Analysis of hsa_circ_0136256 as a biomarker for fibrosis in systemic sclerosis

Abstract

Background

Exploration of whether circRNAs in the skin of systemic sclerosis (SSc) model mice interact with 4E-BP1 protein to mediate the mTOR signaling pathway to regulate SSc fibrosis is crucial to identify homologous human circRNAs as markers to guide the diagnosis and treatment of SSc.

Methods

C57BL/6 mice aged 6–8 weeks and weighing approximately 20 g were subcutaneously injected with bleomycin (BLM) to establish an SSc model. High-throughput sequencing was used to screen the differentially expressed circRNA in the skin of SSc model mice and control mice. RNA immunoprecipitation and RNA pulldown confirmed the interaction between circRNA and 4E-BP1 protein. SSc model mice were treated with empty plasmid (OE-NC), overexpression plasmid of mmu_circ_0005372 (OE-circ_0005372), interference plasmid of mmu_circ_0005372 (sh-circ5372), mutant plasmid of mmu_circ_0005372 (circ5372-MT), mTOR activator (MHY1485), mTOR inhibitor (omipalisib), or JAK1/2 inhibitor (ruxolitinib). Sections of mouse skin tissue were stained with Hematoxylin and eosin and Masson’s stain. The collagen volume fraction (CVF) was calculated as CVF = area of blue collagen/total area with ImageJ. The correlation between homologous human circRNAs and clinical data was analyzed.

Results

Compared to the control group, 21,839 circRNAs were upregulated and 27, 946 circRNAs were downregulated in the skin tissue of mice in the SSc model group. Among them was mmu_circ_0005372, which is derived from the FZD3 gene, is closely related to fibrosis, and is involved in the mTOR signaling pathway. Hsa_circ_0136256 was identified as the homologous human circRNA of mmu_circ_0005372. RT-qPCR confirmed that the expression of mmu_circ_0005372 was significantly reduced in the skin tissue of SSc mice, and the expression of hsa_circ_0136256 was significantly reduced in the peripheral blood mononuclear cells of patients with SSc. The interaction between mmu_circ_0005372 and 4E-BP1 protein was inhibited in the skin tissue of SSc model mice. The results showed that the CVF of OE-circ_0005372 group was significantly lower than that of the sh-circ5372, circ5372-MT, and MHY1485 groups, indicating that OE-circ5372 significantly improved skin fibrosis in the SSc mice. ROC curve analysis was performed on hsa_circ_0136256 (AUC = 0.719, P = 0.035). The expression of hsa_circ_0136256 was negatively correlated with COL IV, RDW-SD, and RDW-CV, and positively correlated with VC, PLT, and PCT. The results suggested that hsa_circ_0136256 may have important roles in the clinical diagnosis of SSc.

Conclusion

Mmu_circ_0005372 and homologous human hsa_circ_0136256 may be biomarkers and therapeutic targets for SSc fibrosis.

Introduction

Systemic sclerosis (SSc) is a connective tissue disease caused by environmental, genetic, vascular, and immune abnormalities, and is characterized by localized or diffuse skin thickening and fibrosis [1]. The mechanism of SSc fibrosis is unclear. Clinically, patients with SSc are typically diagnosed and treated in the middle or late stages of the disease process, at which point conventional treatment is ineffective. Therefore, there is an urgent need to find suitable biomarkers and therapeutic targets for early diagnosis and individualized treatment of patients with SSc, which is the question to be addressed in this study.

Studies have shown that the mTOR complex 1 (mTORC1)/eukaryotic translation initiation factor eIF4E-binding protein 1(4E-BP1) axis plays a key role in the process of fibrosis. When mTORC1 is activated by upstream signals, mTORC1 promotes the activation of renal interstitial fibroblasts and renal fibrosis by phosphorylating 4E-BP1 [2]. In idiopathic pulmonary fibrosis (IPF), TGF-β1 interacts with the mTORC1/4E-BP1 signaling pathway to promote collagen synthesis in human lung fibroblasts [3]. Targeting this axis is therefore a promising anti-fibrosis strategy.

Noncoding RNAs (ncRNAs), including circRNA, lncRNA, and microRNA, do not encode proteins [4]. Due to the unique loop-forming structure of circRNA, which is more stable and conserved than other ncRNAs, circRNA has natural advantages as a biomarker [5]. In addition, circRNA can regulate gene expression through a variety of mechanisms. CircRNA can be used as a transcriptional regulator to regulate RNA polymerase II (RNA Pol II) and promote the expression of parental genes. CircRNAs that contain internal ribosome entry site (IRES) and open reading frames (ORFs) are involved in protein translation. CircRNA can act as an “miRNA sponge,” so that miRNA cannot bind to target genes, and can also directly interact with proteins to regulate the transcription of target genes [6].

