Circular RNA circLOC101928570 suppresses systemic lupus erythematosus progression by targeting the miR-150-5p/c-myb axis

Sixty-two SLE patients and healthy volunteers were recruited from the First Affiliated Hospital of Army Medical University between 2017 and 2020. The SLE patients included in the study fulfilled at least four of the 1997 revised criteria of the American College of Rheumatology (ACR) [20], and all the patients were diagnosed with SLE for the first time or without treatment with glucocorticoid or immunosuppressive agents for one month. The demographic, clinical, and laboratory characteristics of each subject were recorded, and disease activity was evaluated with the SLEDAI [21]. Information on SLE patients can be found in Table 1 and Additional file 1: Table S1. All participants were Han Chinese.

Samples were derived from PBMC, CD4 + T cells and CD8 + T cells in healthy human and SLE patients. And T cells were separately isolated by using EasySepTM Human CD4 + T cell Isolation Kit and EasySepTM Human CD8 + T cell Isolation Kit (STEMCELL, Canada). Total RNA was harvested and separated from PBMCs in samples via TRIzol reagent (Invitrogen, United States), and complementary DNA (cDNA) was synthesized sequentially. Two micrograms of total RNA was used to synthesize cDNA, a portion of which (1 µl, equal to 0.2 µg of cDNA) was used in a PCR assay. After reverse transcription with the PrimeScriptTM RT Reagent Kit (Takara, Japan), cDNA was amplified using SYBR Green Super Mix (Roche, Switzerland). Experimental results were analyzed through the 2-∆∆Ct method. The expression levels of miR-150-5p and nuclear circLOC101928570 were normalized to the expression of U6; in other cases, the expression levels of LOC101928570 and circLOC101928570 were normalized to the expression of β-actin mRNA. The sequences of the primers used for qRT–PCR in this study are shown in Additional file 2: Table S2.

Total RNA (2 μg) of PBMCs was mixed with 3 U/μg ribonuclease R (Epicenter Technologies, United States) at 37 °C for 20 min. The stability of circLOC101928570 and linear LOC101928570 was determined. Relative expression levels were evaluated by qRT–PCR.

The PBMC suspension was pipetted onto autoclaved glass slides, followed by dehydration with 70%, 80% and 100% ethanol. Then, hybridization was performed at 37 °C overnight in a dark, moist chamber using fluorescently labeled probes for circLOC101928570 and miR-150-5p. Briefly, circLOC101928570 was captured with a Cy3-labeled probe (5′-TGGCTTGAATAGATTGGGACTA ATA-3′), while miR-150-5p was captured with a digoxin-labeled probe (5′-TCTCCCAACCCTTGTACCAGTG-3′) for fluorescence in situ hybridization (FISH). MiR-150-5p was further visualized using the anti-digoxin rhodamine-conjugated antibody. After washing, the slices were sealed with parafilm containing DAPI. Specimens were analyzed using a Nikon inverted fluorescence microscope.

Whole blood (10 ml) was collected in EDTA collection tubes from each subject, and human PBMCs were isolated by density-gradient centrifugation using Ficoll-Paque Plus (GE Healthcare Biosciences, United States) and cultured in RPMI 1640 medium (Gibco, United States) supplemented with 10% fetal bovine serum (Gibco, USA), 100 U/mL penicillin, and 100 U/mL streptomycin (Gibco, United States) at 37 °C with 5% CO₂ for 24 h before transfection.

The circLOC101928570 overexpression plasmids and empty vector were constructed and designed by GeneChem (Shanghai, China). The miR-150-3p mimics, miR-150-5p mimics, miR-150-3p inhibitor, miR-150-5p inhibitor, miRNA mimics NC, and miRNA inhibitor NC were chemically synthesized by Riobio (Guangzhou, China). The pCDH-CMV-MCSEF1- GFP + Puro (CD513B-1) vector was used as a negative control plasmid, and the pCDH-MYB plasmid was fragmented and inserted into the pCDH-CMV-MCSEF1-GFP + Puro (CD513B-1) vector with BamHI and NotI restriction sites. The vector construction results were verified by direct sequencing. The sequence used was 5 ‘-CCGGAATTCCGGGAAAGCGTCACTTGGGGAAAA-3’. PLKO.1-puro (Addgene plasmid # 8453) was used to design short hairpin RNA, and the cells were transfected with Lipofectamine 3000 (Invitrogen, United States). The transfection process lasted 48 h.

