| BackgroundHeart failure,a growing major worldwide public health problem,usually results from deficiency of the cardiomyocytes(CMs).Numerous insults,including ischemic coronary artery disease,hypertension,atrial fibrillation and valvular heart disease can cause loss of CMs.The adult human heart have very little regeneration potential to compensate for lost CMs,it repairs itself after injury with noncontractile scar tissue rather than new CMs,which results in adverse remodeling and subsequent heart failure.If there is a robust therapy to stimulate human cardiac regeneration,it may help millions of patients every year.Over the past decade,a lot of efforts have been put into the development of novel therapeutic strategies to induce cardiac regeneration.The approaches for regeneration could be attributed to promote the differentiation of endogenous cardiac progenitor or stem cell and to reactivate the proliferation of pre-existing CMs.Despite numerous types of progenitor/stem cells have been used for cardiac repair,most of the cardiac cell therapy trials in human failed to achieve true cardiac regeneration and adequate improvement in heart function.Notably,there was little evidence that mammalian heart renewal arising from progenitors/stem cell.In contrast,the pre-existing CMs were demonstrated to be the dominant source of cardiac regeneration in normal mammalian myocardial homeostasis and after ischemic or injury.Thus,insights into the stimulating pre-existing CM proliferation might provide feasible strategies to regenerate the injured adult heart.Although adult CMs have limited proliferative potential,fetal CMs are able to proliferate robustly.Cardiac growth during embryonic development is primarily occurs by the increase in CM number.Soon after birth,the CMs lose their ability to proliferate and the subsequent growth of heart switches predominantly to the enlargement of pre-existing CMs.Comprehensive analysis of differential gene expression profiles between fetal and adult heart might reveal the genetic networks that retain proliferative potential of pre-existing CMs,with a goal to reactivate CM proliferation and regenerate the injured adult heart.But,to date and to our knowledge,no comparison of whole transcriptome between human and fetal heart has been reported.Recent advancements in next generation sequencing technologies are enabling a new way to analyze whole transcriptome(including both protein coding and non-coding RNA)and to discover novel transcripts.As part of these advancements,RNA-Seq(also referred to whole transcriptome shotgun sequencing),one of the most powerful methodologies to discover and analyze novel non-coding RNAs,have dramatically changed the field of non-coding RNA research.ChIP-sequencing(ChIP-seq)is another method combineed chromatin immunoprecipitation(ChIP)with massively parallel DNA sequencing to identify the binding sites of DNA-associated proteins.Combining the RNA-seq with Chip-seq technology may help uncover the mystery of lncRNA.A new class of noncoding RNAs with a length of more than 200 nucleotides,defined as long noncoding RNAs(lncRNAs),has received a remarkable level of attention due to their crucial roles in the regulation of multiple biological processes.Increasing studies shown that lncRNAs have function inepigenetic regulation,transcription,translation,splicing,pluripotent stem cells reprogram,cell cycle control and cell differentiation.Currently,although research in the function of lncRNA in the cardiovascular system is still in the initial stage,it has highlighted the important role of lncRNA in the cardiovascular system.A recent study characterized cardiac-enriched lncRNAs.The study identifed 321 cardiac-expressed lncRNAs by comparing with other tissues,most of the abundant and cardiac-enriched lncRNAs were also highly expressed in isolated cardiomyocytes,whereas a few were highly expressed in fbroblasts.Several studies have showed that IncRNAs plays a very important role in heart development.The IncRNA-Fendrr is required for proper development of the heart and body wall.When the mouse lack of lncRNA-Fendrr,intraventricular septal heart defects were obsevred.Another study repored that the IncRNA Bvht played a key role in the regulation of commitment of the embryonic stem cells(ESCs)to cardiovascular