Identification of Novel Long Noncoding RNAs Associated with TGF-β/Smad3-Mediated Renal Inflammation and Fibrosis by RNA Sequencing
2013; Elsevier BV; Volume: 184; Issue: 2 Linguagem: Inglês
10.1016/j.ajpath.2013.10.007
ISSN1525-2191
AutoresQin Zhou, Arthur C.K. Chung, Xiao Ru Huang, Yuan Dong, Xueqing Yu, Hui Y. Lan,
Tópico(s)Circular RNAs in diseases
ResumoWe have previously shown that transforming growth factor-β/Smad3-dependent miRNAs play a critical role in renal inflammation and fibrosis. However, off-target effects of miRNAs limit their therapeutic application. Recently, emerging roles of long noncoding RNAs (lncRNAs) in diseases have been recognized. In this study, we used high-throughput RNA sequencing to identify the Smad3-dependent lncRNAs related to renal inflammation and fibrosis in Smad3 knockout mouse models of unilateral ureteral obstructive nephropathy and immunologically induced anti-glomerular basement membrane glomerulonephritis. Compared with wild-type mice, 151 lncRNAs in the unilateral ureteral obstructive nephropathy kidney and 413 lncRNAs in kidneys with anti-glomerular basement membrane glomerulonephritis were significantly altered in Smad3 knockout mice. Among them, 21 common lncRNAs were up-regulated in wild-type, but down-regulated in Smad3 knockout, kidneys in both disease models in which progressive renal inflammation and fibrosis were abolished when the Smad3 gene was deleted or suppressed. Real-time PCR confirmed these findings and revealed the functional link between Smad3-dependent lncRNAs np_5318/np_17856 and progressive kidney injury. Results demonstrate that the identification and characterization of functional lncRNAs associated with kidney disease may represent a promising research direction into renal disorder and may lead to the development of new lncRNA therapies for kidney diseases. We have previously shown that transforming growth factor-β/Smad3-dependent miRNAs play a critical role in renal inflammation and fibrosis. However, off-target effects of miRNAs limit their therapeutic application. Recently, emerging roles of long noncoding RNAs (lncRNAs) in diseases have been recognized. In this study, we used high-throughput RNA sequencing to identify the Smad3-dependent lncRNAs related to renal inflammation and fibrosis in Smad3 knockout mouse models of unilateral ureteral obstructive nephropathy and immunologically induced anti-glomerular basement membrane glomerulonephritis. Compared with wild-type mice, 151 lncRNAs in the unilateral ureteral obstructive nephropathy kidney and 413 lncRNAs in kidneys with anti-glomerular basement membrane glomerulonephritis were significantly altered in Smad3 knockout mice. Among them, 21 common lncRNAs were up-regulated in wild-type, but down-regulated in Smad3 knockout, kidneys in both disease models in which progressive renal inflammation and fibrosis were abolished when the Smad3 gene was deleted or suppressed. Real-time PCR confirmed these findings and revealed the functional link between Smad3-dependent lncRNAs np_5318/np_17856 and progressive kidney injury. Results demonstrate that the identification and characterization of functional lncRNAs associated with kidney disease may represent a promising research direction into renal disorder and may lead to the development of new lncRNA therapies for kidney diseases. Increasing evidence shows that chronic kidney disease (CKD) is a major problem and health care burden worldwide. More than 10% of the entire population experiences different stages of CKD. Renal inflammation and fibrosis are the common manifestations of CKD.1Eddy A.A. Progression in chronic kidney disease.Adv Chronic Kidney Dis. 2005; 12: 353-365Abstract Full Text Full Text PDF PubMed Scopus (269) Google Scholar Transforming growth factor-β1 (TGF-β1), a well-studied profibrogenic cytokine, exerts its biological effects by activating its downstream mediators, Smad2 and Smad3, which is counterregulated by an inhibitory Smad7.2Roberts A.B. Tian F. Byfield S.D. Stuelten C. Ooshima A. Saika S. Flanders K.C. Smad3 is key to TGF-beta-mediated epithelial-to-mesenchymal transition, fibrosis, tumor suppression and metastasis.Cytokine Growth Factor Rev. 2006; 17: 19-27Abstract Full Text Full Text PDF PubMed Scopus (304) Google Scholar, 3Lan H.Y. Chung A.C. TGF-beta/Smad signaling in kidney disease.Semin Nephrol. 