Up-regulation of circRNA-0003528 promotes mycobacterium tuberculosis associated macrophage polarization via down-regulating miR-224-5p, miR-324-5p and miR-488-5p and up-regulating CTLA4

Plasma hsa_circ_0003528 was elevated in TB patients

In our previous study, we utilized a circRNA microarray to analyze plasma hsa_circ_0003528 in patients with active pulmonary TB in comparison with that in healthy controls. The microarray data were subsequently validated by RT-qPCR of the study population, which included 170 TB patients, 40 pneumonia patients, 40 COPD patients, 40 lung cancer patients, and 150 healthy controls. The clinical characteristics of the study population were presented in the previous literature [25, 26]. Moreover, the validation stage presented good consistency between microarray and RT-qPCR data [25, 26], indicating a significant elevation of hsa_circ_0003528 in TB patients.

Correlation between miRNAs expression and hsa_circ_0003528 expression

To explore the possible miRNAs sponged by hsa_circ_0003528, a computational analysis was performed. As shown in Figure 1, 16 candidate miRNAs were identified, and the according miRNA mimics were transfected into THP-1 cells to observe their effects on the expression of hsa_circ_0003528. Accordingly, the RT-qPCR results presented an evident negative correlation between expression of hsa_circ_0003528 and expression of miR-224-5p, miR-324-5p, miR-488-5p, miR-587, and miR-668, and a significant positive correlation between expression of hsa_circ_0003528 and miR-192 and miR-217.

MiR-224-5p, miR-324-5p and miR-488-5p were down-regulated in TB patients

In our previous study, to validate the microarray profile of hsa_circ_0003528, 50 patients with active pulmonary TB (as the TB group) and 50 healthy controls (as the HC group) were recruited. In this study, we further validated the correlations between the expression of miRNAs (including miR-224-5p, miR-324-5p, miR-488-5p, miR-587 and miR-668) and the expression of hsa_circ_0003528 among these patients. As shown in Figure 2, the relative expression of miR-224-5p (Figure 2A), miR-324-5p (Figure 2B) and miR-488-5p (Figure 2C) was all markedly down-regulated in the TB group, while no evident difference was observed between the TB and HC groups in terms of the expression of miR-587 (Figure 2D) and miR-668 (Figure 2E).

Hsa_circ_0003528 functioned as a sponge of miR-324-5p, miR-224-5p and miR-488-5p

To study the molecular relationship between hsa_circ_0003528 and miR-224-5p, miR-324-5p and miR-488-5p, luciferase assays were conducted to confirm the specific binding between hsa_circ_0003528 and above miRNAs. Different putative binding sites of hsa_circ_0003528 were located on miR-324-5p (Figure 3A), miR-224-5p (Figure 3B) and miR-488-5p (Figure 3C), and the luciferase activity of wild-type hsa_circ_0003528 was evidently suppressed in THP-1 cells transfected with miR-324-5p (Figure 3A), miR-224-5p (Figure 3B) and miR-488-5p (Figure 3C). Therefore, it can be validated that miRNAs including miR-224-5p, miR-324-5p and miR-488-5p could bind to hsa_circ_0003528.

CTLA4 mRNA was targeted by miR-224-5p, miR-324-5p and miR-488-5p

The computational analysis also identified putative binding sites of miR-324-5p (Figure 3D), miR-324-5p (Figure 3E) and miR-488-5p (Figure 3F), respectively, on the 3’UTR of CTLA4 mRNA. The luciferase assay indicated that the expression of wild-type CTLA4 mRNA was markedly decreased in THP-1 cells transfected with miR-324-5p (Figure 3D), miR-224-5p (Figure 3E) and miR-488-5p (Figure 3F). Therefore, it can be validated that CTLA4 mRNA was targeted by miR-224-5p, miR-324-5p and miR-488-5p.

