| BackgroudSystemic lupus erythematosus?SLE?is a multi-system autoimmune disease of unknown etiology,with a complex genetic basis that involves protein-coding genes and non-coding genes.Over the past decades,pervasive studies mainly focused on the roles of the protein-coding genes in disease pathogenesis.By contrast,in the whole genome,much less is known about the effect of non-coding RNA?ncRNA?in the physiological and pathological responses.According to the transcript size,ncRNAs are conventionally classified into two major classes: small ncRNA??200 nt?,such as microRNA?mi RNA?,and long ncRNA?>200 nt?.Many studies have shown that mi RNAs play a critical effect in the pathogenesis of SLE and can be served as biomarkers for the diagnosis and prognosis of SLE.Unlike mi RNAs,long ncRNAs?lncRNAs?are highly diverse and comprise the bulk of the human non-coding transcriptome.LncRNAs can regulate gene expression in many ways,including chromatin remodeling,transcriptional control,post-transcriptional processing and protein metabolism.Moreover,a number of lncRNAs have increasingly been confirmed to play crucial roles in the pathogenesis of many diseases.Recently,two lncRNAs,linc0949 and NEAT1?nuclear enriched abundant transcript 1?,have been reported to be significantly dysregulated in peripheral blood mononuclear cells?PBMCs?of SLE patients,and they were both correlated with systemic lupus erythematosus disease activity index-2000?SLEDAI-2K?score and lupus nephritis?LN?.NEAT1 can regulate expression of inflammatory cytokines and chemokines through the mitogen-activated protein kinase?MAPK?pathway,thus play a role in the pathogenesis of SLE.Due to the heterogeneous clinical manifestations of SLE patients,their unpredictable disease course and difficulty in the early detection,it is urgent to explore novel biomarkers with higher sensitivity and specificity for better diagnosis and prognosis of SLE.A number of studies have confirmed that plasma lncRNAs are stable,and they can be used as early biological effect markers and therapeutic targets in human diseases,such as cancer,cardiovascular diseases,for early screening,diagnosis and prognostic assessment.However,to date,the research on lncRNA in SLE patients is still limited and the potential of plasma lncRNAs as biomarkers for SLE has never been reported.ObjectivesIn this study,we screened plasma lncRNAs and validated the candidate lncRNAs by quantitative reverse transcription polymerase chain reaction?qRT-PCR?,with the intention to identify a panel of lncRNAs that might serve as novel biomarkers for auxiliary diagnosis of SLE.Based on bioinformatics analysis,we predicted the potential target genes and the regulatory network of mRNA-miRNA-lncRNA of the differentially expressed lncRNAs,with the intention to explore their biological functions in the pathogenesis of SLE.MethodsThe study was divided into two parts:Part One: Screening,verification of differentially expressed lncRNAs,and their potential value as biomarkers for SLEHerein,we performed a four-phase case-control study to identify a panel of lncRNAs that might serve as novel biomarkers for auxiliary diagnosis of SLE.Phase 1: In screening phase,we collected plasmas from 24 new-onset SLE patients?12 SLE without nephritis and 12 LN?and 12 normal controls.Then,to screen the differentially expressed lncRNAs,we performed lncRNA profiling using ArrayStar Human LncRNA Microarray in 9 pooled plasma samples?including 3 SLE without nephritis pooling samples,3 LN pooling samples and 3 normal controls pooling samples?.Phase 2: In preliminary validation phase,10 differentially expressed lncRNAs from the screening phase and 5 lncRNAs?ENST00000500597: linc0597,ENST00000500949: linc0949,ENST00000449289: GAS5 or lnc9289,ENST00000587298: lnc-DC,ENST00000495032: HOTAIRM1?from published studies were first tested as candidate lncRNAs in individual sample in the screening phase by qRT-PCR assay.Phase 3: In large-scale independent validation phase,the additional plasma samples from 240 SLE patients and 120 normal controls were collected,and then were randomly divided into training set?including 160 SLE patients and 80 normal controls?and testing set?including 80 SLE patients and 40 normal controls?.Firstly,the identified differentially expressed lncRNAs from preliminary validation phase were evaluated by qRT-PCR in training set.Then,the expression levels of the identified lncRNAs from training set were further assessed by qRT-PCR in testing set.Finally,we explored