lineages:cardiomyocytes,smooth muscle cells and endothelial cells.However,up until now,no ortholog of Braveheart could be identifed in human,whereas human ortholog of Fendrr did exist.Up to date,the role of LncRNA in the human heart development and regeneration remains unclear.In this study,we analyzed the publicly available RNA-seq data of human fetal(gestational 13 to 17 weeks)and normal adult cardiac tissues to identify a comprehensive set of lncRNAs in the human genome that may play a role in the cardiac regeneration,.Furthermore,we profiled chromatin modifications by using the corresponding CHIP-seq data to investigate the stage of differential protein-coding and lncRNAs.Finally,we explored the potential function of lncRNAs in CM proliferation,with a specific focus on those motif structure prediction related to cell proliferation.Methods and materialsRNA-seq data and MappingThe RNA-Seq datasets were acquired from European Nucleotide Archive(ENA).Each data was available as a single-end reads or two paired-end reads.The standard criteria of filtered out low-quality reads that fit following parameters:remove reads with more than 10%unknown bases(N bases);remove reads with more than 50%of low-quality bases in one read.Then clean reads were mapped to the hg19 reference database by using SOAPaligner/SOAP2.The resulting alignment was reconstructed by using Cufflinks.The RefSeq and Ensembl transcript databases were used as the annotation references for mRNA analyses,while the NONCODE v4.0 was used for lncRNA annotation.Novel transcripts predictionWe used Cufflinks to assemble candidate transcripts from the mapped reads using reference gene annotation as a guide.Non-overlapping transcripts were used CPC to assessed for protein-coding-potential.Using a threshold to identify coding/noncoding RNA,all the transcripts with a CPC>-0.5 were called mRNA and CPC<-0.5 called lncRNAs.Analysis of mRNAs and lncRNAs expressionIn our analyses,levels of lncRNAs expression were calculated using the RPKM measure(Reads Per Kilobasetranscriptome per million reads)and RPKM>1 is required for the following analysis.The R-package previously implemented by the authors of NOISeq was used to identify significantly differentially expressed genes.Hierarchical clustering analysesWe used Cluster 3.0 and JavaTreeview software for cluster analysis of the gene expression patterns.Genes with expression differences were clustered by the Hierarchical Complete Linkage Clustering method using Euclidean distance.Before hierarchical clustering analyses,we added two additional steps to genes expression.First,genes with only one read covered were filtered out for lack of reliability.Then,we pretreated the dataset using excel.(1)All data values x by log2(x)to log transform data.[1]Subtract the column-wise mean from the values in each column of data so that the mean value of each column is 0.Red corresponds to fully induced expression and green corresponds to fully repressed expression.Functional analysisThe genes ID lists were submitted online to the DAVID for Gene ontology(GO)enrichment and Kyoto Encyclopedia of Genes and Genomes(KEGG)Pathways analyses.A hypergeometric test with the Benjamini and Hochberg false discovery rate(FDR)was performed using the default parameters to adjust the P value.The significance of the enrichment was defined by the P values<0.05.Functional RNA motifs were performed by using RegRNA2.0.Chromatin-Immunoprecipitation SequencingCHIP-seq data for the fetal and adult human cardiac tissues were obtained from European Nucleotide Archive(ENA).Five markers were considered:H3K4me1,H3K4me3,H3K9me3,H3K27me3 and H3K36me3.A marker was considered present if a non-empty intersection could be detected between the TSS region and a marker peak,in any of the replicates.The intersections were detected using the window command of the BEDTools program,version 2.17.0,with option-w 1000.Correlation of expression between novel IncRNAs and closest coding genesThe coordinates of the lncRNAs were compared to coding genes reference.If the coordinates of the lncRNA overlapped with a known gene(at least I bp),this gene was considered as the closest overlapping gene.If there was no overlap with a known gene,the closest gene was selected and classified as upstream or downstream depending