2012; 32: 236-243Abstract Full Text Full Text PDF PubMed Scopus (190) Google Scholar Although Smad3 is a key transcription factor in response to many fibrogenic mediators, targeting Smad3 may cause autoimmune disease by impairing immunity.2Roberts A.B. Tian F. Byfield S.D. Stuelten C. Ooshima A. Saika S. Flanders K.C. Smad3 is key to TGF-beta-mediated epithelial-to-mesenchymal transition, fibrosis, tumor suppression and metastasis.Cytokine Growth Factor Rev. 2006; 17: 19-27Abstract Full Text Full Text PDF PubMed Scopus (304) Google Scholar, 4Yang F. Huang X.R. Chung A.C. Hou C.C. Lai K.N. Lan H.Y. Essential role for Smad3 in angiotensin II-induced tubular epithelial-mesenchymal transition.J Pathol. 2010; 221: 390-401PubMed Google Scholar, 5Chung A.C. Zhang H. Kong Y.Z. Tan J.J. Huang X.R. Kopp J.B. Lan H.Y. Advanced glycation end-products induce tubular CTGF via TGF-beta-independent Smad3 signaling.J Am Soc Nephrol. 2010; 21: 249-260Crossref PubMed Scopus (157) Google Scholar, 6Meng X.M. Huang X.R. Chung A.C. Qin W. Shao X. Igarashi P. Ju W. Bottinger E.P. Lan H.Y. Smad2 protects against TGF-beta/Smad3-mediated renal fibrosis.J Am Soc Nephrol. 2010; 21: 1477-1487Crossref PubMed Scopus (271) Google Scholar Thus, alternative approaches to inhibit TGF-β actions should be developed in an attempt to suppress renal inflammation and fibrosis without impairing the immune system. Emerging evidence shows that noncoding RNAs (ncRNAs) may play a critical role in the development of kidney diseases. The identification of disease-related miRNAs in nephropathy was thought to be a specific therapy for treating the kidney disease.3Lan H.Y. Chung A.C. TGF-beta/Smad signaling in kidney disease.Semin Nephrol. 2012; 32: 236-243Abstract Full Text Full Text PDF PubMed Scopus (190) Google Scholar, 7Lorenzen J.M. Haller H. Thum T. MicroRNAs as mediators and therapeutic targets in chronic kidney disease.Nat Rev Nephrol. 2011; 7: 286-294Crossref PubMed Scopus (179) Google Scholar, 8Kato M. Natarajan R. MicroRNA circuits in transforming growth factor-beta actions and diabetic nephropathy.Semin Nephrol. 2012; 32: 253-260Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar, 9Wei Q. Mi Q.S. Dong Z. The regulation and function of microRNAs in kidney diseases.IUBMB Life. 2013; 65: 602-614Crossref PubMed Scopus (83) Google Scholar, 10Chung A.C. Yu X. Lan H.Y. MicroRNA and nephropathy: emerging concepts.Int J Nephrol Renovasc Dis. 2013; 6: 169-179PubMed Google Scholar, 11Lan H.Y. Chung A.C. Transforming growth factor-beta and Smads.Contrib Nephrol. 2011; 170: 75-82Crossref PubMed Scopus (92) Google Scholar, 12Meng X.M. Chung A.C. Lan H.Y. Role of the TGF-beta/BMP-7/Smad pathways in renal diseases.Clin Sci. 2013; 124: 243-254Crossref PubMed Scopus (282) Google Scholar, 13Qin W. Chung A.C. Huang X.R. Meng X.M. Hui D.S. Yu C.M. Sung J.J. Lan H.Y. TGF-beta/Smad3 signaling promotes renal fibrosis by inhibiting miR-29.J Am Soc Nephrol. 2011; 22: 1462-1474Crossref PubMed Scopus (452) Google Scholar, 14Zhong X. Chung A.C. Chen H.Y. Dong Y. Meng X.M. Li R. Yang W. Hou F.F. Lan H.Y. miR-21 is a key therapeutic target for renal injury in a mouse model of type 2 diabetes.Diabetologia. 2013; 56: 663-674Crossref PubMed Scopus (286) Google Scholar, 15Zhong X. Chung A.C.K. Chen H.-Y. Meng X.-M. Lan H.Y. Smad3-mediated upregulation of miR-21 promotes renal fibrosis.J Am Soc Nephrol. 2011; 22: 1668-1681Crossref PubMed Scopus (347) Google Scholar, 16Chung A.C.K. Huang X.R. Meng X. Lan H.Y. miR-192 mediates TGF-β/Smad3-driven renal fibrosis.J Am Soc Nephrol. 2010; 21: 1317-1325Crossref PubMed Scopus (309) Google Scholar However, the off-target effects hinder the development of specific miRNA therapy, because one miRNA is able to regulate many target genes.7Lorenzen J.M. Haller H. Thum T. MicroRNAs as mediators and therapeutic targets in chronic kidney disease.Nat Rev Nephrol. 