstract

Background

The clinical response rate to immune checkpoint blockade (ICB) therapy in melanoma remains low, despite its widespread use. Circular non-coding RNAs (circRNAs) are known to play a crucial role in cancer progression and may be a key factor limiting the effectiveness of ICB treatment.

Methods

The circRNAs that were downregulated after coadministration compared with single administration of PD-1 inhibitor administration were identified through RNA-seq and Ribo-seq, and thus the circPIAS1 (mmu_circ_0015773 in mouse, has_circ_0008378 in human) with high protein coding potential was revealed. Fluorescence in situ hybridization (FISH) assays were conducted to determine the localization of circPIAS1 in human and mouse melanoma cells, as well as its presence in tumor and adjacent tissues of patients. Validation through dual-luciferase reporter assay and LC–MS/MS confirmed the ability of circPIAS1 to encode a novel 108 amino acid polypeptide (circPIAS1-108aa). Specific antisense oligonucleotides (ASOs) targeting the junction site of circPIAS1 were developed to reduce its intracellular levels. Proliferation changes in melanoma cells were assessed using CCK8, EdU, and colony formation assays. The impact of circPIAS1-108aa on the ferroptosis process of melanoma cells was studied through GSH, MDA, and C11-BODIPY staining assays. Western Blot, Immunoprecipitation (IP), and Immunoprecipitation-Mass Spectrometry (IP-MS) techniques were employed to investigate the impact of circPIAS1-108aa on the P-STAT1/SLC7A11/GPX4 signaling pathway, as well as its influence on the balance between STAT1 SUMOylation and phosphorylation. Additionally, a melanoma subcutaneous transplanted tumor mouse model was utilized to examine the combined effect of reducing circPIAS1 levels alongside PD-1 inhibitor.

Results

Compared with the group treated with PD-1 inhibitor alone, circPIAS1 was significantly down-regulated in the coadministration group and demonstrated higher protein coding potential. CircPIAS1, primarily localized in the nucleus, was notably upregulated in tumor tissues compared to adjacent tissues, where it plays a crucial role in promoting cancer cell proliferation. This circRNA can encode a unique polypeptide consisting of 108 amino acids, through which it exerts its cancer-promoting function and impedes the effectiveness of ICB therapy. Mechanistically, circPIAS1-108aa hinders STAT1 phosphorylation by recruiting SUMO E3 ligase Ranbp2 to enhance STAT1 SUMOylation, thereby reactivating the transduction of the SLC7A11/GPX4 signaling pathway and restricting the immunogenic ferroptosis induced by IFNγ. Furthermore, the combination of ASO-circPIAS1 with PD-1 inhibitor effectively inhibits melanoma growth and significantly enhances the efficacy of immune drugs in vivo.

Conclusions

Our study uncovers a novel mechanism regarding immune evasion in melanoma driven by a unique 108aa peptide encoded by circPIAS1 in melanoma that dramatically hinders immunogenic ferroptosis triggered by ICB therapy via modulating the balance between SUMOylation and phosphorylation of STAT1. This work reveals circPIAS1-108aa as a critical factor limiting the immunotherapeutic effects in melanoma and propose a promising strategy for improving ICB treatment outcomes.

Introduction

Melanoma, originating from the cancerous transformation of melanocytes, is highly aggressive, metastatic, and often fatal. In recent years, it has emerged as one of the fastest-growing types of cancer in terms of incidence, posing a serious threat to individuals worldwide [1, 2]. Currently, ICB therapy is the primary treatment for melanoma in clinical settings such as inhibitors targeting PD-1/PD-L1, which disrupt the shielding effect of cancer cells on T cells, leading to the activation of CD8+ T cells and subsequent the release of IFNγ to enhance STAT1 phosphorylation to inhibit the downstream transduction of the SLC7A11/GPX4 axis, ultimately triggering cancer cell immunogenic ferroptosis [3, 4]. Despite these advancements, more than half of patients fail to exhibit substantial benefits from ICB treatment, or even show no respond to the therapy [5]. The question of whether there is a key factor within melanoma that maintains tumor progression and impedes the onset of immune checkpoint inhibitors remains unknown and urgent.

Circular non-coding RNAs (circRNAs) are a type of regulatory non-coding RNAs generated by back-splicing of pre-mRNAs to form a circular covalently closed structure. These circRNAs are resistant to degradation by RNA enzymes and maintain stable functionality within the body [6, 7, 8]. CircRNAs often play crucial roles in disease development, particularly in cancer, where they serve as key regulators. It has been reported that circRNAs can influence cancer progression through multiple pathways, such as acting as sponges for miRNA molecules, thereby reducing the suppression of downstream genes by miRNA, or binding directly to target proteins to regulate the latter activity [9,10,11,12]. For instance, hsa_circ_102481 is highly expressed in non-small cell lung cancer (NSCLC) with EGFR-TKI resistance, acting as a molecular sponge to sequester miR-30a-5p and call back ROR1 levels, thereby promoting tumor growth and metastasis [13]. Additionally, circATXN7 transcripts are upregulated in KRAS-mutant tumor cells, which binds directly to the p65 subunit to block its nuclear localization, and leads to immunotherapy resistance [14]. Similarly, circYBX1 is notably abundant in highly invasive hepatocellular carcinoma tissues, inducing liquid–liquid phase separation of YBX1 by interacting with it, subsequently recruiting the HnRNP family to degrade TPM4 mRNA and accelerate liver cancer metastasis [15]. Recent studies have revealed that although circRNAs are typically categorized as non-coding RNAs, a subset of them possess the capability to encode proteins [16, 17]. These peptides have been identified as crucial elements influencing tumor progression [18]. Previous study [19] has identified circ-E-cadherin as a key player in promoting the advancement of glioblastomas. Interestingly, its mechanism of action does not involve miRNA sponge, but rather relies on multiple rounds of ribosomal translation to generate a 254 amino acid peptide. This peptide interacts with the CR2 structural domain of the EGFR, ultimately reactivating the downstream EGFR pathway and counteracting the effects of EGFR inhibitors.

Furthermore, recent studies have revealed that circRNAs play a significant role in tumor immunotherapy, influencing the tumor immune microenvironment, and the release and activation of immune factors [20,21,22,23]. CircFAM53B was found to be highly expressed in breast cancer, utilizing a distinct noncanonical ORFs to encode a 251aa polypeptide. This polypeptide not only plays a critical role in enhancing anti-tumor immunity but also holds potential as a vaccine candidate for treating malignant tumors [24]. It is hypothesized that a special peptide encoded by circRNA may be important for maintaining melanoma progression and reducing the effectiveness of ICB therapy.

In this study, we have identified circPIAS1, a circRNA highly expressed in melanoma, which encodes a 108 amino acid peptide that plays a crucial role in promoting melanoma progression and limiting the efficacy of immunotherapy. Further investigation discovered that circPIAS1-108aa hinders the phosphorylation of STAT1 by enhancing the latter SUMOylation modification, which leads to a significant limitation in IFNγ-induced immunogenic ferroptosis and ultimately impeds immunotherapy. Our findings shed light on a critical mechanism in tumors that obstructs the immune therapy and propose a novel strategy to enhance the effectiveness of ICB therapy in melanoma.

Materials and methods

Cell lines and cell culture

The mouse B16-F10 cells and human A375 cells utilized in this study were sourced from the Shanghai Cell Bank of the Chinese Academy of Sciences. Cells were cultured in DMEM medium supplemented with 10% fetal bovine serum (FBS) and incubated at 37 °C with 5% CO2.