293 T cells in 24-well plates were cotransfected with miR-150-3p/5p mimic, inhibitor, and negative control, and luciferase reporter vectors (SV40-Luc-MCS) with wild-type or mutant circLOC101928570 were designed and constructed by GeneChem (Shanghai, China). For circLOC101928570 and miR-150-3p/5p luciferase reporter assays, the circLOC101928570 sequences containing wild-type miR-150-3p/5p predicted binding sites were inserted into the region directly downstream of a cytomegalovirus (CMV) promoter-driven firefly luciferase cassette in a pCDNA3.1 vector. The IL2RA 3’ UTR sequences containing two predicted wild-type c-myb binding sites were inserted into the region directly downstream of a T7 promoter-driven firefly luciferase cassette in a psiCHECKTM-2 vector (Promega, United States). All constructs were verified by sequencing. 293 T cells were seeded into 24-well plates and cotransfected with a mixture of 1 μg of luciferase reporter plasmid and PCDH and plasmids pCDH-MYB, shNC, and shMYB. The sequence of shRNA-MYB#1 and shRNA-MYB#2 used were 5ʹ-CCGGGCGGCTGAATAGGTTGCTTGTTTCAAGAGAACAAGCAACCTATTCAGCCGCTTTTTTGGTACC-3ʹand 5'-CCGGGCACACGACAGAGATCTTTCCTTCAAGAGAGGAAAGATCTCTGTCGTGTGCTTTTTTGGTACC-3', respectivley. After 48 h, luciferase activity was measured using the Dual-Luciferase® Reporter Assay System (Promega, United States). The relative firefly luciferase activity was normalized to Renilla luciferase activity. All experiments were performed in triplicate.

To stably knock down circLOC101928570 expression, Jurkat cells were cultured and infected with lentivirus carrying shRNA targeting circLOC101928570 and a negative control vector. PLKO.1-puro was used to design short hairpin RNA, and the restriction sites were AgeI (R3552S) and EcoRI (R3101T). After 1300 bp, a single restriction site, KpnI (R0142M), was used. For lentivirus packaging, HEK-293 T cells were transfected with the core plasmid PLKO.1-shRNA, the psPAX2 packaging plasmid and the pMD2.G envelope plasmid for 48 h to obtain the lentivirus supernatant. The shRNA sequences used are shown in Additional file 3: Table S3. All constructs were verified by sequence analysis. No homology to any other human genes was detected.

Double staining of Annexin V and 7-aminoactinomycin-D (7-AAD) was carried out using a PE Annexin V Apoptosis Detection Kit (BD Pharmingen™, United States) and an APC Annexin V Kit (BD Pharmingen™, United States) according to the manufacturer’s recommendations. Briefly, cells were washed with cold phosphate-buffered saline and then resuspended in binding buffer at a concentration of 1 × 10⁶ cells/ml. Then, 100 μl of solution (1 × 10⁵ cells) was transferred to a tube, and 5 μl of PE Annexin V and 5 μl of 7-AAD were added. After incubation for 15 min at room temperature in the dark and the addition of 400 μl of binding buffer, flow cytometric analysis was performed (FACScan, BD Biosciences, San Jose, CA, United States) within 1 h, and the data were analyzed with FlowJo software (Treestar, Ashland, OR). PE Annexin V ( +) or APC Annexin V ( +) and 7-AAD (-) cells represent the early stage of apoptosis, whereas PE Annexin V ( +) or APC Annexin V ( +) and 7-AAD ( +) cells are in the end stage of apoptosis or are already dead.

Concentrations of IL-4 and IFN-γ in peripheral blood serum supernatants were analyzed using a Human IL-4 Instant ELISA Kit (eBioScience, United States) and a Human IFN gamma Platinum ELISA Kit (eBioScience, United States) following the manufacturer’s instructions, respectively. The concentrations were calculated according to their corresponding standard curves.