Establishment of a molecular signaling pathway

THP-1 cells were transfected with hsa_circ_0003528 or an empty vector. It was found that the expression of miR-224-5p (Figure 4A), miR-324-5p (Figure 4B) and miR-488-5p (Figure 4C) was all suppressed while the expression of CTLA4 mRNA (Figure 4D) and protein (Figure 4E) was promoted by the transfection of hsa_circ_0003528 in THP-1 cells. Furthermore, U937 cells were transfected with hsa_circ_0003528 siRNA and a negative control (Figure 4F), and the expression of miR-224-5p (Figure 4G), miR-324-5p (Figure 4H) and miR-488-5p (Figure 4I) was all up-regulated while the expression of CTLA4 mRNA (Figure 4J) and protein (Figure 4K) was down-regulated by the transfection of hsa_circ_0003528 siRNA in U937 cells.

The hsa_circ_0003528 signaling pathway was associated with macrophage polarization induced by IFN-γ + LPS or IL-4

THP-1 cells were grouped as the M(-) group, M(IFN-γ + LPS) group and M(IL-4) group. The expression of hsa_circ_0003528 (Figure 5A) and CTLA4 mRNA (Figure 5E)/protein (Figure 5F) was increased in cells undergoing macrophage polarization induced by IFN-γ + LPS or IL-4. However, although the expression of miR-224-5p (Figure 5B), miR-324-5p (Figure 5C) and miR-488-5p (Figure 5D) was increased in macrophages treated with IFN-γ + LPS, the M(IL-4) group showed reduced expression of these miRNAs.

The hsa_circ_0003528 signaling pathway was associated with M1 to M2 macrophage polarization

The activity of the hsa_circ_0003528 signaling pathway was further compared between the M(IL-4) group, M(IFN-γ + LPS) group, M(M-CSF) group, M(GM-CSF) group, M(IFN-γ + LPS) group, M(IFN-γ + LPS), M(IL-4) group, M(IL-4) group, M(IL-4) group and M(IFN-γ + LPS) group. Accordingly, the expression of hsa_circ_0003528 (Figure 6A and 6B) and CTLA4 mRNA (Figure 6Q and 6R) was increased in the M(IFN-γ + LPS) group and M(GM-CSF) group compared with that in the M(IL-4) group (Figure 6A and 6Q) and M(M-CSF) group (Figure 6B and 6R), respectively. The change in macrophage polarization also reduced the expression of hsa_circ_0003528 (Figure 6C and 6D) and CTLA4 mRNA (Figure 6S and 6T), i.e. the treatment with IL-4 reversed increased expression of hsa_circ_0003528 and CTLA4 mRNA caused by IFN- γ + LPS (Figure 6C and 6D) and the treatment with IFN- γ + LPS reversed increased expression of hsa_circ_0003528 and CTLA4 mRNA caused by IL-4 (Figure 6S and 6T). In addition, we also evaluated the expression of the expression of miR-224-5p (Figure 6E–6H), miR-324-5p (Figure 6I–6L) and miR-488-5p (Figure 6M–6P) in each treatment group. In line with the result about hsa_circ_0003528 (Figure 6A and 6B) and CTLA4 mRNA in each group, it was found that miR-224-5p (Figure 6E–6H), miR-324-5p (Figure 6I–6L) and miR-488-5p (Figure 6M–6P) showed opposite tendency against that of hsa_circ_0003528 and CTLA4 mRNA among these treatment groups.

Overexpression of hsa_circ_0003528 in cells of macrophages polarization induced by IFN-γ + LPS regulated the expression of chemokines