2011; 7: 286-294Crossref PubMed Scopus (179) Google Scholar, 10Chung A.C. Yu X. Lan H.Y. MicroRNA and nephropathy: emerging concepts.Int J Nephrol Renovasc Dis. 2013; 6: 169-179PubMed Google Scholar Since the discovery of Xist, a long noncoding RNA (lncRNA) required for mammalian X chromosome inactivation,17Brown S.D. XIST and the mapping of the X chromosome inactivation centre.Bioessays. 1991; 13: 607-612Crossref PubMed Scopus (62) Google Scholar many lncRNAs have been reported in mammals and other vertebrates.18Guttman M. Rinn J.L. Modular regulatory principles of large non-coding RNAs.Nature. 2012; 482: 339-346Crossref PubMed Scopus (1710) Google Scholar At present, many lncRNAs, defined as RNAs >200 nucleotides (nt) that do not encode any protein, are found to be able to control gene expression.18Guttman M. Rinn J.L. Modular regulatory principles of large non-coding RNAs.Nature. 2012; 482: 339-346Crossref PubMed Scopus (1710) Google Scholar The development of high-throughput RNA sequencing (RNA-Seq) accelerates the discovery of novel ncRNAs because RNA-Seq can achieve single-base resolution and capture all of the transcripts in the samples.19Wang Z. Gerstein M. Snyder M. RNA-Seq: a revolutionary tool for transcriptomics.Nat Rev Genet. 2009; 10: 57-63Crossref PubMed Scopus (8465) Google Scholar, 20Marguerat S. Bahler J. RNA-seq: from technology to biology.Cell Mol Life Sci. 2010; 67: 569-579Crossref PubMed Scopus (381) Google Scholar In this study, we used this technique to search for Smad3-dependent lncRNAs during renal injury in mouse models of immunologically induced anti-glomerular basement membranous glomerulonephritis (anti-GBM GN) and nonimmune disease of unilateral ureteral obstructive nephropathy (UUO). The results of real-time PCR validation, sequence analyses, and chromatin immunoprecipitation (ChIP) assays revealed a functional link between Smad3-dependent lncRNAs and progressive kidney injury. Thus, outcomes from this study suggested a novel research direction into renal disorder and may promote new therapeutic strategies for kidney diseases by targeting lncRNAs. The animal models of UUO and anti-GBM GN were induced in both sexes of C57BL/6J wild-type (WT) and Smad3 knockout (KO) mice at 8 to 10 weeks of age, as previously described.21Huang X.R. Chung A.C. Wang X.J. Lai K.N. Lan H.Y. Mice overexpressing latent TGF-beta1 are protected against renal fibrosis in obstructive kidney disease.Am J Physiol Renal Physiol. 2008; 295: F118-F127Crossref PubMed Scopus (80) Google Scholar, 22Huang X.R. Chung A.C. Zhou L. Wang X.J. Lan H.Y. Latent TGF-beta1 protects against crescentic glomerulonephritis.J Am Soc Nephrol. 2008; 19: 233-242Crossref PubMed Scopus (102) Google Scholar Briefly, for the UUO model, the left ureter from Smad3 WT or KO mice (n = 8 per group, both sexes, 8 weeks of age, 22 to 25 g body weight) was ligated and the UUO kidney was harvested at day 5 after surgery for further analysis. In addition, a group of eight Smad3 WT or KO mice, sham operated on, were used as a control. For the anti-GBM GN model, groups of eight Smad3 WT or KO mice were pre-immunized with sheep IgG in Freund’s complete adjuvant (Sigma Chemical Co, St. Louis, MO), followed 5 days later by i.v. administration of sheep anti-mouse GBM immunoglobulin at a dose of 60 μg/g of body weight (termed day 0). Diseased mice were euthanized at day 10 for further analysis. Groups of age- and weight-matched normal Smad3 WT or KO mice were used as normal controls (n = 8). The experimental procedures for both models were performed following the approved protocol by the Animal Experimentation Ethics Committee at The Chinese University of Hong Kong (Hong Kong). The mixture of Smad7 and tetracycline-on (tet-on) plasmids was transferred into the ligated kidneys of UUO, as described previously.23Chung A.C. Dong Y. Yang W. Zhong X. Li R. Lan H.Y. Smad7 suppresses renal fibrosis via altering expression of TGF-beta/Smad3-regulated microRNAs.Mol Ther. 2013; 21: 388-398Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar The experimental procedures were approved by the Chinese University of Hong Kong’s Animal Experimental Ethics Committee. Mutant mice deficient in exon 1 of the Smad7 gene from a CD1 background (provided by Prof. Rainer L. Heuchel, Karolinska Institutet, Stockholm, Sweden) have been described recently.24Li R. Rosendahl A. Brodin G. Cheng A.M. Ahgren A. Sundquist C. Kulkarni S. Pawson T. Heldin C.H. Heuchel R.L. Deletion of exon I of SMAD7 in mice results in altered B cell responses.J Immunol. 