Full high-throughput sequencing analysis of differences in expression profiles of long noncoding RNAs and their mechanisms of action in systemic lupus erythematosus

Abstract

Background: The specific function of long noncoding RNAs (lncRNAs) in systemic lupus erythematosus (SLE) and the mechanism of their involvement in related pathological changes remain to be elucidated, so, in this study, we analyzed the differences in the expression profiles of lncRNAs and their mechanisms of action in SLE using full high-throughput sequencing, bioinformatics, etc. methods.

Methods: We used high-throughput sequencing to detect differences in the expression profiles of lncRNAs, miRNAs, and mRNAs in PBMCs from patients with SLE at the genome-wide level. Next, we predicted target genes of 30 lincRNAs (long intergenic noncoding RNAs) by constructing a coexpression network of differential lincRNAs and mRNAs and identified the role of lincRNAs. Then, we analyzed the coexpression network of 23 optimized lincRNAs and their corresponding 353 miRNAs, evaluated the cis- and trans-effects of these lincRNAs, and performed GO and KEGG analyses of target genes. We also selected 8 lincRNAs and 2 newly discovered lncRNAs for q-PCR validation and lncRNA-miRNA-mRNA analysis. Finally, we also analyzed respectively the relation between lncRNAs and gender bias in SLE patients using RT-qPCR, the relation between Systemic Lupus Erythematosus Disease Activity Index score and the "IFN signature" using ELISA, and the relation between the differential expression of lncRNAs and a change in the number of a cell type of PBMCs in SLE patients using RT-qPCR.

Results: The profiles of 1087 lncRNAs, 102 miRNAs, and 4101 mRNAs in PBMCs significantly differed between patients with SLE and healthy controls. The coexpression network analysis showed that the network contained 23 lincRNAs and 353 mRNAs. The evaluation of the cis- and trans-effects showed that the 23 lincRNAs acted on 704 target genes. GO and KEGG analyses of the target genes predicted the biological functions of the 23 lincRNAs. q-PCR validation showed 7 lincRNAs and 2 novel lncRNAs were identical to the sequencing results. The ceRNA network contained 7 validated lincRNAs, 15 miRNAs, and 155 mRNAs. In addition, the differential expression of lncRNAs may be gender dependent in SLE patients, SLE patients also exhibit a robust "IFN signature," and PBMCs exhibiting differential expression of lncRNAs may be due to a change in the number of a cell type.

Conclusion: This work determined specific lncRNAs that play important biological functions in the pathogenesis of lupus and provided a new direction for diagnosis and treatment of disease.

Keywords: Expression profiles; High-throughput sequencing analysis; Long noncoding RNA; Mechanisms; Systemic lupus erythematosus.

Fig. 1

High-throughput sequencing to reveal differences in the expression profiles of lncRNAs, mRNAs, and miRNAs between the SLE and control groups. a Differentially expressed lncRNA volcano maps between the SLE and NC groups. b Differentially expressed mRNA volcano plots between the SLE and NC groups. c Differentially expressed miRNA scatter maps between the SLE and NC groups

Fig. 2

Cluster heat map analysis. a Thirty screened lincRNAs. b One hundred thirty-two novel annotated lncRNAs

Fig. 3

The qRT-PCR validation of 8 selected lincRNAs and 2 novel lncRNAs between 30 SLE patients and healthy controls. a Relative expression of 4 upregulated lincRNAs. b Relative expression of 4 downregulated lincRNAs. c The relative expression of upregulated novel lncRNA TCONS00195779_1. d The relative expression of upregulated novel lncRNA TCONS00027049_1

Fig. 4

GO and KEGG analyses of target genes. a–c The highest enrichment GO project among the 30 major entities. d, e KEGG pathway analyses of target genes: 27 pathways with significant functional changes (including 6 parts of cellular processes, namely, environmental information processing, genetic information processing, human diseases, metabolism, and organismal systems)

Fig. 5

SLE patients exhibit a strong gender bias and a robust “IFN signature.” a TSIX expressions in PBMCs of 19 patients with SLE and 37 healthy donors (NC) were analyzed by qPCR. b TSIX expressions in PBMCs of 13 male SLE patients and 26 female SLE patients were analyzed by qPCR. **P < 0.01, ***P < 0.001, ****P < 0.0001. c IFN-α expressions in the serum of 12 patients with inactive SLE, 21 patients with active SLE, and 20 healthy donors (NC) were analyzed by ELISA. ***P < 0.001, ****P < 0.0001. d Nonparametric correlation (Spearman) was performed to assess the correlation between IFN-α and SLEDAI score in patients with SLE (n = 33)

Fig. 6

The relative expression of lncRNA NR_034053.2 was determined by qPCR in SLE patients and healthy donors (NC). a Expression of lncRNA NR_034053.2 in PBMCs of SLE patients and healthy donors (NC) was determined by qPCR analysis. b Expression of lncRNA NR_034053.2 in monocytes, B cells, and T cells from SLE patients. c lncRNA NR_034053.2 expression was upregulated in DCs of SLE patients compared with healthy donors. *P < 0.05, ****P < 0.0001

Fig. 7

The mapping network of lncRNA-miRNA-mRNA interactions between 3 upregulated lncRNAs, 12 downregulated miRNAs, and 57 upregulated mRNAs