Specific ncRNAs are closely related to the occurrence of fibrosis [7]. However, few relevant studies have reported whether circRNA interacts with 4E-BP1 protein to mediate the mTOR pathway and participate in the mechanism of fibrosis in SSc. Therefore, to elucidate the specific mechanism and find circRNAs as markers to guide the diagnosis and treatment of SSc, in this study, we selected C57BL/6 mice for induction of SSc skin fibrosis by subcutaneous injection of BLM, and then performed high-throughput sequencing of skin tissue from SSc mice and control mice. Mmu_circ_0005372, which is involved in the mTOR signaling pathway in SSc model mice and was screened, and showed decreased expression. Mmu_circ_0005372 is derived from the FZD3 gene and is closely related to fibrosis.

The BLAST online website was used to identify homologous human circRNA hsa_circ_0136256. The expression of mmu_circ_0005372 and its human homolog hsa_circ_0136256 was verified by RT-qPCR. The ability of the circRNAs to bind to 4E-BP1 was predicted by the catRNPID online website and verified by RIP and RNA pulldown. To explore the effect of mmu_circ_0005372 and the mTOR signaling pathway on skin fibrosis in SSc mice, the SSc model mice were treated with an mmu_circ_0005372 overexpression vector, mutant vector, or interference vector, and an activator or inhibitor of the mTOR signaling pathway. In addition, a correlation analysis between hsa_circ_0136256 and SSc clinical data provided potential biomarkers for the diagnosis and treatment of patients with SSc. The study design is shown in Fig. 1.

Fig. 1

The study design

Materials and methods

SSc model mice were established

The female C57BL/6 mice aged 6–8 weeks and weighing about 20 g were purchased from Jiangsu Jichi Pharmaceutical Kang Biotechnology Co., Ltd. The feeding conditions of the mice were as follows: temperature, 22℃; humidity, 50%–60%; the indoor air was kept fresh; the cage was sterilized, and the bedding was kept dry and clean; and sterilized purified water and rat food daily was chosen. The mice were randomly divided into the control group and SSc model group (n = 3 mice per group). Mice in the SSc model group were subcutaneously injected with 700 μg bleomycin hydrochloride (Zhejiang Hisun Pharmaceutical Co., Ltd., Taizhou, China), once every other day. The mice in the control group were injected with the same dose of normal saline at the same location. After 28 days, the mice were anesthetized by sodium pentobarbital at a dose of 70 mg/kg via intraperitoneal injection, and the mice were sacrificed by cervical dislocation. Animal experiments were conducted in accordance with the Animal Welfare Act and the Author’s Guide Consensus on Animal Ethics and Welfare.

Hematoxylin and eosin (HE) staining and Masson’s staining

Skin tissue soaked in paraformaldehyde was collected and embedded in paraffin. Hematoxylin and eosin staining (Sigma-Aldrich (Shanghai) Trading Co., Ltd., Shanghai, China) and Masson’s staining (Beijing Solarbio Technology Co., Ltd., Beijing, China) were performed. The collagen volume fraction (CVF) was calculated using ImageJ (National Institutes of Health, Bethesda, MD, USA), following the formula CVF = area of blue collagen/total area [8].

Real-time quantitative PCR (RT-qPCR)

Whole-blood samples were collected from 16 patients with SSc at an active stage and 32 age- and sex-matched healthy controls. All patients were newly diagnosed with SSc and had not been treated with hormones or immunosuppressive drugs before blood samples were collected. The inclusion criteria for patients with SSc complied with the classification criteria of the American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) issued in 2013. Peripheral blood samples were obtained from all patients after they provided informed consent. Reverse transcription of total RNA into cDNA using the ReverTra Ace qPCR RT kit (Toyobo Co., Ltd., Osaka, Japan). The prepared cDNA was amplified by PCR under the following conditions: 95°C, 1 min, 1 cycle; 95°C, 15 s, 40 cycles; 60°C for 30 s. The expression of circRNA was determined by the 2^−ΔΔCt method. The primer sequences used in this study are shown in Table 1:

Table 1 Primer sequences used in this study

High-throughput sequencing

The skin tissues of SSc model mice and control mice were sequenced on an Illumina platform. CIRI and find_circ software were used to predict the distribution of circRNAs on the genome and classify them. The DEGseq method was used to analyze the differentially expressed circRNAs, and the circRNAs with a relative expression (SSc:control) of | log2ratio |> 8 were selected. The heatmap was drawn using the website (http://www.bioinformatics.com.cn/). To further screen for circRNAs related to fibrosis, we began by selecting circRNAs that were downregulated in skin tissues of the SSc model mice and included in the circBase database (http://www.circbase.org/), and then analyzed the signaling pathways involved in their source genes. In addition, the BLAST online website (https://blast.ncbi.nlm.nih.gov/Blast.cgi) was used to determine the mmu_circ_0005372 homologous circRNA. CatRNPID (http://service.tartaglialab.com/page/catrapid_group) was used to forecast the interaction strength between circRNAs and 4E-BP1 protein.