Animal experiments

In this study, 6–8 weeks old C57 BL/6 J female mice (SPF) were used to create a melanoma cancer model by injecting 2 × 105 B16-F10 cells subcutaneously. Once the C57 mice’s tumors reached 50 mm3, they were divided into four groups for treatment (n = 6), with this day designated as day 0. PD-1 inhibitor and IgG negative control were purchased from BioXcell and given once every three days, with 2 mg/kg intraperitoneally injected each time for a total of five injections from day 0. ASO-NC and ASO-circPIAS1, synthesized by TSingke, were used to knock down circPIAS1. These drugs were administered once every three days, with 10 nmol (100 μL, 100 mM ASO in 100 μL of sterile PBS) injected into the tumor each time for a total of four injections from day 3. Tumor tissues from each group were collected on day 14 for further analysis. The animal experimental protocol underwent examination and approval by the Institutional Animal Care and Use Committee (IACUC) of the China Pharmaceutical University Experimental Animal Center (2024–03-020).

RNA-seq and Ribo-seq

Melanoma tumors were collected from C57 BL/6 J mice in the PD-1 inhibitor single administration group and combined administration group. Drug administration commenced when the tumors in the mice reached 50 mm3. The single PD-1 administration group received a liposome-encapsulated PD-1 inhibitor at a dosage of 2 mg/kg, administered once every three days for a total of five doses. Meanwhile, the combined administration group utilized a complex liposome formulation, wherein PD-1 encapsulated the liposome shell, which was subsequently linked to the conjugate composed of oridonin and hyaluronic acid. This group also received a dosage of 2 mg/kg, given once every three days for a total of five doses. Prior to sequencing, the samples underwent processing. For transcriptome sequencing (RNA-seq), the process involved extracting total RNA, eliminating linear RNA and ribosomal RNA, constructing a specific library, and finally sequencing on computer. On the other hand, translation group sequencing (Ribo-seq) required translation inhibition treatment, followed by RNase digestion to recover ribosomes, extraction of ribosome-imprinted RNA, and removal of ribosomal RNA before machine detection. RNA library sequencing was performed on the Illumina HiseqTM 2500/4000 by Gene Denovo Biotechnology Co., Ltd (Guangzhou, China).

Real-time quantitative PCR (qRT-PCR)

qRT-PCR was utilized in this study to detect and quantify various genes of interest. RNA extraction from cells and tissues was carried out using the RNA extraction kit (SHANGHAI YISHAN, ES-R001 and ES-RN002plus). Subsequently, the extracted RNA was reverse transcribed to produce cDNA (HiScript® II Reverse Transcriptase, Vazyme, R211-01). The resulting cDNA, along with primers and SYBRGreen (Taq Pro Universal SYBR qPCR Master Mix, Vazyme, Q311-02), were combined at a specific ratio for qRT-PCR detection. The relative RNA content was determined based on GAPDH using the formula 2−ΔΔCT. The specific primers utilized in this research are listed in Supplementary Table 1.

RNase R tolerance assay

Due to its circular covalently closed structure, circRNA is known to exhibit higher resistance to degradation by RNase R enzyme compared to linear mRNA. To confirm the presence of circRNA, an RNase R tolerance assay was conducted. Total RNA was extracted from the cells, a portion was treated with RNase R enzyme, incubated at 37 ℃ for 30 min, and then purified through column chromatography. Subsequently, 1 μg of the RNA product underwent reverse transcription using random primers, followed by PCR amplification. The relative abundance of linear mRNA and circRNA in the amplified products was assessed through 2% agarose gel electrophoresis after electrophoresis at 120 V for 45 min.

FISH

To determine the localization and content of circRNAs in cells and tissues, we employed FISH for detection. Both cell and tissue samples were fixed to maintain internal material integrity. A PBS solution with 0.5% Triton X-100 served as the permeabilization fluid to enhance permeability of tissues and cells. Following a 5-min wash with 1 × PBS, the samples were treated with prehybridization solution, blocked at 37 °C for 30 min, washed with 1 × PBS, and then incubated with the hybridization probe in a light-proof environment overnight at 37 °C. Subsequently, the samples were washed with hybridization solutions I, II, III, and 1 × PBS in succession at 42 °C in the dark. After DAPI staining, a final wash with 1 × PBS was conducted before imaging. The hybridization probes and kits utilized in this study were provided by RiboBio.

EdU proliferation assay

The EdU assay was employed to measure cell proliferation levels. Cells were plated at a density of 1 × 104 cells per well in a 96-well plate. Following transfection or drug treatment for 48 h, the original culture medium was removed, and added 1 × EdU working solution. The cells were then incubated at 37℃ for 4 h, fixed with 4% paraformaldehyde, and permeabilized using a solution containing 0.5% Triton X, followed by the addition of the working solution and an hour of incubation at room temperature. Subsequently, Hoechst staining was applied before capturing images. The EdU kit used in this experiment was provided by Beyotime (C0071L).

Dual-luciferase reporter assay

The dual-luciferase reporter assay was employed to measure IRES activity. A pcDNA3 RLUC POLIRES FLUC plasmid (Addgene plasmid # 45642; http://n2t.net/addgene:45642) served as a negative control, with the target sequence inserted between the Kpnl and Notl cloning sites. Target sequences included mouse M-IRES-1/2, M-IRES-2-del-1/2, human H-IRES-1/2, H-IRES-2-del-1/2, and an EMCV positive control. Following transfection of these plasmids into cells, fluorescence values of firefly and renilla were measured using a microplate reader after 48 h to determine relative fluorescence intensity. Transfection was carried out using LipofectamineTM 3000 and P3000 (ThermoFisher, L3000015) as the reagent, and the dual-luciferase reporter plasmid was provided by Vazyme (DL101-01).

Western Blot

Western Blot is a common technique used to detect the presence of different proteins. Cells were collected 48 h post transfection or administration and lysed using cell lysis buffer (Beyotime, P0013). The lysate was then centrifuged at 12000 g for 10 min to isolate proteins, which were quantified using a BCA kit (Beyotime, P0013). Each sample was loaded with 50 μg of protein, electrophoresed on a 10% SDS-PAGE, transferred to a PVDF membrane, and blocked with 5% skim milk for 2 h. Subsequently, the membrane was incubated with specific antibodies overnight at 4 °C. Following incubation with secondary antibodies, the signal was developed using ECL (EpiZyme, SQ201). Details of the antibodies used can be found in Supplementary Table 2.

Immunofluorescence(IF)

B16-F10 or A375 cells were seeded at a density of 1 × 104 cells per well prior to transfection or treatment. Following 48 h of the treatment, cells were fixed using 4% paraformaldehyde and permeabilized with a solution containing 0.5% Triton X-100 to enhance cell permeability. Subsequently, cells were washed with PBS, blocked with 5% BSA for 2 h at room temperature, incubated with primary antibody overnight at 4 degrees Celsius. After another PBS wash, the cells underwent a 2 h incubation with secondary antibodies at room temperature. The nuclei were then stained with Hoechst before fluorescence imaging was conducted. Detailed information regarding the antibodies used can be found in Supplementary Table 2.

Immunofluorescence staining on tissues

The melanoma tissues were collected by Sir Run Run Hospital of Nanjing Medical University (2023-SR-036), and the tissues microarray was processed by Bioaitech Co., Ltd. The tissue microarray was deparaffinized with xylene and rehydrated through a graded alcohol series. Antigen retrieval was performed using EDTA buffer (pH 8.0). Sections were blocked with BSA for 30 min at room temperature and incubated overnight at 4 °C with primary antibody. Sections were then incubated with the corresponding fluorescent secondary antibody for 1 h at room temperature in the dark. Samples were incubated with DAPI for 15 min at room temperature in the dark and washed twice with PBS for 5 min each. Autofluorescence quencher was added and incubated for 5 min before rinsing with water for 10 min. Imaging was performed using a fluorescence microscope.

