Blood samples were obtained through the Department of Children's Healthcare from 30 healthy volunteers with no personal or family history of chronic autoimmune, cancer, metabolic, or infectious diseases. The volunteers included 9 males and 21 females, with an average age of 8.82±3.77 years. Blood samples from 30 RF-positive patients with JIA were obtained between May, 2017 and May, 2019 including 11 males and 19 females, with an average age of 8.64±3.58 years at the Children's Hospital of Nanjing Medical University. Peripheral lymphocytes were isolated from blood samples. In brief, peripheral blood was collected from all patients before receiving any therapeutic drugs, and blood was collected from the healthy controls during a physical examination. Peripheral lymphocytes were isolated by adopting the Ficoll-Hypaque density gradient centrifugation method. Children who had previously received disease-modifying anti-rheumatic drug (DMARD) therapy or steroid therapy were excluded. Clinical characteristics were classified according to the detailed diagnostic information obtained from the medical records and physical examinations. All experiments were performed in compliance with government policies and the Helsinki Declaration. All patients or healthy controls had the consent of their legal guardians or parents who signed an informed consent form before collecting blood samples. The present study was approved by the Ethics Committee of the Children's Hospital of Nanjing Medical University.

Human T cells were filtered through a 75 µm strainer and separated by Ficoll centrifugation (800 × g for 20 min at 4°C). The mononuclear cells were resuspended in RPMI-1640 supplemented with 10% FBS. The anti-CD4 magnetic Dynabeads (Invitrogen; Thermo Fisher Scientific, Inc.) were applied to sort the T cells.

Total RNA was isolated from 1×106 T cells and used for the lncRNA/mRNA integrated microarray analysis (CapitalBio). Each group included 3 samples. Sample preparation and microarray hybridization were performed according to the manufacturer's instructions with minor modifications. Briefly, mRNA was purified from total RNA following the removal of rRNA (using the mRNA-ONLY™ Eukaryotic mRNA Isolation kit, EPICENTRE Biotechnologies), amplified and transcribed into fluorescent cRNA along the entire length of the transcripts without 3' bias utilizing a random priming method. The arrays were scanned using an Agilent Scanner (Agilent Technologies, Inc.). Agilent Feature Extraction software (version 11.0.1.1) was used to analyze the acquired array images. Quantile normalization and subsequent data processing were performed using the GeneSpring GX v12.0 software package (Agilent Technologies, Inc.). Following quartile normalization of the raw data, lncRNAs which had flags in present or marginal ('All Targets Value') were selected for further analysis. lncRNA expression patterns were revealed via Hierarchical analysis using Cluster 3.0 software (Stanford University).

The full-length construct of P2X7R, as well as the full-length and mutant constructs of RP11-340F14.6 were synthesized and cloned into pGC-LV plasmid purchased from GenScript Co. Ltd (Nanjing, China). shRNA technology was employed to knockdown the target genes or lncRNA. shRNAs targeting RP11-340F14.6 or P2X7R were designed constructed by GenScript Co. Ltd. and were cloned into the PLL.3.7 vector purchased from GenScript Co. Ltd. and further packaged to produce lentiviral particles, as previously described (23).

The target vectors (20 µg) were mixed with lentiviral packaging 15 µg Δ8.91 (GenScript Co. Ltd.) and envelope expressing 10 µg VSV-G (GenScript Co. Ltd.) plasmids to generate lentiviral particles in 1.2×107/20 ml 293T cells (ATCC) using 100 µl Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.). Viral particles were concentrated by ultracentrifugation and expression vector titers were determined. The plasmids were constructed and transfected using lentivirus. Cells were cultured with 100 µg/ml of human-derived IL-2. The 24-well plates were placed in a cell incubator at 37°C 5% CO₂ and cultured for 48 h. Cells increased in volume after 24 h of stimulation, indicating activation, and were transfected with the lentivirus after 48 h of stimulation. Transfection was performed at a multiplicity of infection (MOI) of 100, using polyberene at a concentration of 5 µg/ml. When the cells reached the optimal transfection state 72 h later, to transfected cells were screened using puromycin. The transfection efficiency was determined by fluorescence intensity and RT-PCR assay.