THP-1 cells undergoing macrophages polarization induced by IFN-γ + LPS were transfected with empty vector (as Group 1), hsa_circ_0003528 (as Group 2), NC siRNA (as Group 3) or circ_0003528 siRNA (as Group 4). As shown in Figure 7, the successful transfection of circ_0003528 or circ_0003528 siRNA were validated by the varied expression of circ_0003528 (Figure 7A). And the expression of CLCL10 (Figure 7B), CXCl11 (Figure 7C) and Nos2 (Figure 7D) was decreased while the expression of CCL17 (Figure 7E) and CCL18(Figure 7F) was increased by the over-expression of hsa_circ_0003528. The concentrations of TNF-α (Figure 8A), IL-6 (Figure 8B) and IL-12 (Figure 8C) were down-regulated while the concentrations of IL-10 (Figure 8D), CCL17 (Figure 8E), CCL18 (Figure 8F) and CCL22 (Figure 8G) were up-regulated in THP-1 cells undergoing macrophages polarization induced by IFN-γ + LPS that were transfected with hsa_circ_0003528. Also, although the co-transfection of NC siRNA and circ_0003528 exhibited no evident effect upon the expression of the above investigated parameters, when replacing NC siRNA with circ_0003528 siRNA, the dysregulated expressions of CLCL10 (Figure 7B), CXCl11 (Figure 7C), Nos2 (Figure 7D), CCL17 (Figure 7E), CCL18(Figure 7F), TNF-α (Figure 8A), IL-6 (Figure 8B) and IL-12 (Figure 8C), IL-10 (Figure 8D), CCL17 (Figure 8E), CCL18 (Figure 8F) and CCL22 (Figure 8G) was all partly recovered by the presence circ_0003528 siRNA.

Patient recruitment

In this study, we first included 50 patients with active pulmonary TB (as the TB group) and 50 healthy controls (as the HC group). We further validated the correlations between the expression of miRNAs (including miR-224-5p, miR-324-5p, miR-488-5p, miR-587 and miR-668) and the expression of hsa_circ_0003528 among these subjects. The clinical characteristics of the study population were summarized and compared among different groups using Student’s t tests. This study was performed based on the Declaration of Helsinki and Chinese GCP and GLP. Written informed consent forms were acquired from all subjects before the study was started. The entire study protocol was approved by the Ethical Committee of our hospital.

Cell culture, polarization and transfection

THP-1 and U937 cells were cultured in RPMI 1640 medium containing 10% FBS and 1.75μL/500 mL β-mercaptoethanol (Sigma) and differentiated into macrophages by treatment with 5ng/mL phorbol-12-myristate-13-acetate (PMA) (Sigma) overnight.

In cellular experiment 1, the THP-1 and U937 cells were transfected with either scramble control or the candidate miRNA mimics to compare their effect on the expression of circRNA_0003528. Those candidate miRNAs include miR-192, miR-215, miR-217, miR-223, miR-224-5p, miR-324-5p, miR-330-3p, miR-370, miR-421, miR-488-5p, miR-495, miR-548, miR-574-5p, miR-587, miR-606, miR-668.

In cellular experiment 2, the THP-1 and U937 cells were divided into 2 groups, i.e., 1. pcDNA group (THP-1 and U937 cells transfected with an empty vector); and 2. pcDNA-hsa_circ_0003528 group (THP-1 and U937 cells transfected with the vector carrying hsa_circ_0003528).

In cellular experiment 3, the THP-1 and U937 cells were divided into 2 groups, i.e., 1. NC group (THP-1 and U937 cells transfected with a scramble siRNA); and 2. hsa_circ_0003528 siRNA group (THP-1 and U937 cells transfected with hsa_circ_0003528 siRNA).

In cellular experiment 4, the THP-1 and U937 cells were divided into 3 groups, i.e., 1. M (-) group (no macrophage polarization control group); 2. M (IFN-γ + LPS) group (cells were treated with 20 ng/ml IFN-γ (R&D systems, USA) + 100 ng/mL of LPS (sigma) for 18 hours); and 3. M (IL-4) group (cells were treated with 20 ng/mL of IL-4 (PeproTech, USA) for 18 hours).

In cellular experiment 5, the THP-1 and U937 cells were divided into 2 groups, i.e., 1. M (M-CSF) group (control group); and 2. M (GM-CSF) group (cells were treated with 20ng/mL GM-CSF for 18 hours).

In cellular experiment 6, the THP-1 and U937 cells were divided into 2 groups, i.e., 1. M (IFN-γ + LPS) group (cells were treated with 20 ng/ml IFN-γ + 100 ng/mL of LPS for 18 hours); and 2. M (IFN-γ + LPS) to M (IL-4) group (cells in groups were subsequently treated with 20 ng/mL of IL-4 for another 18 hours).

