| Muscle development is a very complex process, as it is influenced by a number of genesinvolved in it. In our study, high-throughput transcriptome sequencing, iTRAQ and liquidchromatography-tandem mass spectrometry technologies were employed to explore genemRNA expression and protein expression during prenatal (135days post conception,Emb135d) and postnatal (30months after birth,30M) longissimus muscle development inQinchuan beef cattle, and then we performed Gene Ontology (GO) and KEGG pathwayanalysis by running queries for each expressed gene and protein against the reference database.Besides,47differentially expressed genes were validated by quantitative PCR. At last, weperformed correlation analysis of proteomics and transcriptomics expression. The main resultsare indicated as follows:1. A total of20255767and20427874hight quality clean reads from Emb135d and30Madult cattle groups longissimus muscle RNA-seq were mapped the bovine reference genomedatabase. Mapping efficiency of reads from reference genome were79.11%and77.92%respectively.2. The differentially expressed genes were selected based on the expression profiles andthe following criteria: if the change in gene expression levels between Emb135and30M wasgreater than or equal to two-fold (|log2Ratio|≥1) and if the false discovery rate (FDR) wasless than or equal to0.001. In summary, we restricted6800genes that exhibited differentialexpression between Emb135d and30M adult cattle groups, Emb135group as a control, ofthese, the expression levels of1893genes were up-regulated in the30M adult cattle groups,the remaining4907genes were down-regulated in the30M adult cattle groups. The results ofanalyzing the GO functional annotations are presented that95terms corresponding tobiological processes,26terms corresponding to molecular functions, and71termscorresponding to cellular components. In the biological processes category, the mostimportant enriched terms are related to developmental process. Pathway analysis showed that15pathways were significantly enriched for differentially expressed genes.3. To validate the expression profiles obtained using RNA-Seq, quantitative real-timePCR (qRT-PCR) was performed on47genes, overall, of these47selected genes, those that were differentially expressed showed similar patterns of mRNA abundance in the RNA-Seqanalysis and qRT-PCR. Furthermore, the expression level of47genes in liver, lung, spleen,kidney, heart, stomach, intestines, longissimus muscle and abdominal fat of Emb135d (n=3),newborn (day0at birth, n=3) and30M (n=3) cattle. were detected with qRT-PCR. The resultsshowed that25genes were specifically expressed in the examined (longissimus dorsi muscle)muscle tissue, but displays lower or undetectable level of activity in other tissues of bovine.Similarly,4genes were specifically expressed in the examined (abdominal fat) fat tissue.4. We compared the gene model with the existing annotated gene to extend the gene’s5’-and3’-ends. In the30M group,2157genes were extended at the5’-end, and3531genes hadan extension at the3’-end. In comparison, in the Emb135group,2571genes were extended atthe5’-end, and4625genes had an extension at their3’-end. We also detected an extensivenumber of novel transcript units. In total, we obtained24464and29994novel transcript unitsin the pooled samples for30M and Emb135d, respectively. Four main types of alternativesplicing exist in Qinchuan cattle: exon skipping, intron retention, alternative5’ splice site, andalternative3’ splice site. Alternative3’ splice site is the most common type of alternativesplicing event, accounting for approximately40%of all alternative splicing events in cattle; incontrast, intron retention is the least common type of splicing event in the Emb135d and30Mgroups.5. The SNPs can be identified in the consensus sequences by a comparison with thereference. In total,31334putative SNPs were identified for the30M groups. Similarly, forEmb135groups,30,618putative SNPs were identified. To validate the putative SNPsobtained using RNA-Seq, we selected CSRP family genes as candidate gene. Compared withbovine reference sequence, eight SNPs were identified within the bovine CSRP family genesby sequencing. The eight SNPs of the bovine CSRP family genes were genotyped usingPCR-RFLP and forced PCR-RFLP methods, as a result, the result of restriction endonucleasedigestion was accordance with the sequencing. Furthermore, statistical analysis was onlyperformed on the basis of records of phenotypic traits in Qinchuan beef cattle (n=404), theresults of association analysis demonstrated that there were significant genetic associationsbetween each SNP different genotypes and body height, body length, hip width, slaughterweight, carcass