the association of plasma expression levels of the differentially expressed lncRNAs with clinical parameters of SLE patients.To further evaluate the potential value of identified lncRNAs individually or panel as novel biomarkers for SLE,receiver operating characteristic?ROC?curves analysis were respectively performed on data from training set,testing set and combined set.Based on the data of training set,a discrimination model of lncRNAs panel was constructed to predict the risk of diagnosis with SLE by stepwise logistic regression,and validated in testing set.Finally,according to whether or not nephritis,we divided the 240 SLE patients into LN patients and SLE without nephritis patients.Based on the data of large-scale independent validation phase,a discrimination model of lncRNAs panel was constructed to predict the risk of diagnosis with LN by stepwise logistic regression,and ROC curves analysis were respectively performed to explore the potential value of the differentially expressed lncRNAs individually or panel as novel biomarkers for distinguishing LN from SLE without nephritis.Phase 4: In the external validation phase,to explore the specificity of differentially expressed lncRNAs as candidate biomarkers for SLE,the expression levels of lncRNAs from phase 3 were assessed by qRT-PCR in the plasma samples from 61 disease controls,including 30 rheumatoid arthritis?RA?patients and 31 primary Sjogren's syndrome?pSS?.We further verified the diagnostic performances of the established lncRNAs panel by using ROC curves analysis.Part Two: Bioinformatics study on differentially expressed mRNAs and lncRNAs in SLETo preliminarily explore the biological functions of the differentially expressed lncRNAs in SLE,Firstly,based on the differentially expressed mRNA profile using by ArrayStar Human LncRNA Microarray,we conducted gene ontology?GO?analysis,Kyoto encyclopedia of genes and genomes?KEGG?Pathway analysis.And then in combination with the results of validation by qRT-PCR,we established lncRNA-mRNA co-expression network analysis.Finally,we used competing endogenous RNAs?ceRNAs?analysis to construct the mRNA-miRNA-lncRNA network.ResultsPart One: Screening,verification of differentially expressed lncRNAs,and their potential value as biomarkers for SLEIn the present study,we established the profiles of differentially expressed lncRNAs and mRNAs in SLE without nephritis and LN by ArrayStar Human LncRNA Microarray?fold change ? 2 and P ? 0.05?.Ten differentially expressed lncRNAs via microarray and five lncRNAs from published studies then were subjected to individual qRT-PCR confirmation in original sample.Among 15 candidate lnc RNAs?ENST00000450640: lnc0640;ENST00000583643: lnc3643;ENST00000584688: lnc4688;ENST00000425150: lnc5150;ENST00000506655: lnc6655;NR027074: lnc7074;ENST00000507514: lnc7514;ENST00000458228: lnc8228;ENST00000519603: lnc9603;uc001agf.1: lncagf.1;GAS5;linc0597;linc0949;lnc-DC;HOTAIRM1?,the plasma levels of 9 lncRNAs showed significant change between 24 SLE patients and 12 normal controls?all P < 0.05?.These 9 lncRNAs are GAS5,linc0597,lnc0640,lnc5150,lnc3643,lnc6655,lnc7074,lnc7514 and lncagf.1.To validate the results of these 9 differentially expressed lncRNAs,we conducted a validation study in another larger independent cohort containing training set and testing set.Both in training set and testing set,the levels of plasma linc0597,GAS5,lnc0640,lnc5150 and lnc7074 significantly altered in SLE patients compared with normal controls?all P < 0.05?,and the levels of plasma lnc7074,lnc3643,lnc0640,lnc7514 and lnc6655 were all overexpressed in LN compared with SLE without nephritis?all P < 0.05?.We additionally explored the association of plasma linc0597,GAS5,lnc0640,lnc5150 and lnc7074 levels with clinical characteristics of 240 SLE patients.The expression level of linc0597 and GAS5 were both significantly correlated with C3 level?P < 0.001 and P = 0.009,respectively?;the expression level of lnc5150 level was significantly associated with thrombocytopenia and C3 level?P = 0.008 and P = 0.001,respectively?;the expression level of lnc0640 level was significantly associated with low complement,thrombocytopenia and C3 level?P = 0.001,P = 0.017 and P < 0.001,respectively?;the expression level of lnc7074 level was significantly associated with thrombocytopenia and C3 level?P = 0.028 and P < 0.001,respectively?.From the data of training set,the area under the curve?AUC?of linc0597,GAS5,lnc0640,lnc5150 and lnc7074 was 0.634,0.824,0.598,0.625 and 0.602,respectively.The discrimination model of lncRNAs panel was constructed to predict the risk of diagnosis with SLE by stepwise logistic regression as follows: logit?P = SLE?