In cellular experiment 7, the THP-1 and U937 cells were divided into 2 groups, i.e., 1. M (IL-4) group (cells treated with 20 ng/mL of IL-4 for 18 hours); and 2. M (IL-4) to M (IFN-γ + LPS) group (cells in groups were subsequently treated with 20 ng/ml IFN-γ + 100 ng/mL of LPS for another 18 hours). The treatment was described as previously [25, 26].

In cellular experiment 8, the THP-1 and U937 cells were divided into 4 groups, i.e., 1. M (IFN-γ + LPS) + pcDNA group (cells treated with 20 ng/mL of IL-4 for 18 hours as well as transfection with empty vector as control); 2. M (IFN-γ + LPS) + pcDNA-circ_0003528 group (cells treated with 20 ng/mL of IL-4 for 18 hours as well as transfection with vector containing hsa_circ_0003528); 3. M (IFN-γ + LPS) + pcDNA-circ_0003528, NC siRNA group (cells treated with 20 ng/mL of IL-4 for 18 hours as well as transfection with vector containing hsa_circ_0003528 and NC siRNA); 2. M (IFN-γ + LPS) + pcDNA-circ_0003528, circ_0003528 siRNA group (cells treated with 20 ng/mL of IL-4 for 18 hours as well as transfection with vector containing hsa_circ_0003528 and circ_0003528 siRNA).

All transfection was done using the Lipofectamine ²⁰⁰⁰ reagent (Invitrogen, Carlsbad, CA) in accordance with the standard protocol provided on the product manual of the manufacturer. The transfected cells were harvested 48 h after transfection to assay the expression of target genes.

Target prediction

The candidate circRNAs in our study were selected by utilization of Circular RNA Interactome (https://circinteractome.irp.nia.nih.gov/) and the candidate miRNAs in our study were selected by utilization of some frequently used target gene prediction databases including miRDB and TargetScan.

RNA isolation and real-time PCR

Real-time PCR was done to assay the expression of hsa_circ_0003528, miR-224-5p, miR-324-5p, miR-488-5p, miR-587, miR-668, and CTLA4 mRNA in each sample. In brief, the collected samples were subject to the treatment with a QIAzol Lysis Reagent (Qiagen, Germantown, MD) and subsequent RNA extraction operations carried out using a miRNeasy Mini assay kit (Qiagen, Germantown, MD) or a miRCURY RNA Isolation kit (Qiagen, Germantown, MD) in accordance with the standard protocol provided on the product manual of the manufacturer to isolate total RNA content in each sample for miRNA and mRNA amplification, respectively. Then, the isolated total RNA samples were reverse transcribed into cDNA templates by using a Universal cDNA Synthesis II assay kit (Qiagen, Germantown, MD) in accordance with the standard protocol provided on the product manual of the manufacturer. Finally, in the step of real time PCR, the reactions were carried out using an ExiLENT SYBR Green Master Mix reagent (Qiagen, Germantown, MD) and microRNA LNA PCR primers (Qiagen, Germantown, MD) on a Light Cycler 480 real time PCR machine (Roche, Nutley, NJ) in accordance with the standard protocol provided on the product manual of the manufacturer. The relative expression of hsa_circ_0003528 (Forward primer: 5’-GTAACCAGCAGCCTGGACTC-3; Reverse primer: 5’-GCAACTTGCTGACCAGAACA-3’), miR-224-5p (Forward primer: 5’-CAAGTCACTAGTGGTTCC-3’; Reverse primer: 5’-GAACATGTCTGCGTATCTC-3’), miR-324-5p (Forward primer: 5’- CATCCCCTAGGGCATTG-3’; Reverse primer: 5’-GAACATGTCTGCGTATCTC-3’), miR-488-5p (Forward primer: 5’-CCCAGATAATGGCACTC-3’; Reverse primer: 5’-GAACATGTCTGCGTATCTC-3’), miR-587 (Forward primer: 5’-TTCCATAGGTGATGAGTC-3’; Reverse primer: 5’-GAACATGTCTGCGTATCTC-3’), miR-668 (Forward primer: 5’-TGTCACTCGGCTCGGC-3’; Reverse primer: 5’-GAACATGTCTGCGTATCTC-3’), and CTLA4 mRNA (Forward primer: 5’-ACGGGACTCTACATCTGCAAGG-3’; Reverse primer: 5’-GGAGGAAGTCAGAATCTGGGCA-3’) in each sample was calculated using the Ct value approach.