Following CD4⁺ T cell enrichment, the cells were incubated with human anti-CD3-FITC (cat. no. 557832) and anti-CD4-PerCP monoclonal antibodies (mAbs, cat. no. 564419) in 4°C for 30 min (BD Pharmingen). Cells were fixed and permeabilized with Cytofix/Cytoperm (cat. no. 56422 Human Fc Block from BD Pharmingen) and then intracellularly stained with IL-17A-Phycoerythrin (IL-17A-PE) or IgG-PE as an isotype control. To detect the Treg cell frequency, cells were labeled with anti-CD4-FITC and anti-CD25-APC antibodies (cat. no. 555434, BD Pharmingen). Following fixation and permeabilization, the cells were stained using anti-forkhead box protein 3 (Foxp3)-PE mAb (cat. no. 560046, BD Pharmingen) or IgG-PE control at 4°C for 30 min and were then analyzed using a BD FACSCanto II flow cytometer (BD Pharmingen). The data were analyzed using Cell Quest analysis software, version 5.1 (BD Pharmingen).

RIP assay was carried out using the Magna RIP RNA-Binding Protein Immunoprecipitation kit (EMD Millipore), as previously described (24). Anti-HA antibodies (1:50, ab9110; Abcam) were used for RIP. T cells were either transduced with fixed or varying doses of lentivirus containing RP11-340F14.6 along with lentivirus containing P2X7R. The coprecipitated RNAs were detected by reverse transcription PCR and quantitative (real-time) PCR. Total RNA (input control) and IgG were assayed simultaneously.

Total RNA was isolated from the cells using TRIzol reagent (Invitrogen Life Technologies; Thermo Fisher Scientific, Inc.) and purified using the RNeasy MinElute Clean up kit (Qiagen) according to the manufacturer's instructions. cDNA was synthesized from total RNA using the random priming method using the One step PrimeScript kit (RR064A, Takara Bio, Inc.). Transcript levels were measured in duplicate by qPCR (ABI 7900; Life Technologies; Thermo Fisher Scientific, Inc.). The primer for RP11-340F14.6 was as follows: Forward, 5'-GCC AAG CTT CTT GAA AGG CC-3' and reverse, 5'-TTC CAC GGA GTA GAG CGA GTC-3'. Primer sequences synthesized by GenScript Co. Ltd. The amplification procedure was 95°C for pre-denaturation for 30 sec; 95°C for 5 sec, 60°C for 31 sec (45 cycles); dissolution curves 95°C for 15 sec, 60°C for 60 sec, 95°C 15 sec. The relative expression of lncRNA and mRNA was normalized to GAPDH and was calculated using the 2^−ΔΔCq method as previously described (25). The primer sequences for GAPDH were as follows: Forward, 5'-AAG GTG AAG GTC GGA GTC AAC-3' and reverse, 5'-GGG GTC ATT GAT GGC AAC AAT A-3'.

Whole cell lysates were prepared as previously described (24). Protein was extracted using RIPA lysis buffer (Beyotime Institute of Biotechnology). Protein concentration was determined by the BCA method (Beyotime Institute of Biotechnology). Equal amounts of proteins (20 µg) were boiled, separated on 10% SDS-PAGE and transferred onto PVDF membranes. After blocking with 5% (w/v) non-fat dry milk, the membranes were probed with the primary antibody overnight at 4°C. The secondary antibodies were both horseradish peroxidase (HRP)-conjugated IgG including anti-mouse IgG (ab97040), anti-rabbit IgG (ab7090) and anti-sheep IgG (ab6747) (all from Abcam) and used at a dilution of 1:1,000 and incubation at room temperature for 1 h. Signals were detected by the chemiluminescence procedure (Pierce; Thermo Fisher Scientific, Inc.) with BioMax films (Kodak) and visualized using an ECL kit (EMD Millipore). GAPDH was applied as the reference protein. Antibodies, including P2X7R (ab48871), GAPDH (ab181602), HA-tag (ab18181) were purchased from Abcam and used at a dilution of 1:1,000.