2006; 176: 6777-6784PubMed Google Scholar All mice were from a CD1 background. A mouse model of UUO was induced in groups of six to eight Smad7 KO or WT mice (male, 8 weeks of age, 22 to 25 g body weight), as previously described.21Huang X.R. Chung A.C. Wang X.J. Lai K.N. Lan H.Y. Mice overexpressing latent TGF-beta1 are protected against renal fibrosis in obstructive kidney disease.Am J Physiol Renal Physiol. 2008; 295: F118-F127Crossref PubMed Scopus (80) Google Scholar, 25Chung A.C. Huang X.R. Zhou L. Heuchel R. Lai K.N. Lan H.Y. Disruption of the Smad7 gene promotes renal fibrosis and inflammation in unilateral ureteral obstruction (UUO) in mice.Nephrol Dial Transplant. 2009; 24: 1443-1454Crossref PubMed Scopus (53) Google Scholar The experimental procedures were approved by the Chinese University of Hong Kong’s Animal Experimental Ethics Committee. Kidney tissues of Smad3 KO/WT mice (n = 2 in each group) were collected from a normal, UUO model at day 5, and an anti-GBM GN model at day 10 for RNA sequencing. Total RNA from the kidney tissue was extracted by TRIzol reagent (Life Technologies, Carlsbad, CA), as per the manufacturer’s instructions. Additional DNase I digestion and clear-up steps were performed per the protocol from the Beijing Genomics Institute (Shenzhen, China). RNA quality and purity were determined by the ND-1000 Nanodrop (NanoDrop, Wilmington, DE). For each sample, RNAs were fractionated with a denaturing polyacrylamide gel. To maximize the ncRNA database for analysis, RNAs >150 nt were used for cDNA synthesis. After adenylating the 3′ end of RNAs and ligating the adapter, the RNA samples were preceded to library construction following the protocol at Beijing Genomics Institute (Supplemental Figure S1). The sequencing was performed by the Genome Analyzer (serial No. Hiseq2000; Illumina, San Diego, CA). Low-quality reads, >50% of the sequence reads (including adapters) with a quality score of ≤10 and N >5%, were discarded. Ribosome RNA sequences were filtered from the raw fragments. Remaining clean reads were mapped to a genome (University of California, Santa Cruz, CA, genome browser, mouse genome July 2007, mm9) by using Tophat [1.3.0.Linux_x86_64 (bowtie-0.12.7)] and then assembled with Cufflinks (1.3.0.Linux_x86_64; Berkeley, CA). To discriminate coding from noncoding transcripts, a support vector machine-based classifier (coding potential calculator) was used to predict novel ncRNAs.26Kong L. Zhang Y. Ye Z.Q. Liu X.Q. Zhao S.Q. Wei L. Gao G. CPC: assess the protein-coding potential of transcripts using sequence features and support vector machine.Nucleic Acids Res. 2007; 35: W345-W349Crossref PubMed Scopus (1826) Google Scholar The prediction of novel lncRNA was assembled based on the SwissProt database (Hinxton, Cambridge, UK). All of the ncRNAs were annotated by aligning the genomic sequence against RFAM, a collection of multiple sequence alignments and covariance models representing ncRNA families (http://rfam.sanger.ac.uk, last accessed August 31, 2013), as previously described.27Burge S.W. Daub J. Eberhardt R. Tate J. Barquist L. Nawrocki E.P. Eddy S.R. Gardner P.P. Bateman A. Rfam 11.0: 10 years of RNA families.Nucleic Acids Res. 2013; 41: D226-D232Crossref PubMed Scopus (607) Google Scholar The RPKM value was used to obtain a comparable measure of gene expression, and Pearson correlation was used to analyze RPKM value correlation. ncRNAs >200 nt were selected as lncRNAs for further analysis because approximate 200 nt is the common cutoff to define lncRNA in the literature.18Guttman M. Rinn J.L. Modular regulatory principles of large non-coding RNAs.Nature. 