Construction of OE-circ5372, sh-circ5372, and circ5372-MT vectors

The FASTA sequence of mmu_circ_0005372 was obtained from the circBase database (http://www.circbase.org/). The results predicted by the catRNPID website (http://service.tartaglialab.com/page/catrapid_group) showed that the 1–40 amino acid region of the 4E-BP1 protein might interact with the 1–400 nucleotide region of mmu_circ_0005372, with the 201–300 nucleotide region of mmu_circ_0005372 showing a stronger interaction likelihood. Therefore, the 201–300 bases of the mmu_circ_0005372 FASTA sequence were deleted, and circ5372-MT was constructed. By interfering with the entire FASTA sequence of mmu_circ_0005372, sh-circ5372 was constructed. The sequence of OE-circ5372 was identical to the FASTA sequence of mmu_circ_0005372, and these plasmids were then inserted into the pCD2.1-ciR vector. Overexpression of circ5372 (pCD2.1-ciR-OE-circ5372), interference of circ5372 (pCD2.1-ciR-sh-circ5372), and expression of the circ5372 deletion mutant vector (pCD2.1-ciR-circ5372-MT) were obtained. When the 3T3-L1 cells grew to 80% confluence, the old medium was discarded, and OPTI medium (Thermo Fisher Scientific [Shanghai] Co., Ltd., Shanghai, China) was added. Then 3 L Lipofectamine™ LTX (Thermo Fisher Scientific [Shanghai] Co., Ltd., Shanghai, China) was added to 50 L OPTI medium, and 1 L plasmid was added to another 50 L of OPTI medium. Equal volumes of diluted Lipofectamine™ LTX and plasmid were mixed into transfection complexes, and incubated at 25℃ for 5 min. The transfection complex was added at 100 L/well, and cultured in a cell incubator containing 5% CO₂ at 37℃ for 24 h. In addition, OE-NC (Blank plasmid of overexpression vector) and sh-NC (Blank plasmid of interference vector) transfected 3T3-L1 cells as negative control. The transfected 3T3-L1 cells were collected, and the expression of mmu_circ_0005372 in the OE-circ5372, sh-circ5372, and circ5372-MT groups was detected by RT-qPCR.

RNA immunoprecipitation (RIP) and RNA pulldown

Western blot analysis was used to detect the expression of 4E-BP1 protein before RIP experiments

Cell lysis buffer and PMSF were added to the OE-circ5372 group (3T3-L1 cells transfected with OE-circ5372 plasmid), sh-circ5372 group (3T3-L1 cells transfected with sh-circ5372 plasmid), and 3T3-L1 blank cell group (3T3-L1 cells transfected with WT-circ5372). After centrifugation, the supernatant was removed and the total protein concentration was determined by the BCA method (Beijing Solarbio Science & Technology Co., Ltd., Beijing, China). After denaturation, the proteins were subjected to gel preparation, sample loading, electrophoresis, membrane transfer, and blocking. Then, 4E-BP1 antibody (Abcam [Shanghai] Trading Co., Ltd., Shanghai, China) was added and the membranes were placed at 4°C for 8 h. Subsequently, the membranes were incubated with secondary antibody at 25°C for 1 h, horseradish peroxidase (HRP) conjugated goat anti-rabbit antibody (1:10,000) (Abcam [Shanghai] Trading Co., Ltd., Shanghai, China). Finally, the membrane was washed 3 times with phosphate buffer solution, and immersed in an ECL luminescence solution (Thermo Fisher Scientific [Shanghai] Co., Ltd., Shanghai, China) for approximately 1 min in an exposure box, and exposed in a dark room.

RIP experiment

After obtaining the above cells, the cells were lysed using RIP lysate and stored in a refrigerator at − 80°C. Resuspended magnetic beads were placed into a labeled Eppendorf tube, vortexed after adding the RIP wash buffer, and the supernatant was discarded. Next, approximately 10 μL of the corresponding antibody was added to each sample and incubated at room temperature for 20–30 min. The RIP wash buffer was added again, the beads were vortexed and shaken, and the supernatant was discarded. The samples were placed on a magnetic stand, and the RIP buffer was added. The cell lysates stored at − 80°C were rapidly thawed, the supernatant was aspirated and added to the magnetic beads-antibody complex, and the samples were incubated for 8 h at 4°C. The samples were centrifuged briefly, the supernatant was discarded, and the washes were repeated 3–5 times with the RIP wash buffer. The magnetic beads-antibody-protein complex was then subjected to RNA purification. The RNA was reverse transcribed into cDNA, and finally the enrichment of mmu_circ_0005372 was detected by RT-qPCR. The expression of 4E-BP1 in OE-circ5373 and sh-circ5373 cells was detected by western blotting (WB) after RIP. The expression of 4E-BP1 in immunoprecipitation with IgG antibody (IgG), with 4E-BP1 antibody (IP), and cell lysates without immunoprecipitation (Input) was analyzed.