A mutually targeted method was applied to predict putative ceRNAs for circRNAs. To predict ceRNAs for circLOC101928570, we used circMir1.0 software to identify circLOC101928570-targeting miRNAs.

The biotinylated circLOC101928570 probe was specifically designed and synthesized for binding to the junction site of circLOC101928570. The biotin-coupled RNA complex was pulled down by incubating the cell lysates with PierceTM Streptavidin Magnetic Beads (Thermo Fisher Scientific, United States) following the manufacturer’s instructions. The enrichment of miRNAs in the capture fractions was evaluated by qRT–PCR analysis. After reverse transcription with a Mir-X™ miRNA First Strand Synthesis Kit (Takara, Japan), complementary DNA (cDNA) was amplified using a Mir-X™ miRNA qRT–PCR TB Green® Kit (Takara, Japan), the expression of miRNAs was normalized to the expression of U6, and the probe sequences used are listed in Additional file 2: Table S2.

A RIP assay was performed using a Magna RIP RNA Binding Protein Immunoprecipitation Kit (Millipore, United States) according to the manufacturer’s instructions. Briefly, PBMCs were harvested and lysed in RIP lysis buffer on ice for 30 min. After centrifugation, the supernatant was incubated with 30 μl of Protein G agarose beads (Roche, United States) and 8 μl of AGO2 (ab57113, Abcam, Cambridge, MA) antibody. After overnight incubation, the immune complexes were centrifuged and then washed six times with washing buffer. The bead-bound proteins were further analyzed using Western blotting. The immunoprecipitated RNA was subjected to qRT–PCR analysis.

The complete proteome from PBMCs was extracted after lysis in RIPA lysis buffer (Beyotime, China) supplemented with protease inhibitors (Sigma–Aldrich, United States) and then separated via sodium dodecyl sulfate–polyacrylamide gel electrophoresis. The samples were electrotransferred to polyvinylidene fluoride membranes (Millipore, United States). After blocking with 5% fat-free milk, the membranes were treated with prepared primary antibodies: anti-c-myb (Abcam, England), rabbit IL2Rα antibody (CST, United States), and rabbit GAPDH antibody (CST, United States) were used as controls. Membranes were washed and then treated with an anti-rabbit secondary antibody (CST, United States). The blot signal was examined using Pierce ECL Western blotting Substrate (Thermo Fisher Scientific, United States) with a ChemiDoc™ Touch Imaging System (Bio–Rad, United States). The integrated density values were calculated using Quantity One software (Bio–Rad, United States).

Flow cytometry was performed on freshly isolated PBMCs, as a previous study showed that CD25 expression is affected by freezing–thaw procedures. PBMCs were stained with fluorochrome-conjugated antibodies to identify blood CD4⁺ and CD8⁺ T cell subsets by flow cytometry. The staining procedure was conducted blinded to genotype and was performed simultaneously for each pair. Prior to staining, FcR Blocking Reagent (Miltenyi Biotec, Bergisch Gladbach, Germany) was added to the PBMCs to prevent nonspecific binding. We used monoclonal antibodies specific for CD3, CD8, CD25, IFN-γ, IL-4, FOXP3 and IL17A for flow cytometry. All antibodies were purchased from BioLegend (San Diego, CA, United States). After the samples were stained for surface markers, the cells were further fixed and permeabilized, followed by staining for intracellular indicators. The stained PBMCs were washed 3 times in cold PBS/2% FCS/0.1% NaN₃, and data were acquired on a FACS Canto II 8 color flow cytometer (BD Biosciences, San Jose, CA, United States) aiming for 30,000 acquisitions.

Data were analyzed and visualized with GraphPad Prism 8.0 (GraphPad Software, La Jolla, CA, United States). Receiver operating characteristic (ROC) analysis was performed to differentiate patients from healthy controls, areas under the curve (AUCs), optimal threshold values, sensitivity, and specificity were calculated. Quantitative values are expressed as the means ± standard deviation (SD) of at least three independent repetitions. Statistical differences among groups were tested with an unpaired two-tailed Student’s t test. Spearman’s analysis was used to test correlation. A P value less than 0.05 was chosen to indicate statistical significance.