IP assay

Cells were lysed 48 h post-transfection to extract total protein. Anti-Flag magnetic beads (Selleck, B26102) were incubated with total protein for 4 h at 4℃ on a shaker, followed by adsorption of the magnetic beads to collect immunoprecipitated proteins using a magnetic stand. Subsequently, the beads were washed with PBS, and 20 μL of lysis buffer and 20 μL of protein loading buffer were added. The mixture was then placed in a boiling water bath for 10 min prior to conducting Western Blot analysis.

LC–MS/MS

To confirm the presence of endogenous circPIAS1-108aa, LC–MS/MS was utilized for detection, B16-F10 and A375 cells were individually lysed using ultrasonic treatment, and total protein was extracted. Subsequently, proteins were separated via SDS-PAGE and stained with Coomassie brilliant blue. The band located around 12 kDa was excised, followed by reductive alkylation and enzymatic digestion to extract peptide fragments. These peptide fragments were then separated and analyzed using long gradient LC–MS/MS, with the data searched against a target protein database using Max Quant for identification purposes.

Lipid ROS determination

Following transfection or administration of the cells, the medium is replaced with fresh medium containing C11-BODIPY 581/591 (InvitrogenTM, C10445) at a concentration of 5 μmol/L and incubated at 37℃ for 1 h in the dark. Subsequently, the cells are washed with PBS, Hoechst is added for nuclear staining, and images are captured. When the cell membrane is oxidized by lipid hydroperoxide, lipid peroxidation damage occurs, resulting in a shift of the fluorescence emission wavelength peak from 590 nm (red) to 510 nm (green). The ratio of green to red fluorescence can be used to assess the extent of oxidative damage to the cell membrane.

GSH assay

One million B16-F10 or A375 cells were placed into a culture dish and proceeded with transfection or drug administration until reaching 60% confluence. After 48 h, collect the cells, add 500 μL of PBS, mix thoroughly, and perform ultrasonic lysis. Subsequently, centrifuge at 10,000 g for 10 min. Add working solutions 1 and 2 sequentially to the sample to be tested, then utilize a microplate reader to shake the plate for 1 min, allow it to rest for 5 min, and measure the OD value at 405 nm. Calculate the GSH concentration using the formula y = 0.0043x—0.01191 (R2 = 0.99736), and determine the relative GSH content based on the protein concentration of each sample. The GSH detection kit used was obtained from Elabscience (E-BC-K030-M).

MDA assay

Place 1 × 106 B16-F10 or A375 cells in a culture dish. After 48 h of transfection or drug administration, collect the cells. Add 200 μL of PBS, mix thoroughly, and then sonicate the mixture. Take 100 μL of the sample to be tested and combine it with 150 μL of LTBA diluent, 50 μL of LTBA storage solution, and 3 μL of antioxidant. Subsequently, immerse the mixture in boiling water for 15 min. Once cooled to room temperature, centrifuge at 1000 g for 10 min. Next, add 200 μL of the supernatant to a 96-well plate and measure the absorbance at 532 nm using a microplate reader. Use the formula y = 0.0048x—0.0041 (R2 = 0.9997) to calculate the MDA concentration for each sample, and then determine the relative MDA content based on the protein concentration of each sample. The MDA detection kit used was obtained from Beyotime (S0131M).

Preparation of liposomal drugs

Thermosensitive liposomes containing DPPC were prepared using the thin-film dispersion method and dried in a reduced pressure environment at 45 degrees Celsius. Subsequently, the PD-1 inhibitor was dissolved in PBS (pH = 7.2) and hydrated with liposomes at 45 degrees Celsius for 40 min. This was followed by sonication and dialysis in an ice-water bath to remove free antibody proteins. Conjugates were formed through condensation reactions, oridonin and hyaluronic acid were attached to both sides of BOC-β-Ala, respectively, and the resulting conjugates were collected by freeze-drying. Oridonin was quantitatively detected using HPLC at a wavelength of 238 nm in an LC-10AT system (Shimadzu Corporation, Japan) equipped with a UV detector (SPD-10Avp). The method is cited from a previous study by our group [25].

Statistical analysis

GraphPad Prism 9.5 was utilized for data processing and analysis. Venn diagrams and scatter plots were generated with Origin2021. Statistical analysis was conducted using ANOVA and Student’s t-test to determine group differences. Results are expressed as mean ± standard deviation (SD), and each experiment was replicated at least three times. Statistical significance was indicated as follows: ns (not significant), *P < 0.05 (significant), and **P < 0.01 (highly significant).

Results

CircPIAS1 is highly expressed in melanoma and promotes cancer progression

CircRNAs that hinder cancer immunotherapy tend to be consistently highly expressed in cancer [26]. Based on our previous study [25], tumor tissues from C57 BL/6 J mice treated with PD-1 inhibitor alone and in combination with liposomal drugs were analyzed using RNA-seq and Ribo-seq (Fig. 1a). Among the identified circRNAs with protein-coding potential, 17 were found to be upregulated with PD-1 inhibitor treatment alone and downregulated with the combination therapy. The 17 circRNAs were assessed for internal ribosome entry site (IRES) regional activity using the IRESfinder program [27] and further analyzed with a difference index (Fig. 1b). Our results showed that mmu_circ_0015773 was markedly downregulated after the combination therapy compared with PD-1 inhibitor single treatment, indicating that circPIAS1 may be important for promoting melanoma growth and impeding the effectiveness of immune drugs.

Fig. 1
figure 1

CircPIAS1 is highly expressed in melanoma tissues. a Tumor tissues from C57 BL/6 J mice after single and combined liposome administration of PD-1 inhibitor were subjected to RNA-seq and Ribo-seq to identify potential circRNAs. b Among the 17 candidate circRNAs, mmu_circ_0015773 with high protein coding potential and differential index was further identified using IRES scores and differential index. c Information on mmu_circ_0015773 from circBase and circBank databases was compiled. d Sequence conservation comparison between mmu_circ_0015773 and hsa_circ_0008378. e PCR assays of gDNA and cDNA using convergent and divergent primers to confirm the presence of circPIAS1. f-g RNase R tolerance experiments demonstrated that circPIAS1 actually exists and is resistant to degradation by RNase R enzymes, distinguishing it from linear mRNAs. h FISH assays were performed to detect the presence and intracellular localization of circPIAS1 in murine and human melanoma cells (B16-F10 and A375). Scale bar: 20 μm. i FISH assays were performed to assess circPIAS1 levels in patient melanoma and adjacent tissues. Scale bar: 50 μm. Data are representative of three experiments and presented as the mean ± SD. ns (not significant), *P < 0.05 and **P < 0.01 (significant)

Full size image

The mmu_circ_0015773, generated by back-splicing from the exons 3 to 6 of Pias1, corresponds to the hsa_circ_0008378 of human origin, exhibiting a sequence conservation of 94.71% (Fig. 1c-d). CircRNAs differ from linear non-coding RNAs, which result from the reverse shear cyclization of pre-mRNAs. To distinguish between the linear parent gene and circPIAS1, we designed the convergent primers for the linear host gene and the divergent primers for circPIAS1 to target its junction sites (Supplementary Table 1). Polymerase Chain Reaction (PCR) analyses on gDNA and cDNA confirmed the presence of circPIAS1 in human and mouse melanoma cells (Fig. 1e). Furthermore, RNase R exonuclease tolerance experiments provided additional validation (Fig. 1f-g). Fluorescence in situ hybridization (FISH) analysis revealed that circPIAS1 predominantly localizes in the nucleus (Fig. 1h). In addition, FISH analysis on patient tissues demonstrated a significant increase in circPIAS1 levels in melanoma compared to the adjacent tissues, supporting our initial hypothesis (Fig. 1i and Supplementary Fig. 1a).

To further investigate the pro-oncogenic role of circPIAS1, we developed specific ASO sequences targeting the junction sites of circPIAS1 (Supplementary Fig. 1b and Supplementary Table 3). ASO-circ-1 and 3 effectively reduced circPIAS1 expression without impacting the linear PIAS1 levels (Supplementary Fig. 1c). Functional assays including CCK8, colony formation, and EdU experiments consistently demonstrated a significant inhibition of melanoma cell proliferation upon circPIAS1 knockdown, indicating the pivotal role of circPIAS1 in promoting melanoma progression (Supplementary Fig. 1d-f).