=-2.594 + 7.503 × linc0597-14.253 × GAS5 + 2.853 × lnc5150-7.012 × lnc7074 + 6.233 × lnc0640,and the AUC of the five-lncRNA panel was 0.966.From the data of testing set,the AUC of these five lncRNAs was 0.649,0.851,0.615,0.683 and 0.662,respectively.The AUC of the five-lncRNA panel was 0.968.In external validation phase,these five lncRNAs levels were also tested in RA patients and pSS patients.We found that levels of plasma GAS5 were both decreased dramatically in SLE patients from testing set compared with RA patients and pSS patients?both P < 0.05?;while linc0597 levels was only significantly lower in SLE patients from testing set than in RA patients?P < 0.05?;and lnc7074 levels was only significantly lower in SLE patients from testing set than in pSS patients?P < 0.05?.The five-lncRNA panel also significantly discriminated the SLE patients in testing set from RA patients,pSS patients and disease controls?RA and pSS?.The AUC was 0.683,0.910 and 0.798 respectively.When the 240 SLE patients were divided into LN and SLE without nephritis,the results showed that the levels of lnc0640,lnc3643,lnc5150,lnc6655,lnc7074 and lnc7514 were significantly up-regulated in patients with LN when compared with SLE without nephritis?all P < 0.05?.The AUC of these six lncRNAs was 0.644,0.671,0.601,0.631,0.665 and 0.684,respectively.The stepwise logistic regression model showed that lnc3643,lnc5150 and lnc7514 can serve as a panel of biomarkers for distinguishing LN from SLE without nephritis.The discrimination model of three-lncRNA panel was constructed to predict the risk of diagnosis with LN as follows: logit?P = LN?=-0.139 + 1.387 × lnc3643-2.048 × lnc5150 + 1.647 × lnc7514,and the AUC = 0.716.Part Two: Bioinformatics study on differentially expressed mRNAs and lncRNAs in SLEGO analysis indicated that the highest enriched term was “ion homeostasis?ontology: biological process?”,“cation transmembrane transporter activity?ontology: molecular function?” and “plasma membrane part?ontology: cellular component?” corresponded to down-regulated transcripts;that the highest enriched term was “cell-cell signaling?ontology: biological process?”,“anion: cation symporter activity?ontology: molecular function?” and “cell periphery?ontology: cellular component?” corresponded to up-regulated transcripts.KEGG Pathway analysis indicated that the most enriched network was “axon guidance-Homo sapiens?human?” and “MAPK signaling pathway-Homo sapiens?human?” from the down-regulated transcripts and the up-regulated transcripts,respectively.The lncRNA-mRNA co-expression network analysis revealed that GAS5,lnc0640,lnc5150,lnc7074,lnc3643,lnc6655 and lnc7514 could highly correlate with 174,18,48,30,36,28 and 20 mRNAs,respectively?Pearson correlation coefficient ? 0.935?.The target gene of GAS5,lnc0640 and lnc5150 may be dual specificity phosphatase 4?DUSP4?,arrestin beta 2?ARRB2?and ribosomal protein S6 kinase A5?RPS6KA5?,respectively.Chromosome 4 open reading frame 26?C4orf26?may be the common target gene of lnc0640,lnc5150,lnc6655 and lnc7074.Poly?ADP-ribose?polymerase family member 16?PARP16?may be the common target gene of lnc3643,lnc5150,lnc6655 and lnc7074.The ceRNA analysis pointed out that GAS5,lnc0640,lnc3643,lnc6655 and lnc7074 could act as ceRNA to regulate the expression of predicted target gene by competing for common miRNA-binding sites with mRNAs.Conclusions1.Compared with normal controls,the plasma levels of GAS5 and lnc7074 were significantly down-regulated while linc0597,lnc0640 and lnc5150 were significantly overexpressed in SLE patients;the expression of plasma GAS5 and linc0597 were decreased dramatically in patients with SLE patients compared with RA patients,GAS5 and lnc7074 levels were significantly lower in SLE patients than in pSS patients;2.The five-lncRNA panel?GAS5,lnc7074,linc0597,lnc0640 and lnc5150?in plasma may serve as potential novel biomarkers for SLE;the three-lncRNA panel ?lnc3643,lnc5150 and lnc7514?in plasma may serve as potential novel biomarkers for distinguishing LN from SLE without nephritis.3.The contribution of GAS5,lnc0640 and lnc5150 to the pathogenesis of SLE may through MAPK pathway.4.GAS5,lnc0640,lnc3643,lnc6655 and lnc7074 could act as ceRNA to regulate the expression of predicted target gene by sponging target mi RNAs. |
PDF Full Text Request