Screening Of LncRNA And CircRNA Related To Hair Shedding Traits In Sheep And Construction Of CeRNA Regulatory Network

In this study,5 female Dorper sheep with extreme hair shedding phenotype（S group）and 3 female Dorper sheep with non-shedding phenotype（N group）were used as experimental animals and their skin tissues were sampled on September 27,2019,January 3,2020 and March 17,2020,respectively,followed by RNA-sequencing（RNA-seq）technology,bioinformatics and other approaches.The lncRNAs and circRNAs which associated with hair follicle cyclic growth and development were screened by differential expression analysis,cluster heatmap analysis,expression pattern analysis,target relationship prediction and GO and KEGG enrichment analysis.And construct the corresponding ceRNA regulatory network.Finally validated by qRT-PCR.The results were as follows:1.The three time points of Dorper sheep were divided into the following periods:shedding Dorper sheep in order of anagen（S1）,telogen（S2）and early anagen（S3）;non-shedding Dorper sheep in order of anagen（N1,N2）and slow anagen（N3）.The lncRNAs expression was measured by RNA-seq,and a total of 395 differentially expressed（DE）lncRNAs were screened by differential expression analysis of 9 comparative groups of the same group at different time points（S1-vs-S2,S1-vs-S3,S2-vs-S3,N1-vsN2,Nl-vs-N3,N2-vs-N3）and different groups at the same time points（S1-vs-N1,S2-vs-N2,S3-vs-N3）in the S and N groups.Cluster heatmap analysis of DE lncRNAs revealed two expression patterns,one type of lncRNAs were highly expressed during anagen（A pattern）and the other type of lncRNAs were highly expressed during telogen（T pattern）,and 57 A pattern IncRNAs and 72 T pattern lncRNAs were screened.2.The target genes of the 129 A or T pattern lncRNAs were predicted by cis,trans and antisense,and the Pearson correlation coefficients（PCCs）were calculated between the cis and antisense predicted lncRNAs and the target genes,and the lncRNA-target gene pairs with |PCCs|>0.8 were retained for subsequent analysis.118 lncRNA-target gene pairs were screened by trans analysis,19 lncRNA-target gene pairs were screened by cis analysis,and 19 lncRNA-target gene pairs were screened by antisense analysis.3.A total of 99 target genes predicted by the three methods were analyzed by differential expression analysis and expression pattern analysis based on the pre-mRNA sequencing data.99 target genes were differentially expressed genes（DEGs）,and 63 DEGs with A pattern（Anagen,high expression in anagen）and 28 DEGs with T pattern（Telogen,high expression in telogen）were screened by expression pattern analysis.GO and KEGG enrichment analysis of the 91 DEGs that fit the A or T pattern identified 6 crucial lncRNAs（MSTRG.12818.1,MSTRG.13824.1,MSTRG.13836.1,MSTRG.13826.1,MSTRG.49.1,MSTRG.1616.1）play important roles in anagen of hair follicle by regulating 7 corresponding target genes of the Estrogen signaling pathway and PI3K-Akt signaling pathway（KRT28,KRT33A,KRT40,KRT27,KRT25,KRT31,Lpar6）;4 crucial lncRNAs（MSTRG.15931.1,MSTRG.22447.1,MSTRG.11626.2,MSTRG.23796.1）play important roles in telogen of hair follicle by regulating 3 corresponding target genes（PTPRM,ELMO1,Pip5k1c）distributed in 5 pathways including Cell adhesion molecules（CAMs）,Bacterial invasion of epithelial cells,Chemokine signaling pathway,FcγR mediated phagocytosis and Regulation of the actin cytoskeleton.4.The 10 crucial IncRNAs were predicted to target 2502 miRNAs,and 428 lncRNA-miRNA pairs（10 lncRNAs,228 miRNAs）were screened with PCCs<-0.6 by combining the pre-miRNA sequencing data.These 228 miRNAs were predicted to target 5604 target genes,and 765 miRNA-target pairs（98 miRNAs,262 target genes）were screened for PCCs<-0.6 by combining the pre-mRNA sequencing data.Finally,expression pattern analysis was performed for these 262 target genes,and target genes consistent with the corresponding lncRNAs expression patterns（19 A pattern lncRNAs,116 T pattern lncRNAs）were retained,and GO and KEGG enrichment analyses were performed on these 135 target genes to screen crucial target genes,and finally a ceRNA regulatory network consisting of 10 lncRNAs,11 miRNAs and 10 target genes was constructed.5.The expression levels of 8 DE lncRNAs and 4 DEGs were quantified by quantitative real-time fluorescence PCR（qRT-PCR）,and the results showed that the qRT-PCR quantification levels were basically consistent with the trend of transcriptome sequencing results.6.A total of 1450 differentially expressed（DE）circRNAs were screened by RNA-seq and differential expression analysis.Cluster heatmap analysis and expression pattern analysis were performed on DE circRNAs to analyze the regulatory mechanism of T pattern（Telogen,high expression in telogen）circRNAs,and a total of 78 T pattern circRNAs were screened.GO and KEGG enrichment analysis was performed on 78 source genes conforming to T pattern circRNAs,and 19 circRNAs corresponding to 19 source genes in 25 pathways that were significantly enriched were selected for the construction of ceRNA regulatory network.7.19 candidate circRNAs were predicted to target 463 miRNAs,and combined with the pre-miRNA sequencing data,158 pairs of circRNA-miRNA（1 7 circRNAs,1 13 miRNAs）were screened with PCCs<0.6 as the criteria.Their target genes were predicted for these 113 miRNAs,and 76 miRNAs targeting 365 target genes,and combined with the pre-mRNA sequencing data,433 miRNA-mRNA pairs（44 miRNAs,194 target genes）were screened with PCCs<-0.6 as the criteria.Finally,expression pattern analysis was performed on these 194 target genes,and 96 target genes conforming to T pattern（Telogen,high expression in telogen）were retained,which were screened for crucial target genes by GO and KEGG enrichment analysis,and finally a ceRNA regulatory network consisting of 10 crucial circRNAs（novel_circ₀08094,novel_circ₀15678,novel_circ₀16105,novel_circ₀10394,novel_circ₀10013,novel_circ₀15092,novel_circ₀15374,novel_circ₀01644,novel_circ₀05105,novel_circ₀05706）,16 miRNAs and 10 target genes was constructed.8.qRT-PCR was used to quantify the expression levels of the 8 DE circRNAs,and the results showed that the qRT-PCR quantification levels were generally consistent with the trend of transcriptome sequencing results.In this study,we explored two expression patterns（A pattern and T pattern）during the cyclic growth and development of sheep hair follicles,and screened several lncRNAs and circRNAs related to the cyclic growth and development of sheep hair follicles,and conducted an in-depth study on their regulatory mechanisms,which can help to resolve the molecular regulatory mechanisms related to shedding traits in sheep.