RNA pulldown

The biotin-labeled target probe was synthesized with the probe sequence mmu_circ_0005372 (5 ‘–3’) ACUCAUCUGAGUAACCUUAGACAAGACUAA. The control group sequence was (5 ‘–3’) AAACAGUACUGGUGUGUAGUACGAGCUGAAGC UAC. The target probe labeled with biotin was combined with streptavidin magnetic beads, the liquid was discarded, the magnetic beads were collected, and the RNA protein binding reaction solution was added to bind circRNA to the specific labeled probe. The unbound proteins were then removed by washing, and the RNA–protein complexes were further eluted from the magnetic beads with an eluate, which was collected and subjected to RNA isolation. The RNA was reverse transcribed into cDNA, and the enrichment of mmu_circ_0005372 was detected by RT-qPCR, and the expression of 4E-BP1 protein was detected by WB.

SSc model mice were treated and analyzed

The data were collected from 40 mice. All the mice were randomly divided into eight groups according to their treatment: Healthy control mice (control), SSc model mice treated with empty plasmid (OE-NC), SSc model mice treated with overexpression plasmid of mmu_circ_0005372 (OE-circ_0005372), SSc model mice treated with interference plasmid of mmu_circ_0005372 (sh-circ5372), SSc model mice treated with mutant plasmid of mmu_circ_0005372 (circ5372-MT), SSc model mice treated with mTOR activator (MHY1485) (Selleck Chemicals, Houston, Texas, USA), SSc model mice treated with mTOR inhibitor (omipalisib) (Selleck Chemicals, Houston, Texas, USA), and SSc model mice treated with JAK1/2 inhibitor (ruxolitinib) (Selleck Chemicals, Houston, Texas, USA), with five mice per group. The plasmids were injected at a dose of 0.75 mg/kg via intravenous injection once a day. MHY1485 and omipalisib were administered at a dose of 3 mg/kg via intragastric gavage twice a day. Ruxolitinib was administered at a dose of 20 mg/kg via intragastric gavage once a day. All the above treatments were given once every two weeks for a total of three weeks. The mice were anesthetized by sodium pentobarbital at a dose of 70 mg/kg via intraperitoneal injection, before sacrificing by cervical dislocation.

Correlation analysis of circRNA and clinical data

IBM SPSS 26.0 (SPSS Inc., Chicago, Illinois, USA) and GraphPad Prism 8.0.2 (GraphPad, La Jolla, CA, US) were used in this study to analyze the relationship between circRNA and collagen type IV (COL IV), vital capacity (VC), standard deviation of red blood cell distribution width (RDW-SD), coefficient of variation of red blood cell distribution width (RDW-CV), platelet count (PLT), and platelet crit (PCT). The receiver operating characteristic curve (ROC) was used to diagnose SSc.

Statistical analysis

The data were analyzed by IBM SPSS 26.0 (SPSS Inc., Chicago, Illinois, USA) or GraphPad Prism 8.0.2 (GraphPad, La Jolla, CA, USA). Data are presented as the mean ± SD. The paired sample t-test was used to compare two groups. More than two groups were compared by one-way analysis of variance. The LSD t-test was used when the variance was homogeneous. The Tamhane test was analyzed when the variance was heterogeneous. Pearson or Spearman was used for correlation analysis between the clinical data and circRNA. A P-value < 0.05 was considered statistically significant.

Results

Analysis of differentially expressed circRNAs in the skin tissues of SSc model mice and control mice via high-throughput sequencing

In this study, the Illumina platform was used to screen for circRNAs in the skin tissue of three SSc model mice and three control mice. A total of 22,567, 23,060, and 21,789 circRNAs were screened from the first, second, and third samples of the control group, with 1,038, 1,074, and 1,081 circRNAs in the circBase database for the three samples, respectively. Additionally, 19,499, 17,966, and 23,152 circRNAs were screened from the first, second, and third samples of the SSc group, with 1,081, 1,013, and 1,099 circRNAs in the circBase database, respectively (Fig. 2A). The circRNAs were divided into intragenic circRNAs (introns and exons) and intergenic circRNAs (intergenic circRNAs) by combining the prediction results of the CIRI and find_circ software on the distribution of circRNAs in the genome. Among them, the numbers of circRNAs in the intron regions of the first, second, and third samples of the control group were 696, 466, and 560; the numbers of circRNAs in the exon regions were 18,528, 19,700, and 18,502; and the number of circRNAs in the intergenic region was 3,343, 2,894, and 2,727, respectively. Additionally, there were 498, 451, and 683 circRNAs in the intron regions of the first, second, and third samples of the SSc group; 16,943, 15,663, and 19,819 circRNAs in the exon regions; and 2,058, 1,852, and 2,650 circRNAs in the intergenic region, respectively (Fig. 2B). The differentially expressed circRNAs in the skin tissue of mice were analyzed by DEGseq (a new method based on MA-plot). Compared to the control group, 21,839 circRNAs were upregulated and 27,946 circRNAs were downregulated in the skin tissue of mice in the SSc model group. The number of circRNAs with no difference in expression was 12,491 (Fig. 2C). A heatmap showing differentially expressed circRNAs in the skin tissue of SSc model and control mice is shown in Fig. 2D.

