Bioinformatics Mining Of Genes Associated With Seed Oil Synthesis And Root Nodule Symbiosis In Soybean

Soybean not only has the characteristics of nitrogen fixation but also has high oil content in the seeds,the mechanisms of which have been a hot spot of scientific research.Up to now,the mechanisms of acyl-lipid metabolism and root nodule symbiosis have been widely studied,and a series of key genes in these two pathways have been successfully cloned.However,the synthesis of lipid and the establishment of the root nodule symbiosis both are complex biological processes that involve hundreds of genes and previous studies mostly focused on a single enzyme or transcription factor(TF).Therefore,there is a need for genome-wide study to better investigate the mechanism of lipid synthesis and nitrogen-fixing symbiosis.We combined omics data i.e.genomics,transcriptomics and proteomics data to explore the evolution mechanism of high oil content in high-oil plants and root nodule symbiosis of legumes.The work in this thesis covers two parts as follows:1)We identified high-oil dicots-specific genes that relate to lipid synthesis,differentially expressed genes(DEGs)during seed development,DEGs between high-and low-oil content soybean accessions and TFs that are likely to be associated with oil accumulation.2)We predicted the protein-protein interactions between B.dr azoefficients and G.max by jointly using the interolog and the domain-based methods.Then GO annotation and gene expression data were utilized to validate the quality of predicted PPI network.Lastly,we classified the protein functions using PANTHER overrepresentation test and GO enrichment analysis,identified the hubs in the network and analyzed the subnetworks of protein interactome to identify the candidate proteins that were possibly related to soybean-rhizobium symbiosis.In addition,we discuss how these predicted protein interactomes can help to better understand the mechanism of soybean root nodule symbiosis.The main results are as follows:1)The genomic data of four high-oil dicot species and three low-oil grasses were used to cluster ortholog groups(OGs).As a result,1,534 OGs were defined as four high-oil dicot-specific clusters which include 4,051,2,758,1,731 and 2,152 genes in G.max,G.raimondii,R.communis,and A.thaliana,respectively.In soybean,54 high-oil dicot-specific genes were shown to participate directly in acyl-lipid metabolism,and 93 high-oil dicot-specific genes were found to be related with acyl-lipid metabolism by using gene ontology(GO)enrichment analysis and pathway level co-expression(PLC)analysis.For genes involved in core lipid synthesis pathways,40 were highly up-regulated during seed oil rapid accumulation period and were mainly involved in the initial fatty acid synthesis,triacylglyceride assembly and oil-body formation,for example,ACCase,PP,DGAT1,PDAT1,OLEs and STEROs.Phylogenetic analysis revealed distinct differences of oleosin in patterns of gene duplication and loss between high-oil dicots and low-oil grasses.We also conducted a PLC network analysis to identify TFs that related to lipid synthesis.As a result,seed-specific expressed TFs GmTZF4,GmGRF5 and ABI5 were predicted to regulate seed oil content.Differential gene expression analysis between high-and low-oil soybean accessions was used to validate the above candidates.As a result,17 of the 93 high-oil dicot-specific lipid-related genes,17 of the 40 highly up-regulated genes during seed oil accumulation period,and seven seed-specific TFs were found to be differentially expressed between high-and low-oil soybean accessions.2)Using experimentally verified PPIs of the seven model organisms(Arabidopsis thaliana,Caenorhabditis elegans,Drosophila melanogaster,Escherichia coli K12,Homo sapiens,Mus musculus and Saccharomyces cerevisiae)and combined with interolog and domain-based methods,we predicted 5115 PPIs between soybean and B.drazoefficients USDA 110,including 2291 soybean proteins,290 B.drazoefficients USDA 110 proteins and 60 conserved PPIs.Quality assessment showed that all the proportion of score 1.0 in simJCBP,simICMF,and simJCCC(33.29%,33.50%and 52.48%,separately)were significantly higher in predicted soybean PPIs than that in random protein pairs(6.13%,13.81%and 7.68%);The proportion of PPIs with PCC scores>0.6(23.84%)was significantly larger than that of the random protein pairs(13.80%);71.80%soybean genes in the predicted PPI networks were expressed in nodule with FPKM>5.All the above results suggested that predicted PPIs have a high reliability.To determine whether some biological function biases exist between the B.diazoefficens and soybean proteins in the predicted PPI network,we classified the proteins using PANTHER overrepresentation test,conducted biological process GO enrichment analysis,identified hubs in two species and analyzed the sub-networks possibly related to symbiosis.The results show the following:1)In B.diazoefficens USDA 110,proteins in the PPI networks are mainly ion channel and transporters of carbohydrate and cation.In soybean,these proteins interacted with bacteria implying they are involved in the metabolic process,signalling,response to stimulus and transport.2)In soybean,hubs were mainly ion channel,calcium-binding proteins,tubulin and heat shock proteins;in B.drazoefficients USDA 110,hubs were metalloprotease,aquaporin and cation-transporting ATPase,which may be symbiotic effectors and play important roles in symbiosis.3)Subnetwork analysis showed that B.diazoefficens USDA 110 BAC45924,2 soybean 14-3-3 proteins(SGF14g and SGF14k)and 18 soybean SNARE proteins play roles in the process of symbiosis.SGF14g and SGF14k possibly regulate DctA.Soybean SNARE proteins may regulate the ion channel proteins of B.diazoefficens USDA 110.