weight and dressing percentage traits.6. MS output data from iTRAQ run were searched and analyzed using the Mascotsoftware (2.3.02version) by alignment with the UMD3.1reference protein data fromensemble database. Summary statistics of bovine longissimus muscle proteins were as follows:total number of MS/MS spectra were115380;14250spectra that can aligned to referenceprotein database;11515spectra were uniquely matched to the reference peptide;5327peptides were identified;5045uniquely peptide sequence also were identified; a total of1321 proteins were identified during muscle development.7. The relative molecular weights (MW) of all identified proteins were classified, theresults showed that MW of the most proteins are greater than100kDa (280proteins), MW ofonly3proteins are less than10kDa, the remaining proteins were classified as MW rangedfrom10to100kDa. Besides, we observed distributions of different length peptides, peptideswith11amino acid residues is the most common length accounting for approximately12%ofall peptides in cattle; in contrast, peptides with22or more amino acid residues is the leastcommon length accounting for approximately1%of all peptides in the Emb135d and30Mgroups, similarly, peptides with7or less amino acid residues also is the rare length; theremaining peptides were classified as length ranged from9to20amino acid residues. Wefurther observed that the majority proteins within10peptides and the amount of proteinsdecreased as the increase of peptides.8. The differentially expressed proteins were selected based on the expressionabundances and the following criteria: the change in protein expression levels betweenEmb135d and30M was greater than or equal to1.5-fold (log2fold change) and P-value wasless than0.05. In summary,390proteins exhibited differential expression between Emb135dand30M groups. Of these, the expression levels of67proteins were up-regulated in the30Mgroup with respect to the Emb135d group, the remaining323proteins were down-regulated inthe30M group with respect to the Emb135d group. The results of analyzing the GOfunctional annotations are presented that102terms corresponding to biological processes,12terms corresponding to molecular functions, and21terms corresponding to cellularcomponents. In the biological processes category, the most important enriched terms arerelated to developmental process. Based on KEGG pathway enrichment analysis identified7significantly enriched metabolic pathways or signal transduction pathways related todifferentially expressed proteins. The pathway term showing the highest level of significancewas protein processing in endoplasmic reticulum. Furthermore, after protein processing inendoplasmic reticulum, the terms oocyte meiosis, prion diseases, streptomycin biosynthesis,neurotrophin signaling pathway, citrate cycle (TCA cycle), and ribosome exhibited the highestsignificance in the pathway analysis.9. COG functional classification of all identified proteins were performed, the resultsshowed that those proteins were classified into24clusters of COG categories, the most termsare related to general function prediction only (Cluster R), followed by posttranslationalmodification, protein turnover, chaperones (Cluster O); translation, ribosomal structure andbiogenesis (Cluster J); and energy production and conversion (Cluster C).10. To compare the proteome with the transcriptome, we matched all differentiallyexpressed proteins with mRNAs. We observed four types of protein and mRNA expression pattern, including down-regulated proteins corresponded to down-regulated changes in theabundance of their cognate transcripts, up-regulated proteins corresponded to up-regulatedchanges in the abundance of their cognate transcripts, down-regulated proteins correspondedto up-regulated changes in the abundance of their cognate transcripts, up-regulated proteinscorresponded to down-regulated changes in the abundance of their cognate transcripts. Wealso observed a significantly positive correlation of r=0.6061. At last, cluster analysis ofdifferentially expressed protein and transcript abundance were performed, we observed thevast majority of the up-regulated (down-regulated) proteins was associated with a higher(lower) abundant transcripts, the minority of the down-regulated (up-regulated) proteinscorresponded to up-regulated (down-regulated) changes in the abundance of their cognatetranscripts.Above all, we obtained a high-quality beef cattle transcriptome and proteome data,thereby providing a valuable resource for better understanding the beef cattle gene and proteinexpression. These studies also provides a broad and novel vision of future research at themolecular level in cattle. |