Data are presented as the means ± SEM. Differences between 2 groups were analyzed using the Student's t-test. ANOVA was performed to evaluate differences between multiple groups followed by Tukey's post hoc test. Expression experiments were repeated at least 3 times with samples in triplicates. Pearson's correlation analysis was used to analyze the correlation between the expression of RP11-340F14.6 and that of associated factors [such as retinoic acid-related orphan receptor gamma t (RORγt), Foxp3 and P2X7R]. Statistical analysis was performed using STATA 10.0 software and presented using GraphPad Prism software (GraphPad Software). In all cases, P<0.05 was considered to indicate a statistically significant difference.

It has been previously demonstrated that an imbalance in the immune microenvironment is highly associated with the risk of developing JIA, particularly as regards T cell reprogramming. In the present study, peripheral blood T cells were first extracted from patients with JIA and healthy controls. The distribution of human Th17 and Treg immune cells sorted from JIA blood samples was first analyzed. Using CD4 and IL-17 as markers, a marked increase was identified in the percentage of Th17 cells among the total number of T cells in the JIA samples compared to those from healthy children (Fig. 1A). On the contrary, there was a reduced percentage of Tregs labeled with CD25 and Foxp3 presented in the children with JIA compared to the healthy controls (Fig. 1B). These findings suggest an increase in the Th17/Treg ratio in the immune microenvironment of children with JIA.

Peripheral blood mononuclear cells (PBMCs) were isolated from the JIA samples and matched with the healthy controls. Using anti-CD4 magnetic beads, T cells were sorted in PBMCs that were extracted from 6 pools of paired samples. A high-throughput microarray of lncRNAs was applied to screen for differential expression profiles between the JIA and control samples. The aberrant expression of lncRNA was presented by hierarchical clustering using a heatmap. The profile of the differential expression of lncRNAs in T cells of children with JIA was obtained (Fig. 2A). Among these, lncRNAs were further filtered using the following criteria: i) A fold-change cut-off of 4/0.25; ii) Cq value >25 for PCR detection; iii) detection of at least 75% in all samples. There were 138 lncRNAs that met these criteria. Finally, 20 of these 138 lncRNAs with the most significant P-values and q values were labeled as candidates. A larger sample size including 20 JIA and 20 paired controls was used for further validation. Among the 20 candidate lncRNAs, LINC01225 presented no expression and therefore the data column for this candidate was removed (Fig. 2B). A total of 4 lncRNAs (LINC00471-001, ZPAS1-002, NEAT1 and RP11-340F14.6) exhibited a significantly altered expression in the JIA samples compared to the normal controls, with RP11-340F14.6 expression exhibiting the most significant difference.

To further investigate the aberrant expression of RP11-340F14.6 in JIA, a case-control experiment as performed with 30 RNA samples extracted from children with JIA and 30 paired control samples. An RT-qPCR assay was conducted and the increased expression of RP11-340F14.6 in patients with JIA was validated (Fig. 2C). Furthermore, a higher presentation of RORγt combined with a decreased Foxp3 expression were observed in the JIA group (Fig. 2D and E). Pearson's correction analysis also revealed that RP11-340F14.6 expression positively correlated with RORγt expression and negatively correlated with Foxp3 expression (Fig. 2F). Based on the preliminary data, it was thus hypothesized that RP11-340F14.6 may be associated with the increase in the Th17/Treg ratio in the immune microenvironment of JIA.