2012; 482: 339-346Crossref PubMed Scopus (1710) Google Scholar The sequencing data obtained were deposited in the National Center for Biotechnology Information database (http://www.ncbi.nlm.nih.gov/bioproject/223210; accession number PRJNA223210). We used an evolutionary conserved region (ECR) browser28Ovcharenko I. Nobrega M.A. Loots G.G. Stubbs L. ECR Browser: a tool for visualizing and accessing data from comparisons of multiple vertebrate genomes.Nucleic Acids Res. 2004; 32: W280-W286Crossref PubMed Scopus (403) Google Scholar to detect any conserved Smad binding motif prediction in mouse regulatory regions. Chromosomal location and gene ontology of the 21 common lncRNAs and their linked genes were searched on the EMBL-EBI database (http://www.ebi.ac.uk, last accessed August 31, 2013) and the ECR Browser (http://ecrbrowser.dcode.org, last accessed August 31, 2013).28Ovcharenko I. Nobrega M.A. Loots G.G. Stubbs L. ECR Browser: a tool for visualizing and accessing data from comparisons of multiple vertebrate genomes.Nucleic Acids Res. 2004; 32: W280-W286Crossref PubMed Scopus (403) Google Scholar ChIP was performed with a transcription factor ChIP kit, as previously described.15Zhong X. Chung A.C.K. Chen H.-Y. Meng X.-M. Lan H.Y. Smad3-mediated upregulation of miR-21 promotes renal fibrosis.J Am Soc Nephrol. 2011; 22: 1668-1681Crossref PubMed Scopus (347) Google Scholar, 29Li R. Chung A.C. Dong Y. Yang W. Zhong X. Lan H.Y. The microRNA miR-433 promotes renal fibrosis by amplifying the TGF-β/Smad3-Azin1 pathway.Kidney Int. 2013; 84: 1129-1144Crossref PubMed Scopus (133) Google Scholar In brief, cells were cross-linked with 1% formaldehyde for 10 minutes at 37°C, quenched with glycine, and then sonicated using a Bioruptor (Diagenode, Liège, Belgium) to generate 300- to 600-bp DNA fragments. Immunoprecipitation was performed with the antibody against Smad3 (Upstate/Millipore, Billerica, MA), and a normal IgG was used as a control. Precipitated DNAs were detected by PCR using specific primers: Smad binding site (SBS) for np_17856, 5′-GAAGAGATCTTATACATTTCCCCCTA-3′ and 5′-TTCCTCTTCCCAATGTGGTC-3′; and SBS for np_5318, 5′-CTCTCTCAAACAGCCTGTGG-3′ and 5′-GAAATTTGGAGGTGCAATCAA-3′. lncRNA expression was quantified by real-time PCR with SYBR Green (Life Technologies, Carlsbad, CA) as the fluorescent reporter. Total RNA (1 μg) was used to synthesize the first strand of cDNA using M-MLV Reverse Transcriptase (Life Technologies). The real-time PCR was performed as previously described.23Chung A.C. Dong Y. Yang W. Zhong X. Li R. Lan H.Y. Smad7 suppresses renal fibrosis via altering expression of TGF-beta/Smad3-regulated microRNAs.Mol Ther. 2013; 21: 388-398Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar Primers for 21 common lncRNAs and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) are listed in Table 1. The relative levels of target lncRNAs were calculated using the ΔΔCT method, and the data were normalized with GAPDH.Table 1Real-Time PCR Primers for 21 lncRNAs and GAPDHGeneForward primerReverse primernp_185885′-CGGAGGATTGATAGCCAAAA-3′5′-CCTGGTTGTCGTGTTTCCTT-3′np_269925′-GACCAATGTTCGTAGCAGCA-3′5′-TCCTTCTCTTTGTGCGTGTG-3′na_24145′-CTGCCTCTGTGCTGCATTTA-3′5′-TCTTTGTCACCCAAGCAGTG-3′np_299085′-TGGATTTGGCCTCAGTTTTC-3′5′-CAGCTTCGAGTTTGTCACCA-3′np_165555′-CAGCCTAGCTCACCTTCTGG-3′5′-TGTTCATCTGTGGCTCTTGG-3′np_53405′-TTCCTGGGCTTCACAAGTTT-3′5′-GCGTGGCTTCAGGTATTTGT-3′np_172345′-TGTTTTAACCCATTGTATCATTTCA-3′5′-CAACCCCCTTTGGAAGACTC-3′np_186035′-CCTTTACAAGCAAAAGCCAAA-3′5′-CCCTTGAGACACTGCCTGAC-3′na_98845′-ACCACAGCAGATACCCAAGC-3′5′-CAGCAAGCTCCTTTTTCCAC-3′np_53925′-GGGTGTGAACGGAGACTGTT-3′5′-GGCGAGGAATGTGTTTGTTT-3′np_180045′-CTCCATAAAGAGCCCCATGA-3′5′-TTGGGCCTTCCAACAATAAC-3′np_287305′-TCTGGGTCCAACTCCATTTT-3′5′-CCTGTTTTAATTGGGGGAAA-3′np_43345′-AGAGGGCAAGCAACATTGTC-3′5′-GCTGTGGGGAGAGCAACTAA-3′np_186365′-CAACCTAAGGCTGCTCTCTCC-3′5′-CCATCAAGAAACCTGGCAAT-3′np_178565′-GAAGCCTGCAGGAAGATGAC-3′5′-ACCAGTCACAAAAGCCAACC-3′np_185375′-TTCCAGTAGTGCTGCACAGG-3′5′-TCCAGGTTTCCAACTTTTGC-3′np_53185′-AACTCGCCACAGAAATCCAC-3′5′-ACAACCCCAAACAAGCTGTC-3′np_199025′-TGTTGCTAGAATGGCAGGTG-3′5′-TCCTGCCTGGGAATACTGTC-3′np_282335′-GATGGGGAAGAAACCCATTT-3′5′-CAGGGCCAGTTATGTCCAGT-3′np_21085′-AGATCTCCCCTCTTGGGAAA-3′5′-CTCCCTCTCCACACTGCTTC-3′np_284965′-AGCAGGCTAAGTCAGGGTGA-3′5′-GGGTTGAGCTGCCATGTATT-3′GAPDH5′-TGCTGAGTATGTCGTGGAGTCTA-3′5′-AGTGGGAGTTGCTGTTGAAATC-3′ Open table in a new tab For lncRNA expression by RNA-Seq, significance levels were indicated as P < 0.01 and greater than twofold change. Data obtained from real-time PCR are expressed as the means ± SEM and analyzed using a t-test in Microsoft Excel (Redmond, WA). In this study, high-throughput RNA-Seq was used to search for Smad3-dependent lncRNAs during renal injury in mouse models of anti-GBM GN and UUO. Kidneys of Smad3 KO/WT mice with UUO and anti-GBM GN were collected for sequencing, as described in Supplemental Figure S1. Total RNA of all 12 samples of Smad3 KO/WT kidneys was sequenced by Illumina, each generating >50 million reads. Low-quality reads and rRNA sequences were filtered from the raw fragments. Total clean reads were then mapped to genome by using Tophat, an alignment program that maps RNA-Seq reads across the splice junction without relying on gene annotation.30Trapnell C. Pachter L. Salzberg S.L. TopHat: discovering splice junctions with RNA-Seq.Bioinformatics. 2009; 25: 1105-1111Crossref PubMed Scopus (8995) Google Scholar As shown in Supplemental Table S1, the proportion of clean reads that mapped to the genome ranged from 71.55% to 85.82% in 12 samples. The mapped reads were assembled into transcripts with Cufflinks, putative transcripts containing both coding and noncoding transcripts.31Trapnell C. Williams B.A. Pertea G. Mortazavi A. Kwan G. van Baren M.J. Salzberg S.L. Wold B.J. Pachter L. Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation.Nat Biotechnol. 2010; 28: 511-515Crossref PubMed Scopus (10569) Google Scholar The identification and characterization of lncRNAs are an experimental and computational challenge. In the present study, all assembled transcripts were aligned to a ncRNA database using blastn (1×10−10). The filter was set as identity >0.9 and coverage >0.8. Reads mapped to multinoncoding RNA databases using the previously described criteria were identified as noncoding RNAs. Approximately 20.03% to 35.71% clean reads were mapped to the ncRNA database (obtained from frnadb, lncrnadb, NONCODE, RNAdb, and ensembl) (Supplemental Table S1). All ncRNAs were then annotated by aligning a genomic sequence against RFAM, a collection of multiple sequence alignments and covariance models representing ncRNA families. The resulting hits were then used as supporting evidence for ncRNAs. At last, 12,131 transcripts were aligned to the known ncRNA database, and 32,957 transcripts were predicted to be ncRNAs with a length of >150 nt (Figure 1A). We next identified lncRNAs that differentially expressed, in both models of anti-GBM GN and UUO kidneys, the transcript abundances measured by reads per kb per million (RPKM; RPKM = 109 × C/NL) where C is the reads number of the transcript, L is the length of the transcript, and N is the total reads number of the sample (Supplemental Table S2).32Mortazavi A. Williams B.A. McCue K. Schaeffer L. Wold B. Mapping and quantifying mammalian transcriptomes by RNA-Seq.Nat Methods. 2008; 5: 621-628Crossref PubMed Scopus (9850) Google Scholar Pearson correlation analysis showed that the RPKM values between these groups correlated moderately (Supplemental Figure S2), which may be due to low expression levels of lncRNAs within the diseased kidney because higher expression levels of mRNAs (0.8 to 1.0) in these diseased kidney tissues were detected by the Pearson correction analysis (data not shown). Further studies revealed that, compared with normal mice, many lncRNAs were differentially expressed in the diseased kidney of Smad3 WT mice, but this was less abundant in Smad3 KO mice (Figure 1B). Interestingly, deletion of Smad3 largely altered the expression levels of lncRNAs in both disease models (Figure 1B). Of these, 151 lncRNAs (plus or minus twofold to eightfold) in UUO and 413 lncRNAs (plus or minus twofold to ninefold) in anti-GBM GN were differentially expressed in WT and Smad3 KO kidneys (Figure 1C). Among them, 21 lncRNAs were altered in both UUO and anti-GBM GN models, being Smad3-dependent common lncRNAs (Figure 1C). Expression