In this study, mmu_circ_0005372 was found to be downregulated in SSc by high-throughput sequencing. Mmu_circ_0005372 was found in the circBase database, and its source gene was Fzd3. Using the circbank database (http: //www.circbank.cn/), hsa_circ_0136256 was found, and mmu_circ_0005372 was its conserved circRNA in mice. The homology between mmu_circ_0005372 and hsa_circ_0005372 was found to be 0%, while the homology between mmu_circ_0005372 and hsa_circ_0136256 was 91%. This suggested that mmu_circ_0005372 was highly homologous to hsa_circ_0136256 (Fig. 2E). In the genome structure map of mmu_circ_0005372 and hsa_circ_0136256, it can be seen that mmu_circ_0005372 was derived from exons 5, 6, and 7 of Fzd3 (Fig. 2F). The genomic structure of hsa_circ_0136256 consisted of exons 6, 7, and 8 of FZD3 (Fig. 2G)

Fig. 2

Analysis of differentially expressed circRNAs in the skin tissues of SSc model mice and control mice. A The number of circRNAs in the circBase database in the skin tissues of SSc model mice and control mice was classified by high-throughput screening (n = 3). B The circRNAs were divided into intragenic circRNAs (introns and exons) and intergenic circRNAs. C The differentially expressed circRNAs in the skin tissue of mice were analyzed by DEGseq (n = 3). Red represents upregulation, blue represents downregulation, and gray represents no difference. D Heatmap showing hierarchical cluster analysis of differentially expressed circRNAs in the skin tissue of SSc model mice and healthy control mice. The X-axis represents 16 samples from the cluster analysis, and the Y-axis represents differential circRNAs. Color represents expression normalized by the z-score. Red indicates upregulation and blue indicates downregulation (n = 3). E Mmu_circ_0005372 was highly homologous to hsa_circ_0136256 on the BLAST website. Range: the length of the alignment sequence. Score: alignment score, in which the longer the matching fragment, the higher the similarity, the higher the score. Expect: likelihood of a random match, with smaller expect values indicating more similar sequences. Identities: sequence similarity, number of matched bases as a percentage of total sequence length. Gaps: allows for gaps (base deletions or insertions) in the aligned sequence. F Genome structure map of mmu_circ_0005372. G Genome structure map of hsa_circ_0136256

.

Mmu_circ_0005372 and 4E-BP1 protein interaction is inhibited by reduced or absent expression of mmu_circ_0005372

The interaction of 4E-BP1 protein with mmu_circ_0005372, and hsa_circ_0136256 was predicted by catRNPID. The red box region in the heatmap indicates that the 1–40 amino acid region of 4E-BP1 protein and the 200–400 nucleotide region of mmu_circ_0005372 may interact (Fig. 3A). The red region in the heatmap indicates that the 25–50 amino acid region of 4E-BP1 protein may interact with the 200–400 and 600–800 nucleotide regions of hsa_circ_0136256 (Fig. 3B). The results showed that compared to the cells transfected with sh-NC, the expression of mmu_circ_0005372 in the cells transfected with the sh-circ5372 expression vector was significantly decreased (P < 0.001). Compared to the cells transfected with OE-NC, the expression of mmu_circ_0005372 in the cells transfected with the OE-circ5372 vector was significantly increased (P < 0.01). There was no significant difference in the expression of mmu_circ_0005372 between the cells transfected with the circ5372-MT vector and OE-NC (P > 0.05) (Fig. 3C). The results showed that the OE-circ5372, sh-circ5372, and circ5372-MT vectors were successfully constructed, and could be used for further experiments.

The expression of 4E-BP1 in the sh-circ5372, OE-circ5372, and WT-circ5372 groups was detected by WB before RIP experiments, and indicated that 4E-BP1 protein was expressed in 3T3-L1 cells, allowing RIP and RNA pulldown experiments to proceed (Figs. 3D and E). In 3T3-L1 cells transfected with OE-circ5372, the expression of mmu_circ_0005372 enriched in the 4E-BP1 antibody immunoprecipitation was significantly higher than that in the IgG immunoprecipitation (P < 0.01) (Fig. 3F and G). A WB after the RIP assay showed that 4E-BP1 was not enriched in the IgG immunoprecipitation but was enriched in the 4E-BP1 immunoprecipitation. In addition, the expression of 4E-BP1 in the 4E-BP1 immunoprecipitation was significantly higher in 3T3-L1 cells transfected with OE-circ5372 than in cells transfected with sh-circ5372 (Fig. 3H). RT-qPCR after the RNA pulldown assay showed that in 3T3-L1 cells transfected with sh-circ5372 and in 3T3-L1 cells transfected with OE-circ5372, the level of mmu_circ_0005372 enriched by the mmu_circ_0005372 probe was significantly higher than that of the negative control probe (P < 0.01) (Fig. 3I and J). WB results after the RNA pulldown assay showed that 4E-BP1 was not enriched by mmu_circ_0005372 or NC probes in 3T3-L1 cells transfected with sh-circ5372. However, the mmu_circ_0005372 probe showed that 4E-BP1 was enriched in 3T3-L1 cells transfected with OE-circ5372, and the NC probe did not enrich 4E-BP1 (Fig. 3K). This indicated that mmu_circ_0005372 interacted with 4E-BP1 protein, but when the expression of mmu_circ_0005372 was reduced or not expressed, the interaction between mmu_circ_0005372 and 4E-BP1 protein was inhibited