The pro-oncogenic effect of circPIAS1 relies on the specific polypeptide it encodes

The database [27, 28] predicted that either mmu_circ_0015773 or hsa_circ_0008378 contain highly active IRES sequences and could translate a unique polypeptide consisting of 108 amino acids (circPIAS1-108aa) (Fig. 2a). Through dual-luciferase reporter experiment, we identified the precise IRES sequences of mouse and human circPIAS1, located at 81–254 bp and 21–194 bp, respectively (Fig. 2b-c). Subsequent truncation of the sequences revealed that the primary active regions are concentrated in the latter half (Fig. 2d-e).

Fig. 2
figure 2

CircPIAS1 is capable of encoding a specific 108aa-length polypeptide. a According to circPrimer 2.0, both human and murine circPIAS1 are predicted to encode a specific polypeptide of 108aa in length with a specific tail of 16 amino acids. b-c Dual-luciferase reporter assay to detect the activity and relative levels of IRES region of human and murine circPIAS1. d-e Dual-luciferase reporter assay to further detect the major active regions of IRES of human and murine circPIAS1. f LC–MS/MS detection of the protein fragment of circPIAS1-108aa to determine whether it is truly present in melanoma cells. g-h Overexpression and knockdown efficiency of circPIAS1 at the protein level were assessed through Western Blot analysis by using a polyclonal antibody specific to circPIAS1-108aa after transfection with different plasmids and ASOs. Data are representative of three experiments and presented as the mean ± SD. ns (not significant), *P < 0.05 and **P < 0.01 (significant)

Full size image

To validate the presence of circPIAS1-108aa, endogenous protein components from mouse and human melanoma cells were analyzed using high-precision LC–MS/MS. Multiple peptide fragments of circPIAS1-108aa, including its specific tail, were successfully identified (Fig. 2f). Polyclonal antibodies targeting the specific tail of circPIAS1-108aa were generated and used to assess the expression levels before and after ectopic regulation of circPIAS1(Fig. 2g-h).

To investigate whether the role of circPIAS1 in promoting melanoma proliferation is dependent on circPIAS1-108aa, two sets of overexpression plasmids: (1) circ-control, circPIAS1, circPIAS1-ATG-mut, and (2) linear-control, linear-ORF-108aa were transfected into cancer cells. qPCR and Western Blot experiments confirmed the overexpression at the RNA and protein levels, respectively (Figs. 2g and 3a). CircPIAS1-ATG-mut differs from circPIAS1 plasmid in that the former is unable to produce the specific protein upon transfection into cancer cells and can only be upregulated at the RNA level. The CCK8 assay indicated that melanoma proliferation was enhanced only when the circPIAS1-108aa protein level was increased (Fig. 3b). Consistent findings were observed in EdU and colony formation experiments (Fig. 3c-f), and were further supported by the IHC data from patients (Fig. 3g-h). These results corroborated that the oncogenic function of circPIAS1 relies on the specific 108aa polypeptide it encodes.

Fig. 3
figure 3

The promotion of melanoma cell proliferation by circPIAS1 is dependent on the specific peptide encoded by it. a qRT-PCR assay to measure the overexpression efficiency of circPIAS1 at the RNA level after transfection with various plasmids. b CCK8 assay to detect the impact of both circPIAS1 at RNA level and circPIAS1-108aa at protein level on melanoma cell viability. c-f EdU and colony formation assays were employed to examine the regulatory effects and relative quantification of circPIAS1 at RNA level and circPIAS1-108aa at protein level on the proliferative capacity of melanoma cells. Scale bar for EdU: 100 μm. g-h IHC staining analysis of patients’ melanoma and adjacent tissues for circPIAS1-108aa using specific polyclonal antibody. Scale bar: 100 μm. Data are representative of three experiments and presented as the mean ± SD. ns (not significant), *P < 0.05 and **P < 0.01 (significant)

Full size image

CircPIAS1-108aa activates the SLC7A11/GPX4 pathway by inhibiting STAT1 phosphorylation to avoid ferroptosis in melanoma cells

It is known that PIAS1 acts as a specific inhibitor of STAT1, preventing the formation of activated dimers of P-STAT1 and the subsequent downstream signaling pathways [29]. Therefore, it is hypothesized that circPIAS1 may be closely linked to STAT1. Through immunoblotting experiments, we observed alterations in STAT1 and its downstream SLC7A11/GPX4 signaling cascades following the knockdown or overexpression of circPIAS1 (Fig. 4a-b and Supplementary Fig. 2a-b). Interestingly, circPIAS1-108aa did not alter the total protein levels of STAT1 but notably suppressed its phosphorylation. Besides, SLC7A11 and GPX4 were positively modulated by circPIAS1-108aa. The SLC7A11/GPX4 axis, a critical downstream target of STAT1, is primarily involved in the transport of reducing glutathione (GSH) and the clearance of lipid peroxides from the cell membrane to mitigate oxidative damage and shield cells from ferroptosis [30, 31]. The study utilized the C11-BODIPY probe to directly measure lipid peroxidation damage on the cancer cell membrane. Results indicated that circPIAS1-108aa effectively reduced oxidative damage and protected cancer cells from ferroptosis (Fig. 4c and Supplementary Fig. 2c). Furthermore, detection of GSH and MDA levels confirmed that circPIAS1-108aa activated the SLC7A11/GPX4 cascades, leading to increased GSH content and decreased MDA levels (Fig. 4d and e).

Fig. 4
figure 4

CircPIAS1-108aa protects melanoma against ferroptosis by inhibiting the P-STAT1/SLC7A11/GPX4 signaling pathway. a-b Western Blot experiments were conducted to investigate the impact of circPIAS1 knockdown or overexpression on the P-STAT1/SLC7A11/GPX4 signaling pathway. c C11-BODIPY 581/591 probe was used to assess the impact of overexpression or knockdown of circPIAS1 on lipid peroxidation damage on cell membrane. Scale bar: 50 μm. d Relative levels of GSH in cells after knockdown or overexpression of circPIAS1. e Relative levels of MDA in cells after knockdown or overexpression of circPIAS1. f IF staining of patient tissue microarrays was performed to evaluate the expression levels of circPIAS1, P-STAT1, and GPX4 in tumor tissues compared with normal tissues. Scale bar: 50 μm. g Relative levels of circPIAS1 in melanoma tissues in comparison to normal tissues. h Relative levels of circPIAS1 and P-STAT1 in melanoma tissues and normal tissues and their correlation. i Relative levels of circPIAS1 and GPX4 in melanoma tissues and normal tissues and their relevance. Data are representative of three experiments and presented as the mean ± SD. ns (not significant), *P < 0.05 and **P < 0.01 (significant)

Full size image

To further investigate whether circPIAS1-108aa contributes to cancer progression by directly influencing ferroptosis or other forms of cell death, we conducted a CCK8 assay utilizing the inducer and the inhibitor of ferroptosis (Supplementary Fig. 3a-b). The results demonstrated that the tumor-killing effect of the ferroptosis inducer Erastin, was significantly enhanced, while the tumor-protective activity of the ferroptosis inhibitor Ferrostatin-1 was notably reduced in the presence of ASO-circPIAS1. Concurrently, knockdown of circPIAS1 did not lead to significant changes in the cell cycle (Supplementary Fig. 3c-d).

The examination on patient tissues also confirmed that the circPIAS1 levels were obviously higher in tumor tissues compared to normal tissues (Fig. 4f). In addition, the expression of circPIAS1 was negatively correlated with P-STAT1 and positively correlated with GPX4 in tumor tissues (Fig. 4g-i).