To identify circRNAs that contribute to SLE pathogenesis, we analyzed the circRNA RNA-seq data [22] to identify circRNAs with significant differences (P < 0.01) between SLE patients and healthy controls (Fig. 1A and Additional file 4: Table S4). We selected and identified 2 upregulated circRNAs and 2 downregulated circRNAs (|fold change|> 2, P < 0.01) that were significantly related to SLE (Additional file 4: Table S4. labeled in red), among which circLOC101928570 showed a lower level of expression in SLE patients. The circLOC101928570 was derived from the LOC101928570 gene locus and was predicted to target miR-150, a well-known SLE-associated miRNA [23, 24] and experimentally confirmed that their expression was consistent with the circRNA RNA-seq data (data not shown).

To further investigate the effects of circLOC101928570 on SLE, we verified that it was markedly downregulated in PBMCs of 62 SLE patients versus 62 healthy controls, we also checked the expression of circLOC101928570 in CD4 + and CD8 + T cells isolated by magnetic activated cell sorting (MACS) from the PBMCs of SLE patients and healthy controls. Results showed that circLOC101928570 was downregulated in the CD4 + and CD8 + T cells from SLE patients compared with the healthy controls (Fig. 1B). Liner regression analysis demonstrated that patients with lower expression of circLOC101928570 had a significantly higher SLEDAI score (Fig. 1C) and a lower complement C3 level (Fig. 1D). To assess the diagnostic value of circLOC101928570 for SLE, receiver operating characteristic (ROC) curve analysis was performed to determine the relative circLOC101928570 expression between the 62 patients and 62 healthy controls (Fig. 1E). The area under the curve (AUC) was 0.8985, and the 95% confidence interval (95% CI) was 0.8441 − 0.9530. We next evaluated the circular structure of circLOC101928570, which originated from exon 3 of the LOC101928570 gene (chr6: 77271340–77273307). Sanger sequencing validated the back-spliced junction of circLOC101928570 (Fig. 1F). Software verified that 2 CPG islands existed in the transcriptional core region of circLOC101928570 (http://​www.​ebi.​ac.​uk/​emboss/​cpgplot/​), as shown in Fig. 1G. Then, we used the exonuclease RNase R to examine the stability of circLOC101928570. circLOC101928570 showed strong resistance to digestion by RNase R, whereas the linear RNAs of LOC101928570 and β-Actin were highly degraded (Fig. 1H). FISH results showed that circLOC101928570 was predominantly localized in the cytoplasm (Fig. 1I). Together, these data suggest that circLOC101928570 was significantly expressed at low levels in SLE patients and correlated with SLE.

To explore the possible mechanism of functional circLOC101928570, we identified its ceRNAs with circMir1.0 software (http://​www.​bioinf.​com.​cn/​). The results showed that circLOC101928570 has two targets, i.e., the two mature products (miR-150-5p and miR-150-3p) of pre-miR-150 [25] (Fig. 2A). The miRNAs were extracted after pull-down assay, and the levels of 10 candidate miRNAs were detected by real-time PCR. MiR-150-3p/5p was abundantly pulled down by circLOC101928570 in PBMCs (Fig. 2B). To determine whether circLOC101928570 functions as a miRNA sponge, we next performed AGO2 immunoprecipitation to determine whether circLOC101928570 served as a platform for AGO2 and miR-150-5p. The results showed that circLOC101928570 binds to AGO2, and qRT–PCR results further supported this observation (Fig. 2C). To confirm that circLOC101928570 could be regulated by miR-150-3p/5p, we constructed luciferase reporters containing wild-type and mutated putative binding sites of circLOC101928570 transcripts (Fig. 2D) and then cotransfected them with miR-150-3p/5p mimics or inhibitors into 293 T cells. Luciferase reporter assays showed that the luciferase activities of the circLOC101928570 wild-type reporter were significantly reduced when transfected with miR-150-3p/5p mimics compared with the control reporter or mutated luciferase reporter (Fig. 2E). However, the miR-150-3p/5p inhibitor significantly increased the luciferase signal of the wild-type circLOC101928570 reporter compared with the control reporter or the mutated luciferase reporter (Fig. 2F). qRT–PCR further confirmed that circLOC101928570 knockdown could increase the miR-150-5p level and that circLOC101928570 overexpression had the opposite effect in Jurkat cell lines (Fig. 2G). However, miR-150-3p/5p did not significantly influence circLOC101928570 levels (Fig. 2H). Moreover, RNA FISH assays revealed that circLOC101928570 and miR-150-5p were colocalized in the cytoplasm (Fig. 2I). These results showed that circLOC101928570 can bind to miR-150-5p.