Fig. 3

The mmu_circ_0005372 and 4E-BP1 protein interaction is inhibited by reduced or absent expression of mmu_circ_0005372. A The interaction matrix of 4E-BP1 and mmu_circ_0005372 predicted by catRNPID, in which the red shaded part of the figure is the binding site. B The interaction matrix of 4E-BP1 and hsa_circ_0136256 was predicted by catRNPID online software, and the red shaded part in the figure is the binding site. C The expression of mmu_circ_0005372 in 3T3-L1 cells transfected with sh-circ5372, OE-circ5372, and MT-circ5372 was detected by RT-qPCR. D, E WB was used to detect the expression of 4E-BP1 and GAPDH in OE-circ5372, sh-circ5372, and WT-circ5372 groups before RIP. F, G The enrichment of mmu_circ_0005372 in transfected 3T3-L1 cells after RIP was detected by 19 RT-qPCR. H The expression of 4E-BP1 in 3T3-L1 cells transfected with sh-circ5372 and OE-circ5372 was detected by WB after RIP. I, J The enrichment of mmu_circ_0005372 in transfected 3T3-L1 cells after RNA pulldown by RT-qPCR. K The expression of 4E-BP1 in 3T3-L1 cells transfected with sh-circ5372 and OE-circ5372 after RNA pulldown was detected by WB. *P < 0.05; *P < 0.01. All the blots were from different gels without cropping. IgG means the enrichment of 4E-BP1 protein in immunoprecipitation with IgG antibody; IP means the enrichment of 4E-BP1 protein in immunoprecipitation with 4E-BP1 antibody; Input means cell lysates without immunoprecipitation

.

Skin fibrosis was inhibited by mmu_circ_0005372 overexpression in SSc mice

HE staining showed that dermis thickening and inflammatory cell infiltration were noticeable in the OE-NC group. Compared to the OE-NC group, the OE-circ5372, omipalisib, and ruxolitinib groups had significant reductions in inflammatory cells, significant reductions in collagen fiber bundles, and recovery of skin accessory glands and adipose tissue (Fig. 4A). Masson’s staining results showed that blue-stained collagen deposition was significantly reduced in the OE-circ5372, omipalisib, and ruxolitinib groups, but not in the OE-NC group (Fig. 4B). By analyzing the CVF in each treatment group, it was found that the CVF in the OE-circ5372 group was significantly lower than that in the sh-circ5372 group, circ5372-MT group, and MHY1485 group (P < 0.05) (Fig. 4C). The results indicated that MHY1485, as an activator of the mTOR signaling pathway, promoted the formation of skin fibrosis in SSc mice. When mmu_circ_0005372 was overexpressed in SSc mice, it inhibited the formation of skin fibrosis in SSc mice. When mmu_circ_0005372 was inhibited or mutated, it promoted the formation of skin fibrosis in SSc mice.

Fig. 4

Histopathological analysis and Masson’s staining of the skin tissue of SSc mice after treatment. A HE staining of skin tissue from treated mice (scale bar: 100 μm). B Masson’s staining of skin tissue of treated mice. Muscle fibers are shown in red and collagen fibers in blue (scale bar: 100 μm). C The collagen volume fraction (CVF) in the OE-circ5372 group was significantly lower than that in the sh-circ5372 group, circ5372-MT group, and MHY1485 group *P < 0.05

Hsa_circ_0136256 may be a biomarker for the clinical diagnosis and treatment of SSc

The demographic and clinical data of patients with SSc is presented in Table 2. The expression of hsa_circ_0136256 was significantly decreased in the PBMCs of patients with SSc (P < 0.05) (Fig. 5A). ROC curve analysis was performed on hsa_circ_0136256 (AUC = 0.719, P = 0.035). These results suggested that hsa_circ_0136256 may have key roles in the clinical diagnosis of SSc (Fig. 5B). We further analyzed the correlation between hsa_circ_0136256 and clinical data. The results showed that the expression of hsa_circ_0136256 was negatively correlated with COL IV (r = –0.588, P = 0.035), RDW-SD (r = − 0.354, P = 0.047), and RDW-CV (r = − 0.381, P = 0.031), and positively correlated with VC (r = 0.786, P = 0.036), PLT (r = 0.362, P = 0.042), and PCT (r = 0.423, P = 0.016) (Figs. 5C–H). Hsa_circ_0136256 may thus be a potential biomarker for the clinical diagnosis and treatment of SSc.