CircPIAS1-108aa limits IFNγ-induced immunogenic ferroptosis by suppressing STAT1 phosphorylation

Recent studies have shown that following treatment with ICB, activated CD8+ T cells produce high levels of IFNγ, which in turn promotes STAT1 phosphorylation, restrains the SLC7A11/GPX4 signaling pathway, and induces ferroptosis in melanoma [4, 32]. Interestingly, our study revealed that circPIAS1-108aa blocked these signaling pathways and prevented ferroptosis, suggesting that circPIAS1-108aa may counteract the effects of IFNγ by restraining this specific pathway (Fig. 5a-b and Supplementary Fig. 4a-b).

Fig. 5
figure 5

CircPIAS1-108aa resists IFNγ-induced immunogenic ferroptosis by inhibiting STAT1 phosphorylation. a-b Administration of IFNγ alongside knockdown or overexpression of circPIAS1 followed by Western Blot analysis to assess changes in protein levels within the pathway. c-d Impact of IFNγ administration while manipulating circPIAS1 levels on cellular GSH and MDA levels was evaluated. e C11-BODIPY 581/591 probes were utilized to assess lipid peroxidation damage on cell membrane surface in response to IFNγ and the manipulation of circPIAS1 expression. Scale bar: 50 μm. f CCK8 assay to detect the impact of IFNγ administration while manipulating circPIAS1 levels on cell viability. Data are representative of three experiments and presented as the mean ± SD. ns (not significant), *P < 0.05 and **P < 0.01 (significant)

Full size image

Reducing the levels of circPIAS1-108aa enhanced the phosphorylation and activation of STAT1 by IFNγ and strengthened the inhibition of the SLC7A11/GPX4 axis. Conversely, upregulating circPIAS1-108aa can significantly reverse the original effect of IFNγ on activating STAT1 phosphorylation, leading to the restoration of SLC7A11 and GPX4 levels. This was also evident in the GSH and MDA detection results (Fig. 5c-d).

Besides, through the detection of C11-BODIPY-labeled cell membranes, it was observed that IFNγ in combination with ASO-circPIAS1 promoted oxidative damage, whereas IFNγ combined with circPIAS1 upregulation reduced the oxidative damage (Fig. 5e and Supplementary Fig. 4c). Furthermore, CCK8 assay demonstrated that the presence of circPIAS1 significantly hindered the effectiveness of IFNγ. It was only after reducing circPIAS1 that IFNγ can exhibit its maximum killing effect (Fig. 5f). This observation was further supported by the outcomes of EdU and colony formation assays (Supplementary Fig. 4d-e).

These findings provided strong evidence that circPIAS1-108aa plays a crucial role in limiting the response of melanoma to IFNγ by inhibiting the phosphorylation and activation of STAT1.

CircPIAS1-108aa hinders STAT1 phosphorylation by promoting the latter SUMOylation

CircPIAS1-108aa is in the PINIT domain of the parent protein PIAS1, a crucial region for regulating PIAS1 nuclear localization (Supplementary Fig. 5a). IF experiments revealed that nuclear localization of PIAS1 was not disturbed after circPIAS1 knockdown (Supplementary Fig. 5b). Besides, no changes were observed in the levels of STAT1 phosphorylation upon PIAS1 knockdown (Supplementary Fig. 5c-d), implying that the inhibition of STAT1 phosphorylation by circPIAS1-108aa is not dependent on PIAS1. Furthermore, the phosphorylation of STAT1 was significantly inhibited by circPIAS1 overexpression regardless of PIAS1 knockdown (Supplementary Fig. 5e). Thus, our findings conclusively demonstrated that the inhibitory effect of circPIAS1-108aa on STAT1 phosphorylation is independent of PIAS1.

Given that STAT1 phosphorylation is regulated by various factors, including kinases and phosphatases [33, 34], we employed IP-MS technique to investigate whether circPIAS1-108aa influences this process by recruiting specific factors (Fig. 6a). Interestingly, circPIAS1-108aa did not recruit any kinases or phosphatases, but instead tightly bound to Ranbp2, a SUMO E3 ligase (Fig. 6b). IP results confirmed that circPIAS1-108aa effectively recruited Ranbp2 (Fig. 6c). In addition, circPIAS1-108aa and Ranbp2 were observed to co-localize in nucleus, creating favorable spatial conditions for their interaction (Fig. 6d). Previous studies have indicated that elevated levels of Ranbp2 can enhance the SUMOylation of STAT1, leading to the inhibition on STAT1 phosphorylation at Tyr701 [35,36,37,38]. Based on this, we hypothesized that circPIAS1-108aa enhances STAT1 SUMOylation by recruiting Ranbp2, thereby impeding STAT1 phosphorylation. Western Blot illustrated that circPIAS1 overexpression while Ranbp2 inhibition resulted in the restoration of P-STAT1 to its basal level (Fig. 6e and Supplementary Fig. 6a-b).

Fig. 6
figure 6

CircPIAS1 inhibits STAT1 phosphorylation by recruiting Ranbp2 to enhance STAT1 SUMOylation. a Proteins from B16-F10 cells transfected with vector-flag or C-P-108aa-flag were collected for IP, and the proteins interacting with circPIAS1-108aa were analyzed using LC–MS/MS technique. b LC–MS/MS analysis revealed that circPIAS1-108aa recruits SUMO E3 ligase Ranbp2. c IP assay confirmed the interaction between circPIAS1-108aa and Ranbp2. d IF assay demonstrated that circPIAS1-108aa and Ranbp2 are co-localized in the nucleus. Scale bar: 20 μm. e Overexpression of circPIAS1 and knockdown of Ranbp2, followed by Western Blot analysis, were conducted to assess the impact on STAT1 phosphorylation. f IP and Western Blot analysis were conducted to assess the impact of circPIAS1-108aa on the SUMOylation and phosphorylation levels of STAT1. g-h Recruitment of Ranbp2 by circPIAS1-108aa promotes the STAT1 SUMOylation at specific sites (K703 and E705), affecting the latter phosphorylation process. STAT1 mutations reverses the effect of circPIAS1-108aa on STAT1 phosphorylation, as confirmed by IP and Western Blot experiments

Full size image

To investigate the impact of circPIAS1-108aa on enhancing the SUMOylation of STAT1, STAT1-flag, SUMO1-His, and circPIAS1 overexpression plasmids were transfected into cells. The level of SUMOylation was assessed by IP and the results demonstrated that circPIAS1-108aa markedly promoted the SUMOylation and suppressed the phosphorylation of STAT1 (Fig. 6f and Supplementary Fig. 6c). Furthermore, mutations were introduced (Fig. 6g) and confirmed that the phosphorylation at Tyr701 was primarily influenced by the SUMOylation at Lys703 and Glu705 in STAT1. The SUMOylation was facilitated, and the phosphorylation was obstructed by circPIAS1 in the wild-type STAT1, but not in the mutant types (Fig. 6h and Supplementary Fig. 6d). These findings supplied the evidence that circPIAS1-108aa enhances the SUMOylation of STAT1 by recruiting Ranbp2, consequently inhibiting STAT1 phosphorylation.

Inhibiting circPIAS1 enhances the anti-melanoma effect of PD-1 inhibitor

To further investigate the role of circPIAS1 in melanoma and its potential as a therapeutic target, a melanoma subcutaneous inoculation model was established in C57 BL/6 J mice, followed by the administration of different treatments (Fig. 7a). ASOs were specifically designed to target and knock down circPIAS1 at its junction site (Supplementary Fig. 7a and d). PD-1 inhibitor and IgG-negative control were used to assess the impact of ASO-circPIAS1 on immunotherapy in vivo. Results from the animal experiments demonstrated that the health of the mice, including body weights and tissues, remained unaffected by the treatments (Fig. 7b and Supplementary Fig. 7c). Moreover, tumor volume was decreased following the administration of ASO-circPIAS1 or PD-1 inhibitor alone. The combined treatment showed the most pronounced inhibition (Fig. 7c-d). These findings suggested that the inhibition of circPIAS1 in conjunction with the PD-1 inhibitor significantly enhances the efficacy of immunotherapy, highlighting the role of circPIAS1 in limiting the tumor-killing effect of immune checkpoint inhibitors.