To explore the function of miR-150 in SLE, we then detected the expression of miR-150-5p using qRT–PCR in PBMCs obtained from 36 patients with SLE and 25 healthy controls. The results showed that the expression of miR-150-5p was upregulated in SLE (Fig. 3A). MiR-150-5p expression was also contrarily correlated with complement C3 levels (Fig. 3B). There was a strong positive correlation between miR-150-5p expression and the SLEDAI score in patients with SLE (Fig. 3C). To assess the diagnostic value of miR-150-5p in SLE, we also performed ROC curve analysis with the relative miR-150-5p expression in the 36 patients and 25 healthy controls (Fig. 3D). The AUC was 0.7889, and the 95% CI was 0.6643–0.9135. These results demonstrated that miR-150 was highly expressed in SLE patients and correlated with SLE.

It has been widely reported that c-myb is the target protein of miR-150. MiR-150 is highly expressed in mature lymphocytes, and c-myb is an important transcription factor for regulating lymphocyte development and participating in the pathogenesis of SLE [26‐29]. To further study the role of circLOC101928570 in progression, we constructed two short hairpin RNAs (shRNA-circLOC101928570#1, shRNA-circLOC101928570#2). Jurkat cells with stable circLOC101928570 knockdown with lentiviral shRNA and circLOC101928570 overexpression plasmid with lentiviral were then established. We found that shRNA-circLOC101928570 successfully knocked down circLOC101928570 expression but had no effect on LOC101928570 mRNA expression in Jurkat cells (Fig. 3E, F). Similarly, circLOC101928570 was successfully overexpressed in Jurkat cells, while LOC101928570 mRNA expression had no obvious change (Fig. 3G, H). To explore whether circLOC101928570 could regulate c-myb by competitively binding with miR-150-5p, we transfected miR-150-5p mimics and circLOC101928570 overexpression plasmids into Jurkat cells. CircLOC101928570 overexpression increased the expression of c-myb, while transfected miR-150-5p mimics significantly attenuated the circLOC101928570-induced increase in the expression of c-myb (Fig. 3I). Downregulation of circLOC101928570 resulted in decreased expression of c-myb. Furthermore, transfection with the miR-150-5p inhibitor promoted the decreased expression of c-myb (Fig. 3J). These data demonstrated that circLOC101928570 regulates c-myb by competitively binding miR-150-5p to mediate the immune inflammatory response in SLE.

As a proto-oncogene, c-myb plays crucial roles in the processes of cell development, differentiation and apoptosis. Downregulated expression of c-myb will cause cell apoptosis [30‐32]. Furthermore, previous studies have reported that PBMCs from individuals with SLE showed significantly decreased levels of c-myb [28, 29]. The expression of c-myb has a positive correlation with the number of immune complexes (ICs) and clinical disease activity [28]. Our study has shown that the downregulation of circLOC101928570 decreased c-myb expression by sponging miR-150-5p in SLE, suggesting that circLOC101928570 might be involved in SLE pathogenesis by regulating c-myb expression.

To further explore the function of circLOC101928570, we examined cell apoptosis in stable Jurkat cells and found that the percentage of apoptotic cells was significantly increased in the shRNA-circLOC101928570#1 group compared with the shRNA-NC group (Fig. 4A, B). The proportion of apoptotic cells was lower in the circLOC101928570 overexpression group than in the NC group (Fig. 4A and C). These findings suggested that circLOC101928570 negatively regulated the early apoptosis of Jurkat cells. Identifying the factors contributing to the enhanced apoptosis of SLE T cells will deepen our understanding of SLE pathogenesis.