Table 2 Demographic and clinical data of patients with SSc

Fig. 5

Hsa_circ_0136256 may be a biomarker for the clinical diagnosis and treatment of SSc. A The expression of hsa_circ_0136256 was significantly decreased in the PBMCs of patients 24 with SSc (n = 16). HC: Healthy control, SSc: Patients with systemic sclerosis (SSc). *P < 0.05. B ROC curve analysis of hsa_circ_0136256 suggested that hsa_circ_0136256 plays key roles in the clinical diagnosis of SSc (n = 16). (C–H) hsa_circ_0136256 expression was negatively correlated with COL IV, RDW-SD, and RDW-CV, and positively correlated with VC, PLT, and PCT

Discussion

Exploring the pathogenesis of SSc and finding markers to guide the diagnosis and treatment of the disease is particularly important. CircRNA shows specific expression in specific tissues and different developmental stages, so it is regarded as a potential indicator for predicting disease diagnosis and staging [5]. At present, increasingly more studies have shown that circRNA regulates fibrosis of the lung, liver, heart, kidney, and other organs through a variety of mechanisms [9]. Given that circRNA can regulate fibrosis in multiple organs, there may also be a potential link between circRNA and fibrosis in SSc. MTOR plays an important regulatory role in fibrosis [10]. We selected circRNAs that were downregulated in the skin of SSc mice, and involved in mTOR signaling pathway as listed in the circBase database. To further screen for circRNAs related to fibrosis, the source genes of the circRNAs selected above were analyzed. The source gene of mmu_circ_0005372 was found to be frizzled homolog 3 (Fzd3). Several reports have suggested that FZD-mediated Wnt activates the mTOR signaling pathway to regulate the occurrence and development of fibrosis [11]. Therefore, we selected mmu_circ_0005372 derived from the FZD3 gene for subsequent study. As a downstream target of the mTORC1 signaling pathway, 4E-BP1 is closely related to fibrosis of diseases [12]. This study verified the interaction between mmu_circ_0005372 and 4E-BP1 protein in 3T3-L1 cells. In 3T3-L1 cells, when mmu_circ_0005372 was overexpressed, the enriched 4E-BP1 protein level was greater than that in 3T3-L1 cells transfected with sh-circ5372. These results suggest that mmu_circ_0005372, which was downregulated in SSc mice, and its homologous hsa_circ_0136256, which was downregulated in human PBMCs, caused the phosphorylation of 4E-BP1 protein by reducing their interaction with 4E-BP1 protein, which in turn activated mTORC1/4E-BP1 pathway, and then promoted the formation of fibrosis.

As an activator of mTOR, MHY1485 promotes the phosphorylation of S6K and 4E-BP1, the downstream targets of mTOR [13]. Omipalisib, a dual inhibitor of PI3K and mTOR, inhibits TGF-β1-induced synthesis of collagen, fibronectin, and α-SMA in lung fibroblasts [14]. Ruxolitinib is a JAK1/2 inhibitor with antifibrotic effects, and JAK1/2 inhibitors are more effective than JAK inhibitor monotherapy [15]. To further investigate whether mmu_circ_0005372 affected SSc fibrosis through the mTOR signaling pathway, SSc model mice were divided into eight groups that were administered different treatments. The results revealed that activation of the mTOR signaling pathway promoted skin fibrosis in SSc mice, and the anti-fibrosis effect of mmu_circ_0005372 overexpression in SSc mice was greater than that of direct inhibition of either the mTOR or JAK1/2 signaling pathway. The vector OE-circ5372 did not have a significant impact on the alleviation of the histopathologic symptoms compared to the healthy control and blank plasmid control (OE-NC). The expression of mmu_circ_0005372 was mutated (circ5372-MT), or inhibited (sh-circ5372), led to the exacerbation of the symptoms. The results suggested that overexpression of mmu_circ_0005372 in SSc mice may inhibit the formation of skin fibrosis. In addition, overexpression of mmu_circ_0005372 inhibited the mTOR/4E-BP1 signaling pathway in SSc mice, while the mTOR/4E-BP1 signaling pathway was activated when mmu_circ_0005372 expression was low. This may be caused by the inhibition of mmu_circ_0005372 binding to 4E-BP1.

Interstitial lung disease (ILD) is the most common cause of death in SSc, and is characterized by diffuse pulmonary parenchymal, alveolar inflammation, and pulmonary interstitial fibrosis, resulting in decreased lung function and ventilatory dysfunction. The quality of life of patients with ILD is seriously reduced and the mortality rate is high. Therefore, early detection, diagnosis, and treatment can delay the development of the disease. COL IV, which is a vital component of the lung basement membrane, plays an important role in the formation of lung epithelium and blood vessels, and also promotes the development of alveolar myofibroblasts [16]. Merl-Pham et al. [17] showed that the expression of COL IV was increased in the serum of patients with pulmonary fibrosis. In addition, Stefano et al. [18] found that COL IV could also be used as a biomarker to predict the presence or absence of liver fibrosis in patients with nonalcoholic fatty liver disease. In this study, the expression of hsa_circ_0136256 was negatively correlated with COL IV, indicating that hsa_circ_0136256 may be used as a biomarker to monitor the progression of fibrosis in SSc.