on its position.For gene expression,the same data as in Expression heatmap was used.The correlation of expression was calculated between the lncRNA and closest coding gene using Pearson correlation.Cell Proliferation AssayThe cardiomyocyte proliferation was detected by Click-iT(?)EdU Imaging Kits and immunofluorescence labeling with cell nuclear antigen Ki-67.Injection of lentiviral vectors in neonatal and adult ratsThe neonatal rats(post-natal day 1)were intramyocardially injected with adenovirus Vector-Mediated lncRNA-PTTG1 overexpression and vector control at a dose of 1 × 1011 viral genome particles per animal by using an insulin syringe with incorporated 30-gauge needle.Proliferation of cardiac cells was analysed 6 days after injection.Statistical and Computational MethodsFor gene expression analysis,differential gene expression was identified using the NOISeq method with a cutoff probability threshold of 0.8 as described above.For GO enrichment and Pathways analysis,Benjamini and Hochberg false discovery rate was performed to as described above.Other computational procedures were carried out using in-house programs written in R or analyzed in Excel(Microsoft)and Stata 12.0.SPSS 13.0 software was used for statistical analysis.Measurement data was analyzed by independent-samples T test.Results were described as mean ± standard deviation(x±s),P<0.05 was considered as significant difference.A value of P<0.05 was considered significant.Results:Globalidentification of mRNA and LncRNATo uncover the different transcriptome profiling of fetal and adult heart,the RNA-seq data of fetal and adult human cardiac tissues were analyzed.The 4 RNA-seq data generated a total of 189 million clear reads,of which over 170 million(?89.7%)were uniquely aligned to the human genome(hg19).Among the uniquely mapped reads,87 million(51.1%)reads mapped to intergenic regions,69 million(40.6%)reads mapped within exons,and 14 million(8.2%)reads mapped to introns.The proportion of reads mapped to introns and exons were distinctly different between fetal and adult heart.The distribution of the mapped reads mapped to the chromosome in fetal heart is similar to that in adult heart.Further,we identified 152130(70.9%)transcripts were annotated to RefSeq genes,33073(15.4%)were annotated to Ensemble genes,and 28075(13.1%)were annotated to NONCODEv4 genes.Compared with the proportion of lncRNAs in adult,lncRNAs accounted for less percentage of total genes in fetal heart.The rest of reads underwent transcriptome reconstruction using Cufflinks,which identified 3958 novel transcripts that do not overlap with any known coding or non-coding genes,3830 of which with low coding potentials were defined as novel lncRNAs.Among these 3830 novel lncRNAs,1832 were located in intergenic regions,1595 were located in intron and 403 intersect with known genes.Besides,3594 of these novel lncRNAs were single-exon lncRNAs,236 were multi-exon lncRNAs.Among these multi-exon lncRNAs,63 were expressed at>0.5 Reads PerKilobase of exon per Million mapped reads(RPKM)and 18 were expressed at>1 RPKM in either fetal or adult hearts.Conservation analysis revealed that novel and known-lncRNA were less conserved than mRNA.In detail,Novel and known-IncRNA exons were less conserved than coding exons although introns and promoters were equally conserved.In addition,novel lncRNAs and known-lncRNAs were are less abundant and were shorter in length than the mRNA coding genes.Our results were consistent with previously reported human lncRN A expression data.Differentially expressed mRNA in fetal and adult heart.To identify genes that might involved in heart development,we compared the gene expression profiles between fetal and adult heart.For this analysis,we retained only those transcripts that were transcribed>1 RPKM in either fetal or adult hearts.Hierarchical clustering revealed significant variations in the expression of differentially expressed protein-coding RNAs in fetal and adult heart.Volcano plots comparing fold change and probability value identified 3211 significantly increased and 578 significantly decreased mRNAs.Further,we performed hierarchical clustering of differentially expressed protein-coding RNAs in fetal and adult heart.One of the most significant differentially expressed genes modules