Association of LncRNA‐GAS5 gene polymorphisms and PBMC LncRNA‐GAS5 level with risk of systemic lupus erythematosus in Chinese population

Abstract Growth arrest‐specific 5 (GAS5) is a kind of long non‐coding RNAs (lncRNAs). Previous studies showed that down‐regulation of LncRNA‐GAS5 was involved in the development of systemic lupus erythematosus (SLE). However, the regulatory mechanism of down‐expressed LncRNA‐GAS5 in SLE remains obscure. In this study, we aimed to investigate the association of LncRNA‐GAS5 polymorphism with SLE risk. And further explore how LncRNA‐GAS5 is involved in the occurrence of SLE. Here, we evaluated the relationship between the risk for the development of SLE and the 5‐base pair (AGGCA/‐) insertion/deletion (I/D) polymorphism (rs145204276) in the LncRNA‐GAS5 promoter region. A custom 36‐Plex SNPscan kit was used for genotyping the LncRNA‐GAS5 polymorphisms. The LncRNA‐GAS5 and miR‐21 target prediction was performed using bioinformatics software. Enzyme‐linked immunosorbent assay (ELISA) and quantitative real‐time PCR (qRT‐PCR) were performed to assess GAS5 and miR‐21 mRNA expression and PTEN protein expression. The results revealed that rs145204276 resulted in a decreased risk of SLE (DD genotypes vs II genotypes: adjusted OR = 0.538, 95% CI, 0.30‐0.97, P = .039; ID genotypes vs II genotypes: adjusted OR = 0.641, 95% CI, 0.46‐0.89, P = .007; ID/DD genotypes vs II genotypes: adjusted OR = 0.621, 95% CI, 0.46‐0.84, P = .002; D alleles vs I alleles: adjusted OR = 0.680, 95% CI, 0.53‐0.87, P = .002). A reduced incidence of renal disorders in SLE was found to be related to ID/DD genotypes and D alleles (ID/DD genotypes vs II genotypes: OR = 0.57, 95% CI, 0.36‐0.92, P = .020; D alleles vs I alleles: OR = 0.63, 95% CI, 0.43‐0.93, P = .019). However, no significant association of rs2235095, rs6790, rs2067079 and rs1951625 polymorphisms with SLE risk was observed (P > .05). Additionally, haplotype analysis showed that a decreased SLE risk resulted from the A‐A‐C‐G‐D haplotype (OR = 0.67, 95% CI, 0.49‐0.91, P = .010). Also, patients in the SLE group showed a down‐regulated expression of LncRNA‐GAS5 and PTEN than the healthy volunteers; however, patients with rs145204276 ID/DD genotypes showed up‐regulated expression of LncRNA‐GAS5 and PTEN compared with patients carrying the II genotype. Furthermore, the miR‐21 levels were considerably up‐regulated in the SLE group than the healthy volunteers, and patients with rs145204276 ID/DD genotype had lower miR‐21 levels than the ones with the II genotype. Thus, we found that the LncRNA‐GAS5/miR‐21/PTEN signalling pathway was involved in the development of SLE, where LncRNA‐GAS5 acted as an miR‐21 target, and miR‐21 regulated the expression of PTEN. These findings indicated that the rs145204276 ID/DD genotypes in the LncRNA‐GAS5 gene promoter region may be protected against SLE by up‐regulating the expression of LncRNA‐GAS5, which consecutively regulated miR‐21 and PTEN levels.