Pulmonary function is a routine examination for ILD, which plays an important role in assessing the severity and prognosis of the disease, evaluating the therapeutic effect, diagnosing the lesion site, and identifying the causes of dyspnea [19]. Due to long-term airflow limitation, ILD causes restrictive ventilatory dysfunction, which can further cause the reduction of VC, total lung capacity (TLC) and residual volume (RV) [20]. Lv et al. [21] found that pulmonary artery systolic pressure (PASP) was significantly negatively correlated with the diffusing lung capacity for CO (DLCO), VC%, and FVC% in patients with ILD. It was suggested that ILD patients with pulmonary hypertension may aggravate the impairment of lung function. In this study, the expression of hsa_circ_0136256 was negatively correlated with VC, suggesting that hsa_circ_0136256 may serve as a biomarker for monitoring lung function in SSc.

Red cell distribution width (RDW) is often used for the identification and diagnosis of anemia. However, recent studies have shown that RDW may be a marker of disease activity such as Sjogren’s syndrome, rheumatoid arthritis, and systemic lupus erythematosus [22,23,24]. Petrauskas et al. [25] showed that due to the lack of specific symptoms in the early stage, the diagnosis of pulmonary hypertension is easily delayed and can be life-threatening, but RDW can be used as a biomarker for patients with SSc with pulmonary hypertension, which may help patients to receive timely treatment. Ebata et al. [26] found that within 1 year of the diagnosis of SSC-ILD, an increase in the RDW value was correlated with a deterioration in the lung function of patients with SSc within 5 years. In this study, the expression of hsa_circ_0136256 was found to be negatively correlated with the RDW-SD and RDW-CV, suggesting that hsa_circ_0136256 could serve as an indicator to evaluate the severity of SSc in patients. SSc, when combined with other autoimmune diseases or renal crisis symptoms, can easily lead to a reduction in PLT [27, 28]. Gordon et al. [29] have demonstrated that thrombocytopenia is a risk factor for future renal crisis in patients with SSc. When thrombocytopenia occurs, the PCT decreases. However, the present study showed that the expression of hsa_circ_0136256 was positively correlated with PLT and PCT, suggesting that it may serve as a marker for the tendency of SSc with renal crisis.

The pathogenesis of SSc is complex, and the complications are serious. Moreover, as there is currently no effective anti-fibrosis treatment, it is crucial to identify a plasma marker for early detection and intervention of SSc. In recent years, many studies have shown that circRNA plays an important role in the diagnosis of diseases [30]. Chen et al. [31] found that the expression of hsa_circ_0000190 in the plasma of patients with gastric cancer was correlated with tumor diameter, lymphatic metastasis, and distant metastasis. Therefore, hsa_circ_0000190 is a potential noninvasive diagnostic biomarker. Ouyang et al. [32] showed that the expression of circRNA_002453 in the plasma of patients with lupus nephritis was positively correlated with 24-h proteinuria and the renal systemic lupus erythematosus disease activity index (SLEDAI score); thus, circRNA_002453 may be a biomarker for the diagnosis of kidney severity. In this study, we found that hsa_circ_0136256 was negatively correlated with COL IV, RDW-SD, and RDW-CV, and positively correlated with VC, PLT, and PCT, suggesting its potential use in monitoring disease progression. These findings suggest that hsa_circ_0136256 levels in peripheral blood mononuclear cells can serve as a non-invasive diagnostic biomarker for the early detection of fibrosis in patients with SSc.

This study has several limitations that should be acknowledged. The small human sample sizes may not be representative of the broader population, which limits the generalizability of the findings. Small sample sizes can also lead to overestimation or underestimation of effect sizes. Additionally, the human study was conducted with patients from a single center, which may limit the diversity of the patient population and potentially affect the external validity of the findings. Future research with larger, multi-center cohorts, longer follow-up durations, and additional validation studies would help to overcome these limitations and confirm the findings. While this study provides a promising insight into the role of hsa_circ_0136256 in SSc and its potential therapeutic applications, further research is needed to fully understand its mechanisms and develop effective therapeutic strategies. This includes clinical trials to assess safety and efficacy in humans, and the development of delivery systems for gene therapy or small molecule drugs targeting hsa_circ_0136256. A long-term follow-up for patients with SSc will be conducted to assess the persistence of the biomarker’s expression levels and its long-term correlation with disease progression or treatment response.

Conclusions

The expression of mmu_circ_0005372 was reduced in the skin tissue of SSc model mice, and the interaction between mmu_circ_0005372 and the 4E-BP1 protein may be inhibited, thereby inducing skin fibrosis by activating 4E-BP1/mTOR pathway. Hsa_circ_0136256, which is homologous to mmu_circ_0005372, is predicted to interact with the 4E-BP1 protein, and it was found to be inhibited in the PBMCs of patients with SSc. The expression of hsa_circ_0136256 was negatively correlated with COL IV, RDW-SD, and RDW-CV, and positively correlated with VC, PLT, and PCT, indicating its potential as a diagnostic biomarker and therapeutic target for SSc.