Fig. 7
figure 7

Targeting circPIAS1 significantly enhances the anti-melanoma effect of immune checkpoint inhibitors. a Melanoma models were established by subcutaneously inoculating B16-F10 cells into C57 BL/6 J mice, which were then divided into four groups upon tumor formation. These groups included: 1. ASO-NC + IgG negative control; 2. ASO-circPIAS1 + IgG negative control; 3. ASO-NC + PD-1 inhibitor; 4. ASO-circPIAS1 + PD-1 inhibitor. PD-1 inhibitor and IgG negative control were administered from day 0, every three days, for a total of five doses, while ASO-NC and ASO-circPIAS1 were administered from day 3, every three days, for a total of four doses. b Changes in body weight of mice across different treatment groups. c Melanoma tissues of mice from different treatment groups. d Changes in melanoma volume size over time in mice across different treatment groups. e IHC and C11-BODIPY staining were conducted on mouse melanoma tissues to assess various markers. Scale bar: 100 μm. f Changes in the levels of various proteins within the P-STAT1/SLC7A11/GPX4 signaling pathway in mouse tumors were detected using Western Blot assay. Data are representative of three experiments and presented as the mean ± SD. ns (not significant), *P < 0.05 and **P < 0.01 (significant)

Full size image

To investigate whether circPIAS1 could attenuate immunogenic ferroptosis by inhibiting the STAT1 phosphorylation process, we conducted IHC staining analysis on tumor tissues (Fig. 7e). The combination of ASO-circPIAS1 and PD-1 inhibitors led to a significant increase in P-STAT1 levels compared with the single administration, while GPX4 levels showed the opposite trend (Supplementary Fig. 7b). In addition, we observed alterations in the levels of the aforementioned proteins within mouse tumor tissues via Western Blot assays, which were consistent with those obtained from IHC (Fig. 7f). Previous researches have shown that cancer cells undergoing ferroptosis can modulate the immune microenvironment and enhance tumor immune infiltration [32, 39]. Therefore, we assessed the ferroptosis and immune microenvironment in tumor tissues (Fig. 7e). The results revealed that the ferroptosis was notably increased in cancer cells, and the CD8+ T cells and IFNγ levels were significantly risen in the immune environment after the combined treatment (Supplementary Fig. 7b).

In conclusion, the presence of circPIAS1-108aa hindered the efficacy of immune drugs. Knockdown of circPIAS1 using ASOs led to a marked enhancement in the anti-melanoma effect of PD-1 inhibitor without causing additional harm to the host. These findings suggested that circPIAS1 knockdown can enhance the effectiveness of immune checkpoint inhibitors, presenting a novel approach for improving clinical melanoma immunotherapy.

Discussion

ICB therapy for melanoma has been limited by poor efficacy and low response rates in recent years, prompting a search for factors that impede immune drug effectiveness [40,41,42]. CircRNAs have been identified as playing a significant role in tumorigenesis and cancer progression. Sometimes, they encode specific peptides to regulate tumor growth [24, 43]. Our research focuses on circRNA-encoding proteins and aims to uncover a mechanism regarding the limitations of immune checkpoint inhibitors in melanoma. Specifically, our investigation concentrated on circPIAS1, a product formed through back-splicing of PIAS1 pre-mRNA, which exhibited significant downregulation following coadministration. It has been found that circPIAS1 acts as an oncogene and encodes a specific 108aa peptide (circPIAS1-108aa) through multiple-round translation using the highly active IRES region. Interestingly, the oncogenic function of circPIAS1 is not RNA-dependent like a miRNA sponge, but rather relies on the novel peptide circPIAS1-108aa. Further investigation discovered that circPIAS1-108aa decreases STAT1 phosphorylation independent of PIAS1, resulting in the reactivation of the SLC7A11/GPX4 signaling pathway. This ultimately protects melanoma cells from immunogenic ferroptosis induced by IFNγ. The finding successfully sheds light on how circPIAS1-108aa hinders the effectiveness of ICB therapy in melanoma.

Further experiments revealed that circPIAS1-108aa facilitated the SUMOylation of STAT1 at Lys703 and Glu705 through the recruitment of the SUMO E3 ligase Ranbp2, instead of kinase or phosphatase, consequently impeding the phosphorylation of STAT1 at Tyr701. This process severely impaired the activation of STAT1 (Tyr701) by IFNγ and reduced the efficacy of IFNγ in killing tumors. Additionally, the combination of ASO-circPIAS1 with PD-1 inhibitor showed a stronger anti-melanoma effect compared to the single treatment of PD-1 inhibitor in vivo. Previous studies have reported that 4-Hydroxynonenal (4-HNE), an α-unsaturated fatty aldehyde, is closely related to the process of ferroptosis [44, 45]. In our investigation, we detected this indicator in mouse tumor tissues and found that, compared with the single administration, the content of 4-HNE increased markedly following combined administration. This finding is highly consistent with the results of this study (Fig. 7e and Supplementary Fig. 7b). Furthermore, we also assessed the levels of circPIAS1 in normal skin cells and various tumor cell lines. Interestingly, we observed a notable increase in circPIAS1 levels in tumor cells compared to normal skin cells, with the most pronounced elevation occurring in melanoma and colorectal cancer. These findings further support the characterization of circPIAS1 as a universal cancer-promoting factor (Supplementary Fig. 7f).

Obviously, neither the treatment of IFNγ at the cellular level nor the administration of PD-1 inhibitor at the animal level resulted in a down-regulation of circPIAS1 levels. In fact, there was even an upregulation observed (Supplementary Fig. 7d-e). This suggests that circPIAS1 functions as an oncogene that is consistently expressed in melanoma, maintaining high levels and impeding the efficacy of immune drugs.

The discovery of the function and regulatory mechanism about the specific peptide circPIAS1-108aa presents a very innovative approach to enhance melanoma treatment with ICB therapy, although exploring various strategies such as discovering more targets of ICB and optimizing drug structures and delivery systems are attractive. The development of specific antibodies, chemicals or ASOs targeting circPIAS1-108aa or circPIAS1, and the formulation of a combination regimen with immune checkpoint inhibitors hold promise for significantly improving ICB efficacy in melanoma.

Conclusions

In conclusion, our study has identified a unique 108aa-length peptide (circPIAS1-108aa) encoded by circPIAS1 in human and murine melanoma cells, which diminishes the anti-cancer effect of IFNγ. Besides, our research uncovers a novel mechanism of immune evasion in melanoma, where circPIAS1-108aa suppresses STAT1 (Tyr701) phosphorylation by recruiting SUMO E3 ligase Ranbp2 to enhance STAT1 SUMOylation at Lys703 and Glu705 (Fig. 8). This process reactivates the SLC7A11/GPX4 signaling pathway and hinders IFNγ-induced immunogenic ferroptosis, ultimately reducing the effectiveness of ICB therapy in melanoma.