| INTRODUC TI ON
Systemic lupus erythematosus (SLE) is a chronic autoimmune disorder that may damage multiple tissue/organs by generating autoantibodies and immune complexes. 1,2 It is a heterogeneous disease with diverse clinical characteristics. 3 Most SLE patients exhibit a range of symptoms, including malar rash, nephritis, arthritis and neurologic disorders, which significantly reduce the quality of life. 4,5 Previous studies have shown that the development of SLE results from an imbalanced immune system, including aberrant excretion of cytokines, along with the modified immune cellular responses. [6][7][8][9][10] The pathogenesis of SLE involves disorders related to genetics, immunity, environment and hormones; however, the precise mechanism of pathogenesis is unclear. 2 Thus, Understanding the pathways of development of SLE would help in the identification of novel therapies.
LncRNAs are non-protein coding transcripts that are > 200 nucleotides in length. 11 Recent studies have shown that lncRNAs, which were previously considered as mere transcriptional noise, regulate gene expression at the transcriptional levels. 12 The pathogenesis of several autoimmune inflammatory disorders, such as rheumatoid arthritis, SLE, multiple sclerosis, ulcerative colitis, hyperthyroidism and chronic liver disease, are known to involve lncRNAs. [13][14][15] The aberrant expression of GAS5, a new type of ln-cRNAs, has been reported in SLE patients and animal models. 16,17 The lncRNA-mRNA co-expression analysis showed that LncRNA-GAS5, lnc0640 and lnc5150 were involved in SLE pathogenesis through the mitogen-activated protein kinase pathway (MAPK), and the LncRNA-GAS5 in plasma could be used as SLE biomarkers. 16 In addition, the chromosomal locus of LncRNA-GAS5, 1q25, has been shown to be related with human SLE development in genetic studies. 18 MicroRNAs (miRNAs) are non-coding RNAs molecules, containing approximately 21-23 nucleotides, which promote translational repression or degradation of their target mRNAs to regulate gene expression. 6 miRNAs have been identified as novel biomarkers and potential therapeutic targets in several diseases, and they perform diverse immunoregulatory functions. 19 Growing evidence has shown that the abnormal expression of miR-21 is a contributor to SLE. 6,20 Here, we performed a bioinformatics analysis for target gene prediction and identified the probable binding site between miR-21 and LncRNA-GAS5. Furthermore, we identified the phosphatase and tensin homolog (PTEN) as the direct miR-21target.
Several genome-wide association studies (GWAS) have revealed the frequent use of single nucleotide polymorphisms (SNPs) as genetic markers. 21 SNPs associated with lncRNAs are known to impact an individual's susceptibility to autoimmune diseases. For example, the lnc0640 gene rs13039216 TT genotype resulted in a reduced risk of rheumatoid arthritis (RA) and the G allele of rs141561256 in lnc5150 gene was significantly associated with the rheumatoid factor in RA patients. 22 Furthermore, the variation at rs13259960 A>G, which weakened the STAT1 recruitment to the enhancer that looped to the SLEAR promoter, caused a reduced SLEAR expression in SLE patients. 23 However, there are no published reports on the link between the LncRNA-GAS5 gene polymorphisms and the SLE risk. This case-control study investigated whether LncRNA-GAS5 gene polymorphisms contributed to the development of SLE. Additionally, we assessed the impact of LncRNA-GAS5 polymorphisms on LncRNA-GAS5, miR-21 and PTEN levels.

| Study subjects
We recruited 302 SLE patients (63 male and 239 female patients, average age 38.77 ± 14.41 years) and 396 age and gender-matched and PTEN than the healthy volunteers; however, patients with rs145204276 ID/DD genotypes showed up-regulated expression of LncRNA-GAS5 and PTEN compared with patients carrying the II genotype. Furthermore, the miR-21 levels were considerably up-regulated in the SLE group than the healthy volunteers, and patients with rs145204276 ID/DD genotype had lower miR-21 levels than the ones with the II genotype. Thus, we found that the LncRNA-GAS5/miR-21/PTEN signalling pathway was involved in the development of SLE, where LncRNA-GAS5 acted as an miR-21 target, and miR-21 regulated the expression of PTEN. These findings indicated that the rs145204276 ID/DD genotypes in the LncRNA-GAS5 gene promoter region may be protected against SLE by up-regulating the expression of LncRNA-GAS5, which consecutively regulated miR-21 and PTEN levels.

K E Y W O R D S
LncRNA-GAS5, miR-21, PTEN, single nucleotide polymorphism, systemic lupus erythematosus healthy volunteers (106 male and 290 female volunteers, average age 39.85 ± 11.79 years) from the Affiliated Hospital of Youjiang Medical University for Nationalities and People's Hospital of Baise between June 2016 and September 2018. Written informed consent was obtained from all patients before study initiation. The study design was sanctioned by the ethics committee of the hospital. All patients met the SLE classification standards revised by the American rheumatology Society (ACR) in 1997. 24 The clinical characteristics were collected based on medical records or questionnaires and were reviewed by senior doctors. We collected clinical data for diverse features, including malar rash, arthritis, leucopenia, renal disorder, thrombocytopenia, photosensitivity, neurological disorder, ribonucleoprotein antibody (Anti-RNP), antinuclear antibody (ANA), smith antibody (Anti-Sm), double-stranded DNA antibody (Anti-dsDNA), complement 3 (C3) and complement 4 (C4) antibodies. The healthy volunteers were unrelated and did not have any history of SLE, autoimmune diseases, cancer or other inflammatory diseases.
We isolated genomic DNA from peripheral blood samples and stored at −80°C. Online primer v3.0 was used for designing the PCR primers (http://prime r3.ut.ee/) and synthesized by Shanghai Genesky Biotechnologies Inc (Table S1). The total volume of ligation reaction cycles; 68°C 60 minutes and 4°C forever. The multiplex PCR product was diluted with 10 times of ddH 2 O, mixed with 1.0 μL of liz500 and 8.9 μL Hi-Di, denatured at 95°C for 5 minutes. ABI3500 sequencer (ABI) was used for genotyping, and GeneMapper 4.1 software was used to analyse the original data. For quality control (QC), we randomly selected 5% of the samples for Sanger sequencing and found the results to be 100% accurate and consistent.