was related to cell cycle function.For instance,CDC6,CCNB1,PTTG2,and CDCA5 were example genes known to promote cell cycle progression and proliferation.What's more,Gene Ontology(GO)classification of up-regulated genes in fetal hearts showed enrichment for gene categories that regulate cell cycle,while down-regulated genes were enriched in other functional categories.Besides,Kyoto Encyclopedia of Genes and Genomes(KEGG)pathway analysis further revealed that up-regulated genes were enriched in the proliferation,such as "cell cycle","DNA replication" and "Oocyte meiosis".Differentially expressed lncRNA in fetal and adult heart.Recently,tens of thousands of IncRNA genes with diverse functional roles were discovered in the mouse and human genomes.To identity a distinct IncRNA signature for human developing hearts,we carry out hierarchical clustering analysis of lncRNAs expression across fetal and adult heart revealed a distinct expression signature of IncRNA in the fetal hearts.A total of 3092 lncRNAs exhibited significant differences(probability>0.8)in expression between fetal and adult heart,while 1343 IncRNAs were up-regulated and 1749 IncRNAs were down-regulated in fetal heart.Further,previous studies revealed that lncRNAs can epigenetically regulate mRNA/protein coding genes by interacting nearby in cis or distally in trans.To explore the potential cis-and trans-regulatory roles of human IncRNAs on protein coding genes.Firstly,we computed the interactions of IncRNAs-rnRNA pairs that are located within 10 kb upstream and downstream representing cis-correlation of expression.34%of the tested lncRNA:cis-mRNA gene pairs showed positive correlation(Pearson's r>0.5),whereas 44%showed negative correlation(Pearson's r<-0.5).In contrast,mRNA:cis-mRNA gene pairs showed more frequent positive than negative correlations(54%versus 29%,respectively).Gene Ontology(GO)classification of up-regulated IncRNA:cis-mRNA pairs(r>0.95)showed enrichment for gene categories that regulate mRNA metabolic process,mRNA processing,RNA splicing and cell cycle,while down-regulated lncRNA:cis-mRNA pairs(r>0.95)showed also enriched in heart development-related functional categories.Besides,we calculated interactions of IncRNA-mRNA pairs that are located at least 1 Mb distance or on different chromosomes representing trans-correlation of expression.We found that the percentage of lncRNA:trans-mRNA pairs showing a positive correlation is similar to that shows a negative correlation(35%with Pearson r>0.5,33%with Pearson r<-0.5).Figure S3 illustrates an example of alncRNA:tans-mRNA pair that is unregulated in fetal heart and shows a strong positive correlation.Which might involve in the cell cycle.In line with the mRNA:cis-mRNA gene pairs,the mRNA:tans-mRNA gene pairs showed more frequent positive than negative correlations(54%versus 36%,respectively).This observation suggested that the potential trans-regulatory of IncRNAs is lower than the extent of background mRNA:trans-mRNA interactions.Gene Ontology(GO)classification of up-regulated IncRNA:cis-mRNA pairs showed enrichment for heart development,while down-regulated lncRNA:cis-mRNA pairs showed enrichment for other functional categories.These results indicated that the lncRNAs identified in our study have similar potential cis and tans-regulatory of coding genes.Long non-coding RNAs are associated with specific chromatin states in the fetal heartDistinct chromatin states can be used to annotate genomic sequence,and can demarcate regulatory and transcribed genomic sequences.We investigated chromatin states at the lncRNAs and compared with those associated with protein-coding genes and lncRNAs utilizing publically available ChIP-Seq catalogues.We calculated the frequencies of each chromatin state in promoters of all transcripts specifically in the fetal and adult human heart.Novel lncRNAs were H3K4me1 in fetal heart when compared with known-coding and lncRNA genes,Conversely,novel lncRNAs were significantly less associated with the polycomb repressed mark(H3K27me3).Besides,Genes in fetal heart were more frequently associated with active enhancer states(H3K4me1)alone when compared with adult.To explore whether the differences in chromatin states between fetal and adult were functional specific,we performed GO analysis to the