Fig. 8
figure 8

Melanoma treated with immune checkpoint blockade (ICB) activates CD8+ T cells, leading to the release of IFNγ. This cytokine triggers the phosphorylation of STAT1, which in turn induces immunogenic ferroptosis by inhibiting the SLC7A11/GPX4 signaling pathway. The occurrence of immunogenic ferroptosis enhances the immune microenvironment response, establishing a positive feedback loop. However, circPIAS1, derived from back-splicing of exons 3 to 6 of PIAS1 pre-mRNA, exhibits high expression levels in melanoma. It encodes a unique 108aa-length peptide (circPIAS1-108aa, C-P-108aa) that recruits SUMO E3 ligase Ranbp2 to enhance STAT1 SUMOylation, inhibiting its phosphorylation and consequently counteracting the suppressive effect of phosphorylated STAT1 on the SLC7A11/GPX4 pathway. This mechanism allows melanoma cells to regain resistance to lipid peroxidation damage, thus significantly restricting the immunogenic ferroptosis induced by IFNγ and ultimately impeding ICB therapy for melanoma

Full size image

Availability of data and materials

The raw sequence data of RNA-seq and Ribo-seq reported in this paper have been deposited in the Genome Sequence Archive (Genomics, Proteomics & Bioinformatics 2021) in National Genomics Data Center (Nucleic Acids Res 2022), China National Center for Bioinformation / Beijing Institute of Genomics, Chinese Academy of Sciences (GSA: CRA016589) that are publicly accessible at https://ngdc.cncb.ac.cn/gsa.

Abbreviations

circRNA:

Circular RNA

RNA-seq:

Transcriptome sequence

Ribo-seq:

Translation group sequence

qRT-PCR:

Real-time quantitative PCR

FISH:

Fluorescence in situ hybridization

WB:

Western blot

IF:

Immunofluorescence

IP:

Immunoprecipitation

IP-MS:

Immunoprecipitation-Mass Spectrometry

LC-MS:

Liquid chromatography tandem mass spectrometry

IHC:

Immunohistochemistry

HE:

Hematoxylin-eosin

GSH:

Glutathione

MDA:

Malondialdehyde

EdU:

5-Ethynyl-2’-dexyuridine

CCK8:

Cell counting Kit-8

References

  1. The rapid rise in cutaneous melanoma diagnoses. N Engl J Med. 2021;384. Available from: http://www.nejm.org/doi/10.1056/NEJMc2101980. Cited 2024 May 15.

  2. Carvajal RD, Sacco JJ, Jager MJ, Eschelman DJ, Olofsson Bagge R, Harbour JW, et al. Advances in the clinical management of uveal melanoma. Nat Rev Clin Oncol. 2023;20:99–115.

    Article  PubMed  Google Scholar 

  3. Kalaora S, Nagler A, Wargo JA, Samuels Y. Mechanisms of immune activation and regulation: lessons from melanoma. Nat Rev Cancer. 2022;22:195–207.

    Article  CAS  PubMed  Google Scholar 

  4. Wang W, Green M, Choi JE, Gijón M, Kennedy PD, Johnson JK, et al. CD8 + T cells regulate tumour ferroptosis during cancer immunotherapy. Nature. 2019;569:270–4.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Long GV, Swetter SM, Menzies AM, Gershenwald JE, Scolyer RA. Cutaneous melanoma. Lancet. 2023;402:485–502.

    Article  PubMed  Google Scholar 

  6. Misir S, Wu N, Yang BB. Specific expression and functions of circular RNAs. Cell Death Differ. 2022;29:481–91.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Ngo LH, Bert AG, Dredge BK, Williams T, Murphy V, Li W, et al. Nuclear export of circular RNA. Nature. 2024;627:212–20.

    Article  CAS  PubMed  Google Scholar 

  8. Zhang XY, Li SS, Gu YR, Xiao LX, Ma XY, Chen XR, et al. CircPIAS1 promotes hepatocellular carcinoma progression by inhibiting ferroptosis via the miR-455-3p/NUPR1/FTH1 axis. Mol Cancer. 2024;23:113.

  9. Kristensen LS, Jakobsen T, Hager H, Kjems J. The emerging roles of circRNAs in cancer and oncology. Nat Rev Clin Oncol. 2022;19:188–206.

    Article  CAS  PubMed  Google Scholar 

  10. Vo JN, Cieslik M, Zhang Y, Shukla S, Xiao L, Zhang Y, et al. The landscape of circular RNA in cancer. Cell. 2019;176:869–881.e13.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Kristensen LS, Andersen MS, Stagsted LVW, Ebbesen KK, Hansen TB, Kjems J. The biogenesis, biology and characterization of circular RNAs. Nat Rev Genet. 2019;20:675–91.

    Article  CAS  PubMed  Google Scholar 

  12. He J, Xie Q, Xu H, Li J, Li Y. Circular RNAs and cancer. Cancer Lett. 2017;396:138–44.

    Article  CAS  PubMed  Google Scholar 

  13. Yang B, Teng F, Chang L, Wang J, Liu D-L, Cui Y-S, et al. Tumor-derived exosomal circRNA_102481 contributes to EGFR-TKIs resistance via the miR-30a-5p/ROR1 axis in non-small cell lung cancer. Aging. 2021;13:13264–86.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Zhou C, Li W, Liang Z, Wu X, Cheng S, Peng J, et al. Mutant KRAS-activated circATXN7 fosters tumor immunoescape by sensitizing tumor-specific T cells to activation-induced cell death. Nat Commun. 2024;15:499.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Liu B, Shen H, He J, Jin B, Tian Y, Li W, et al. Cytoskeleton remodeling mediated by circRNA-YBX1 phase separation suppresses the metastasis of liver cancer. Proc Natl Acad Sci. 2023;120: e2220296120.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Chen R, Wang SK, Belk JA, Amaya L, Li Z, Cardenas A, et al. Engineering circular RNA for enhanced protein production. Nat Biotechnol. 2023;41:262–72.

    Article  CAS  PubMed  Google Scholar 

  17. Pamudurti NR, Bartok O, Jens M, Ashwal-Fluss R, Stottmeister C, Ruhe L, et al. Translation of CircRNAs. Mol Cell. 2017;66:9–21.e7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Wen S, Qadir J, Yang BB. Circular RNA translation: novel protein isoforms and clinical significance. Trends Mol Med. 2022;28:405–20.

    Article  CAS  PubMed  Google Scholar 

  19. Gao X, Xia X, Li F, Zhang M, Zhou H, Wu X, et al. Circular RNA-encoded oncogenic E-cadherin variant promotes glioblastoma tumorigenicity through activation of EGFR–STAT3 signalling. Nat Cell Biol. 2021;23:278–91.

    Article  CAS  PubMed  Google Scholar 

  20. Wang J, Zhao X, Wang Y, Ren F, Sun D, Yan Y, et al. circRNA-002178 act as a ceRNA to promote PDL1/PD1 expression in lung adenocarcinoma. Cell Death Dis. 2020;11:32.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Morad G, Helmink BA, Sharma P, Wargo JA. Hallmarks of response, resistance, and toxicity to immune checkpoint blockade. Cell. 2021;184:5309–37.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Jiang W, Pan S, Chen X, Wang Z, Zhu X. The role of lncRNAs and circRNAs in the PD-1/PD-L1 pathway in cancer immunotherapy. Mol Cancer. 2021;20:116.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Feng Z, Zhang X, Zhou J, Li Q, Chu L, Di G, et al. An in vitro-transcribed circular RNA targets the mitochondrial inner membrane cardiolipin to ablate EIF4G2+/PTBP1 + pan-adenocarcinoma. Nat Cancer. 2023;5:30–46.

    Article  PubMed  Google Scholar 

  24. Huang D, Zhu X, Ye S, Zhang J, Liao J, Zhang N, et al. Tumour circular RNAs elicit anti-tumour immunity by encoding cryptic peptides. Nature. 2024;625:593–602.

    Article  CAS  PubMed  Google Scholar 

  25. Xiao Q, Li X, Liu C, Jiang Y, He Y, Zhang W, et al. Improving cancer immunotherapy via co-delivering checkpoint blockade and thrombospondin-1 downregulator. Acta Pharm Sin B. 2023;13:3503–17.