levels of these common lncRNAs were up-regulated in WT (plus twofold to approximately sixfold), but suppressed in Smad3 KO, mice with UUO (−twofold to approximately fivefold) and anti-GBM GN (−twofold to approximately sixfold) (P < 0.01) (Table 2). To confirm the expression levels of these 21 common lncRNAs in vivo, real-time PCR assays were used. The relative expression levels of 21 lncRNAs were significantly up-regulated in WT, but inhibited in Smad3 KO, kidneys in both anti-GBM GN and UUO models (Figure 1D).Table 2Comparison of Expression Levels of 21 Common lncRNAs in S3WT/S3KO Kidneys with UUO and Anti-GBM GNlncRNALength (nt)Log2 fold changeS3WT (UUO vs Nor)S3KO (UUO vs Nor)S3WT (GBM vs Nor)S3KO (GBM vs Nor)np_185885742.146∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−3.660P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.2.834∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−4.600P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.np_2699212282.174P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−2.982P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.2.419∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−4.922P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.na_24145282.196P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−3.245∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.2.320P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−2.085∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.np_2990814132.196P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−4.075P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.2.213P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−5.015∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.np_165558182.229P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−4.245∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.3.132∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−5.185∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.np_534021572.360∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−3.708∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.3.104∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−3.384∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.np_172342692.381∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−3.397∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.2.789∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−3.337∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.np_186036062.381∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−5.535∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.2.497P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−2.474P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.na_98844362.474∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−2.175∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.3.156∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−3.337∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.np_539229172.544∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−2.075P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.3.320∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−5.600∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.np_180046212.644∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−3.883P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.2.294∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−4.822P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.np_287304362.703∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−4.776P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.2.227∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−4.715P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.np_433413932.722P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−2.277P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.4.337∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−3.538∗P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.np_1863614952.866P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison between groups.−2.535P < 0.01, ∗∗P < 0.001, and ∗∗∗P < 0.0001 for the comparison
Referência(s)