Luciferase assay

To study the molecular relationship between hsa_circ_0003528 and miR-224-5p, miR-324-5p and miR-488-5p, we first constructed wild type luciferase vectors by cloning the sequences of hsa_circ_0003528 into separate pcDNA3.1 vectors (Promega, Madison, WI) between BamHI and HindIII. The correctness of inserted sequence was confirmed by using direct sequencing with common primer: 5' CGCAAATGGGCGGTAGGCGTG 3'. In addition, we used a Quick Change II mutagenesis assay kit (Stratagene, San Diego, CA) in accordance with the standard protocol provided on the product manual of the manufacturer to introduce site directed mutations in the hsa_circ_0003528 binding sites of miR-324-5p, miR-224-5p, and miR-488-5p. In the next step, THP-1 and U937 cells were co-transfected with wild type/mutant type hsa_circ_0003528 using Lipofectamine 2000 together with miR-324-5p, miR-224-5p, and miR-488-5p mimics respectively. At 48 post transfection, the relative luciferase signals of transfected THP-1 and U937 cells were measured on a luminometer using a Dual Luciferase reporter gene assay kit (Promega, Madison, WI) in accordance with the standard protocol provided on the product manual of the manufacturer. Similarly, to study the molecular relationship between CTLA4 and miR-324-5p, miR-224-5p, and miR-488-5p, we first constructed wild type luciferase vectors of CTLA4 3’ UTR by cloning the sequences of CTLA4 3’ UTR containing the miR-324-5p, miR-224-5p, and miR-488-5p binding sites into separate pcDNA3.1 vectors. Then, we used the Quick Change II mutagenesis assay kit to introduce site directed mutations in the miR-324-5p, miR-224-5p, and miR-488-5p binding sites of CTLA4 3’ UTR. In the next step, THP-1 and U937 cells were co-transfected with wild type/mutant type CTLA4 3’ UTR vectors along with miR-324-5p, miR-224-5p, and miR-488-5p mimics using Lipofectamine 2000. At 48 post transfection, the relative luciferase signals of transfected THP-1 and U937 cells were measured on a luminometer using the Dual Luciferase reporter gene assay kit.

Western blot analysis

The collected samples were first lysed in a RIPA lysis buffer (Sigma Aldrich, St. Louis, MO) in accordance with the standard protocol provided on the product manual of the manufacturer to extract total protein content, which was then separated on a 10% SDS-PAGE gel and transferred onto nitrocellulose membranes. After being blocked in 5% skim milk for 1 h, the nitrocellulose membranes were sequentially probed with primary anti-CTLA4 antibodies and HRP-conjugated secondary antibodies (Abcam, Cambridge, MA, used in accordance with the incubation conditions suggested by the manufacturer). After protein band development using an enhanced chemiluminescence reagent (Clearness ECL substrate, Bio-Rad Laboratories, CA), the relative protein expression of CTLA4 in each sample was quantified by utilizing the ImageJ v1.4.9 software.

ELISA

The levels of CXCL10, CXCL11, Nos2, CCL17, CCL18, CCL22, TNF-α, IL-6, IL-12, and IL-10 in collected clinical samples were assayed by using commercial ELISA kit (Thermo Fisher Scientific, Waltham, MA) in accordance with the standard protocol provided on the product manual of the manufacturer.

Statistical analysis

The Student’s tests were done using SPSS 20.0 software (IBM, Chicago, IL). P values of < 0.05 were deemed statistically significant.