genes with fetal specific chromatin states.The enrichment of genes with fetal specific chromatin states were associated with development and cell cycle.Further,we analyzed chromatin state frequencies at the promoters and genes body of differentially expressed genes,either up or down-regulated between fetal and adult.Coding genes and known lncRNAs exhibited no obvious differential enrichments in chromatin states associated with the up-or down-regulated genes in both fetal and adult heart assessed.In contrast,novel IncRNAs exhibited a inactive chromatin states enrichment associated with genes that were down-regulated.Long non-coding RNA gene potentially involved in cell cycleFunctionally related genes involved in the same biological pathways or protein interaction networks are often regulated by similar transcription factors or other gene regulators.Thus,one approach to infer the potential function of novel genes is by determining whether their expression patterns correlate with those of known genes of certain function,based on coexpression analysis.Note that such analyses only provide tentative functional indications of a gene.However,they may help to formulate hypotheses for further experimental studies.We applied this scheme to examine the potential functions of lncRNAs.Based on the cell cycle related coding mRNAs,we found a IncRNA NONHSAG04200 insert PTGG1,which was associated with several important cell cycle related genes,such as CDK6,MAD2L1,E2F1 and so on.RNA folding analyses of this 3' region indicated a stem-loop structure,which might provide the necessary spatial conformation for the interaction.Further,motif analyses of this lncRNA showed that it may interacted with several cell proliferation related genes,such as JUN,LEF1 and NFATC2.Intriguingly,the chromatin state of this IncRNA was significant increased enrichment of H3K4me1,H3K4me3 and H3K36me3 compared with the loci in adult.Overall,this lncRNA may played a important role in cell proliferation during human heart development.LncRNA-PTTG1 was conserved between human and ratThe reslut of Real time PCR showed that the level of lncRNA-PTTG1 was significantly higher in fetal rat heart than in adult rat heart.In addition,we found that the IncRNA-PTTG1 was expressed in isolated rat cardiac myocytes,and the level gradually decreased with increases of days after birth.The levels of IncRNA-PTTGI in post natal-1 were higher than post natal-1,post natal-3 and post natal-7(both P<0.001).lncRNA increase CM proliferation in vitro and vivo Cultures of neonatal rat cardiomyocytes were transfected with Adenovirus Vector Overexpressing lncRNA-PTTG1.After 72 h,the cells were stained for sarcomeric a-actin in to distinguish cardiomyocytes,for the proliferation antigen Ki-67 and for 5-ethynyl-2'-deoxyuridine(EdU).Analysis was performed to selectively quantify the number of proliferating cardiomyocytes(a-actinin+,Ki-67+ or EdU+).The percentage proliferating of cardiomyocytes in the group of Ad-lncRNA was significantly higher than that in vector control group and control group(32%vs.3%;and 32%vs.6%,both P<0.05).Intramyocardial injection of neonatal rats with Adenovirus Vector Overexpressing IncRNA-PTTG1,EDU marked positive cardiomyocytes and Ki-67 marker-positive cells was significantly increased cardiac overexpression lncRNA EDU markers of myocardial cells positive cells of the total cell 33%of the number,and the virus control group and the control group proliferation of cardiomyocytes were 5%and 6%,and there are significant gaps between the virus control group and control group(P<0.001).Further,we explored whether IncRNA-PTTG1 promoted the proliferation of cardiomyocytes in vivo.Analysis of EdU incorporation and Ki-67 revealed a marked increase in the number of EdU+ cells in the neonatal rat hearts after transfected with Adenovirus Vector Overexpressing IncRNA-PTTG1(P<0.001).ConclusionWe firstlyprovided a genome-wide profiling of differential expressed mRNA and IncRNAs between fetal and adult heart.These mRNA and lncRNAs were related to function of cell cycle.In addtion,we provided evidence that IncRNA-PTTG might play important roles in heart regeneration.The InRNA-PTTG1 was conserved between human and rat,and promoted the proliferation of cardiomyocytes in vivo and in vivo. |
PDF Full Text Request