    Article  CAS  PubMed  Google Scholar 

  26. Miao Z, Li J, Wang Y, Shi M, Gu X, Zhang X, et al. Hsa_circ_0136666 stimulates gastric cancer progression and tumor immune escape by regulating the miR-375/PRKDC Axis and PD-L1 phosphorylation. Mol Cancer. 2023;22:205.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Zhao J, Wu J, Xu T, Yang Q, He J, Song X, IRESfinder. Identifying RNA internal ribosome entry site in eukaryotic cell using framed k-mer features. J Genet Genomics. 2018;45:403–6.

    Article  PubMed  Google Scholar 

  28. Zhong S, Feng J. CircPrimer 2.0: a software for annotating circRNAs and predicting translation potential of circRNAs. BMC Bioinformatics. 2022;23:215.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Li C, McManus FP, Plutoni C, Pascariu CM, Nelson T, Alberici Delsin LE, et al. Quantitative SUMO proteomics identifies PIAS1 substrates involved in cell migration and motility. Nat Commun. 2020;11:834.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Liang D, Feng Y, Zandkarimi F, Wang H, Zhang Z, Kim J, et al. Ferroptosis surveillance independent of GPX4 and differentially regulated by sex hormones. Cell. 2023;186:2748–2764.e22.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Lei G, Zhuang L, Gan B. Targeting ferroptosis as a vulnerability in cancer. Nat Rev Cancer. 2022;22:381–96.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Zhao L, Zhou X, Xie F, Zhang L, Yan H, Huang J, et al. Ferroptosis in cancer and cancer immunotherapy. Cancer Commun. 2022;42:88–116.

    Article  Google Scholar 

  33. Li Y-J, Zhang C, Martincuks A, Herrmann A, Yu H. STAT proteins in cancer: orchestration of metabolism. Nat Rev Cancer. 2023;23:115–34.

    Article  PubMed  Google Scholar 

  34. Philips RL, Wang Y, Cheon H, Kanno Y, Gadina M, Sartorelli V, et al. The JAK-STAT pathway at 30: much learned, much more to do. Cell. 2022;185:3857–76.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Li J, Su L, Jiang J, Wang YE, Ling Y, Qiu Y, et al. RanBP2/Nup358 mediates Sumoylation of STAT1 and antagonizes Interferon-α-Mediated antiviral innate immunity. Int J Mol Sci. 2023;25: 299.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Demel UM, Böger M, Yousefian S, Grunert C, Zhang L, Hotz PW, et al. Activated SUMOylation restricts MHC class I antigen presentation to confer immune evasion in cancer. J Clin Invest. 2022;132: e152383.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Begitt A, Droescher M, Knobeloch K-P, Vinkemeier U. SUMO conjugation of STAT1 protects cells from hyperresponsiveness to IFNγ. Blood. 2011;118:1002–7.

    Article  CAS  PubMed  Google Scholar 

  38. Rogers RS, Horvath CM, Matunis MJ. SUMO modification of STAT1 and its role in PIAS-mediated inhibition of Gene activation. J Biol Chem. 2003;278:30091–7.

    Article  CAS  PubMed  Google Scholar 

  39. Jiang X, Stockwell BR, Conrad M. Ferroptosis: mechanisms, biology and role in disease. Nat Rev Mol Cell Biol. 2021;22:266–82.

    Article  PubMed  PubMed Central  Google Scholar 

  40. Wolf Y, Bartok O, Patkar S, Eli GB, Cohen S, Litchfield K, et al. UVB-induced tumor heterogeneity diminishes immune response in melanoma. Cell. 2019;179:219–235.e21.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Fillon M. Fecal microbiota transplants may aid melanoma immunotherapy resistance. CA Cancer J Clin. 2021;71:285–6.

    Article  PubMed  Google Scholar 

  42. Frampton AE, Sivakumar S. A New Combination Immunotherapy in Advanced Melanoma. N Engl J Med. 2022;386:91–2.

    Article  CAS  PubMed  Google Scholar 

  43. Okholm TLH, Sathe S, Park SS, Kamstrup AB, Rasmussen AM, Shankar A, et al. Transcriptome-wide profiles of circular RNA and RNA-binding protein interactions reveal effects on circular RNA biogenesis and cancer pathway expression. Genome Med. 2020;12:112.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Liu L, Pang J, Qin D, Li R, Zou D, Chi K, et al. Deubiquitinase OTUD5 as a Novel Protector against 4-HNE‐Triggered ferroptosis in myocardial Ischemia/Reperfusion Injury. Adv Sci. 2023;10:2301852.

    Article  CAS  Google Scholar 

  45. Mao C, Lei G, Horbath A, Wang M, Lu Z, Yan Y, et al. Unraveling ETC complex I function in ferroptosis reveals a potential ferroptosis-inducing therapeutic strategy for LKB1-deficient cancers. Mol Cell. 2024;84:1964–79.

Download references

Acknowledgements

This work was supported by the Guangxi Key Laboratory of Early Prevention and Treatment for Regional High Frequency Tumor (GKE-KF202402), National Natural Science Foundation of China (Grant numbers 81973524), Special funds for science and technology development under the guidance of the central government (2021Szvup163, ZY20198020) and the "Double First-Class" University Project (CPU2018GF03).

Author information

Authors and Affiliations

  1. Jiangsu Key Laboratory of Bioactive Natural Product Research and State Key Laboratory of Natural Medicines, School of Traditional Chinese Pharmacy, China Pharmaceutical University, 24 Tong Jia Xiang, Nanjing, 210009, China

    Xin Zang, Xiao-Yu He, Cheng-Mei Xiao, Qing Lin, Meng-Yue Wang, Cheng-Yan Liu, Ling-Yi Kong & Yuan-Zheng Xia

  2. Department of Orthopaedics, Sir Run Run Hospital, Nanjing Medical University, 109 Long Mian Avenue, Nanjing, 211100, China

    Zhong Chen

  3. Guangxi Key Laboratory of Early Prevention and Treatment for Regional High Frequency Tumor and Key Laboratory of Early Prevention and Treatment for Regional High Frequency Tumor(Guangxi Medical University), Ministry of Education, Nanning, 530021, China

    Yuan-Zheng Xia

Authors
  1. Xin Zang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Xiao-Yu He
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Cheng-Mei Xiao
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Qing Lin
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Meng-Yue Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Cheng-Yan Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Ling-Yi Kong
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Zhong Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Yuan-Zheng Xia
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

X. Z.: Performing the experiments, Formal analysis, Investigation, Data collection and validation, Methodology, Visualization, Writing–original draft. X.Y.and C.M.: Performing the experiments, Validation, investigation, Methodology. Q. L.: Software, Methodology. M.Y.: Investigation, Visualization. C.Y.: Investigation, Visualization. L.Y.: Supervision, resources, writing–review and editing. Z. C.: Clinical cohort, resources, data curation, investigation, writing–review and editing. Y.Z.: Conceptualization, design the project, supervision, funding acquisition, resources, methodology, project administration, experimental validation, writing–review and editing.

Corresponding authors

Correspondence to Ling-Yi Kong, Zhong Chen or Yuan-Zheng Xia.

Ethics declarations

Ethics approval and consent to participate

Patient tissue collection was approved by the Sir Run Run Hospital of Nanjing Medical University (2023-SR-036). The animal experimental protocol underwent examination and approval by the Institutional Animal Care and Use Committee (IACUC) of the China Pharmaceutical University Experimental Animal Center (2024–03-020).

Consent for publication

All authors have agreed to publish this manuscript.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Zang, X., He, XY., Xiao, CM. et al. Circular RNA-encoded oncogenic PIAS1 variant blocks immunogenic ferroptosis by modulating the balance between SUMOylation and phosphorylation of STAT1. Mol Cancer 23, 207 (2024). https://doi.org/10.1186/s12943-024-02124-6

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12943-024-02124-6

Keywords

Molecular Cancer

ISSN: 1476-4598