| Extraction of RNA and quantitative real-time polymerase chain reaction (qRT-PCR)
We collected peripheral blood samples (3 mL) from each patient in an EDTA-containing tube, followed by the isolation and purification of peripheral blood mononuclear cells (PBMCs) using the Ficoll-Hypaque density gradient centrifugation. The TRIzol reagent was used for total RNA extraction from the PBMCs (Invitrogen), and a Hangzhou Nano-300 spectrophotometer was used to measure the

| Serum PTEN determination
We obtained blood samples from both SLE patients and healthy volunteers. After clotting at room temperature, the serum was isolated and stored at −80°C for subsequent testing. The concentration of serum PTEN was measured using appropriate ELISA kits (Human PTEN ELISA Kit), and absorbance (OD value) was measured at 450 nm using an RT-6000 enzyme micro-plate reader. A standard calibration curve was used to determine the concentration of serum PTEN ( Figure S3).

| Statistical analysis
The SPSS software v17.0 was used for data analysis.
After Bonferroni correction, no significant association between the rs145204276 DD genotype and the risk of SLE was observed, the significance of multiplex detection at the P < .01 (0.05/5) level was retained for rs145204276 ID genotype, D allele and dominant model.

| Association of rs145204276 polymorphism with clinical features
We further stratified the distribution of rs145204276 polymorphism in thirteen specific clinical features in both SLE positive and negative

| Haplotype analysis of the LncRNA-GAS5 polymorphisms with risk of SLE
Next, we assessed the haplotype frequencies of the five SNPs in the LncRNA-GAS5 gene in both the SLE patients and the healthy volunteers using the online SHEsis software (Table 4). Approximately, 32.1% and 28.3% of the maximum haplotype (G-G-C-G-I) were observed in SLE patients and the healthy volunteers, respectively. We also found that reduced risk for SLE was related to the presence of the haplotype (A-A-C-G-D) (OR = 0.67, 95% CI, 0.49-0.91, P = .010).
After Bonferroni correction, the significant association between the haplotype (A-A-C-G-D) and the risk of SLE was observed.

| Target gene prediction and LncRNA-GAS5, miR-21 and PTEN expression
The potential miR-21 targets were predicted through an online bioinformatics software. We identified that there were several base pair binding sites between miR-21 and LncRNA-GAS5 (Figure 2A).
We further identified PTEN as the miR-21 target protein ( Figure 2B).
Next, qRT-PCR and ELISA were performed to compare LncRNA-GAS5, miR-21 and PTEN levels in SLE patients having different rs145204276 genotypes to explore the molecular mechanism between rs145204276 polymorphism and the development of SLE.
SLE patients showed considerably reduced LncRNA-GAS5 levels than the healthy volunteers (P = .001; Figure 3A). We further analysed whether rs145204276 genetic polymorphism could affect the expression level of the LncRNA-GAS5 gene in SLE patients.

Additionally, PTEN expression followed a trend similar to that of
LncRNA-GAS5, that is SLE patients showed a substantially reduced .085

TA B L E 3 Association between genotypes and alleles frequencies in rs145204276 with clinical features in SLE patients
Note: The P* value indicates statistical significance after Bonferroni correction.
Abbreviations: CI, confidence interval; OR, odds ratio. PTEN expression than the healthy volunteers (P < .001; Figure 3C) and the rs145204276 ID/DD genotype had elevated PTEN levels than the rs145204276 II genotype (P < .001; Figure 3D). On the contrary, in SLE patients, we observed significantly up-regulated miR-21 expression compared with the healthy volunteers (P < .001; Figure 3E). Furthermore, rs145204276 ID/DD genotypes had lower miR-21 levels compared with the rs145204276 II genotype (P < .001; Figure 3F). The expression levels of GAS5, miR-21 and PTEN in the controls and SLE patients were not significantly different between male and female (P > .05; Figure S4). In addition, spearman's rank correlation analysis showed that there was a correlation between the expression levels of GAS5, miR-21 and PTEN (Table S2). will enhance our understanding of the molecular interaction involved in this field more clearly and draw more powerful conclusions.

| D ISCUSS I ON
Thus, we demonstrated that the rs145204276 ID/DD genotypes in the promoter region of the LncRNA-GAS5 gene acted as a protective factor towards the development of SLE, most probably by elevating LncRNA-GAS5 expression. Additionally, LncRNA-GAS5 may contribute to SLE in the pathogenesis by targeting PTEN through competitive binding to miR-21. Thus, our findings provide a new direction to further our understanding of the pathogenesis mechanism and the role of LncRNA-GAS5 in SLE. In the future, a larger sample size and different ethnic groups will be studied to confirm these findings.