We then further explored the pathogenesis of circLOC101928570 involvement in SLE. Software-based prediction of transcription factor c-myb target genes (http://​cistrome.​org/​db/​#/) revealed that IL2RA might bind c-myb in Th1 and Th2 CD4⁺ T cells and T lymphocytes (Fig. 5A). CD4⁺ T lymphocytes are an important factor in the pathogenesis of SLE, mainly manifested by the immune imbalance of CD4⁺ T lymphocytes, and the differentiation of CD4⁺ T lymphocytes is regulated by IL-2 [33]. However, the specific mechanism remains to be clarified. Normal serum IL-2 levels are an important condition to maintain the normal function of CD4⁺ T lymphocytes, B cells and NK cells. As a transcription factor, c-myb may regulate target gene expression at the transcriptional level, thereby exerting biological functions. We analyzed the potential binding DNA sequence loop of c-myb and found two theoretical binding sites in the top 2000 nt of the promoter domain of the IL2RA gene (http://​jaspar.​genereg.​net). Hence, we speculated that c-myb may regulate IL2RA expression at the transcriptional level. To confirm this supposition, a luciferase plasmid with the top 2000 nt of the promoter domain of the IL2RA gene (psicheck2-WT) and a luciferase plasmid with mutant sequences in both binding sites of the top 2000 nt of the promoter domain (psicheck2-Mutant) were generated (Fig. 5B). In addition, we designed two short hairpin RNAs (shRNA-MYB#1 and shRNA-MYB#2), and an MYB overexpression plasmid was constructed with a PCDH vector. We found that shRNA-MYB successfully knocked down MYB expression and that MYB was successfully overexpressed in 293 T cells (Fig. 5C and D). Luciferase reporter assays demonstrated that c-myb enhanced the luciferase activity of psicheck2-WT in a dose-dependent manner but not that of psicheck2-Mutant (Fig. 5E, F), suggesting that c-myb enhanced IL2RA expression by directly binding to the promoter domain of IL2RA. These data demonstrated that c-myb may suppress SLE progression by positively regulating IL2RA expression at the transcriptional level. To explore whether circLOC101928570 could regulate the expression of IL2RA, we transfected circLOC101928570 overexpression plasmid, circLOC101928570-NC, shRNA-circLOC101928570#1, and shRNA-NC into Jurkat cells. Upregulated circLOC101928570 increased the expression of IL2RA (Fig. 5G). Downregulation of circLOC101928570 resulted in the decreased expression of IL2RA (Fig. 5H). These data indicated the function of circLOC101928570 in regulating IL-2RA expression.

We analyzed the cell cytokine expression in the serum from 55 different SLE and 33 healthy groups by ELISA. IL-4 expression was decreased in the SLE group compared with the healthy group (Fig. 6A). There was no significant difference in IFN-γ expression in SLE patients compared with healthy controls (Fig. 6B). The IL-4/IFN-γ level in the SLE group was lower than that in the healthy group (Fig. 6C). The IL-4/IFN-γ ratio of SLE was negatively correlated with the expression of circLOC101928570 (Fig. 6D). Meanwhile, the Th17/Treg ratios negatively correlated with the expression of circLOC101928570 (Fig. 6E). We analyzed the Th17 and Treg percentages in PBMCs from the SLE and healthy groups by flow cytometry (Fig. 6F, G). The complete gating strategy regarding the flow cytometry for Th17 and Treg cells was shown in the Additional file 5: Figure S1-2. By linear regression analysis, the Th1/Th2 and Th17/Treg ratios were correlated with the expression of circLOC101928570. Next, we analyzed the expression of IL2RA on Th1, Th2, Tc1 and Tc2 cells in the T cell populations between SLE patients and healthy controls by flow cytometry. The results demonstrated that the expression of IL2RA on Th1, Th2, Tc1 and Tc2 cells in SLE patients was lower than that in healthy controls (Fig. 6H, I). We analyzed the percentages of IL2RAs among Th1, Th2, Tc1 and Tc2 cells in PBMCs from different SLE and healthy groups by flow cytometry (Fig. 6J, K). The complete gating strategy regarding the flow cytometry was shown in the Additional file 5: Figure S3. These data indicated that circLOC101928570 suppresses IL2RA expression in T cell subsets of SLE.