Subsequently, a series of in vitro experiments were conducted to investigate the function of RP11-340F14.6 in CD4⁺ T cells. CD4⁺ T cells were sorted by flow cytometry from PBMCs of patients with JIA and RP11-340F14.6 was overexpressed using a lentivirus in these cells. It was found that the overexpression of RP11-340F14.6 induced the expression of IL-17 and increased the percentage of Th17 cells, which is defined by a CD4-positive cell subgroup (Fig. 3A). The ectopic expression of RP11-340F14.6 also resulted in a decreased expression of Foxp3 and a decrease in the number of Tregs which were labeled with CD25 (Fig. 3B). In the RP11-340F14.6^high CD4⁺ T cells from PBMCs, endogenous RP11-340F14.6 expression was silenced using an shRNA lentivirus. The cells in which RP11-340F14.6 was silenced exhibited a lower percentage of Th17 cells and a greater distribution of Tregs (Fig. 3C and D).

According to the results described above, a functional role of RP11-340F14.6 in JIA was identified. However, since RP11-340F14.6 has been poorly investigated in human diseases, the detailed mechanisms of action of this lncRNA remain unclear. Thus, in the present study, detailed information on RP11-340F14.6 was obtained using the UCSC genome browser (http://genome.ucsc.edu/). Based on the FLANK10K theory of lncRNA-mRNA interaction (26), neighbors of RP11-340F14.6 labeled with the 10K region were identified, including P2X7R, P2X4R and CAMKK2 (Fig. 4A). These neighbors are likely to interact with RP11-340F14.6 through a cis-regulation approach. Subsequently, the expression of these neighbors was detected in RP11-340F14.6^high CD4⁺ T cells derived from PBMCs following transfection with RP11-340F14.6 shRNA. Of note, it was found that P2X7R expression was decreased with the loss of RP11-340F14.6 (Fig. 4B). In human clinical samples, the mRNA expression of P2X7R was increased in the JIA immune microenvironment (Fig. 4C) and its expression positively correlated with RP11-340F14.6 expression (Fig. 4D).

The phenotype for RP11-340F14.6 expression in T cell differentiation was also measured to examine whether the function of RP11-340F14.6 was mediated through interaction with P2X7R. The expression of P2X7R was knocked down using shRNA technology and the percentage of Th17 and Treg cells was determined. The overexpression of RP11-340F14.6 increased the amount of Th17 cells; however, this increase was attenuated by the loss of P2X7R expression (Fig. 5A). The percentage of Tregs was decreased by the overexpression of RP11-340F14.6, but was restored with the loss of P2X7R expression (Fig. 5B). These results indicated that the overexpression of RP11-340F14.6 may have increased the Th17/Treg ratio via a P2X7R-dependent mechanism.

The expression of P2X7R mRNA and protein was detected in cells following overexpression and/or silencing of RP11-340F14.6. The increased expression of P2X7R was observed in cells which overexpressed RP11-340F14.6, and this increase in P2X7R mRNA expression was suppressed by the knockdown of RP11-340F14.6 expression (Fig. 6A and B). CatRAPID, a bioinformatics software (http://service.tarta-glialab.com/page/catrapid_group), was used to predict the potential binding fragment of P2X7R in RP11-340F14.6 (Fig. 6C). The full length of lncRNA nucleotide sequence and the full length of P2X7R amino acid peptide were used as input. The RP11-340F14.6 region between amino acids 301 to 352 was predicted to bind to P2X7R. For P2X7R, the intracellular (IC) domain and amino acids between 501 and 522 were suggested to be the most probable binding domain for RP11-340F14.6. RIP assays revealed that anti-HA (wild-type P2X7R) antibodies specifically precipitated RP11-340F14.6; however, deletions in the amino acids 501-522 of P2X7R abrogated its ability to bind to RP11-340F14.6 (Fig. 6D). These results suggested that RP11-340F14.6 interacted with P2X7R in a highly specific manner and that the predicted region (301-352) of RP11-340F14.6 is crucial for its ability to interact with P2X7R.