2012 Annual Report
1a.Objectives (from AD-416):
Objective 1: Analyze mRNA and storage oil profiles of cotyledons over the course of seed maturation to infer biological networks underlying soybean seed oil composition and content, and to predict their key gene candidates.
Objective 2: Analyze small RNA profiles of cotyledons over course of the seed maturation, and further predict small RNA candidates regulating soybean seed oil composition and content.
1b.Approach (from AD-416):
Objective 1: Soybean oil synthesis and deposition occur mainly in cotyledons and are differentially regulated over the course of seed maturation. The oil synthesis and deposition are accomplished through the concerted activities of many gene products and biological pathways that are primarily regulated at transcription levels. Gene expression patterns change very quickly over the course of an organism’s evolution if it is not subjected to functional constraints. Although it is not a universal rule, evidence suggests that expression patterns of many genes are intended to co-evolve with their biological functionalities. The co-evolution is reflected by a correlation of gene expression pattern with the related biological functions, and co-expression of functionally related genes such as those that encode proteins that reside in the same metabolic or signal pathways, or in the same cellular complexes under a variety of biological conditions (Stuart et al., 2003; Wei et al., 2006). The expression pattern correlation is widely used as criteria to predict biological functions of genes, functional relatedness between genes, and gene regulatory networks (Horan et al., 2008). We have examined transcriptomes and storage lipid profiles of cotyledons at six distinct developmental stages over the course of seed maturation.
Objective 2: We have conducted deep sequencing of small RNA populations in the same RNA preparations used for transcriptome analysis. It is believed that miRNAs function as master regulators in gene regulatory networks underlying diverse biological processes in Arabidopsis. However, much less is known about soybean miRNA species and their accumulation patterns over the course of seed maturation. As part of our effort to delineate regulatory networks and identify key genetic components controlling oil composition and content, we will conduct genome-wide characterization of small RNAs, particularly miRNAs, in cotyledons over the course of seed maturation. We will use a bioinformatic approach to analyze the small RNA sequences we have conducted to discover miRNA species and their accumulation patterns. MiRNA species recognize their target mRNA by high sequence complementation, and function mainly as suppressors of the accumulation of the target mRNA by directing the degradation of its functional target mRNAs in plants. Sequence complementarities and negative correlation of mRNA and miRNA accumulation patterns should offer a more effective approach to identify the functional target genes, and can be used to delineate the topology of small RNAs in the TF networks inferred in Objective 1. The miRNA species that locate in TF networks enriched with oil related genes or target oil related genes would be strong candidates for future investigations for oil composition and content improvement.
Production of storage oil in seeds requires concerted activities of many genes and biological pathways over the course of seed maturation. However, lack of knowledge on the intricate biological network and the availability of its key regulatory components has been a bottleneck in the effective application of both breeding and biotechnological approaches for soybean oil quality improvement. To address this problem we have undertaken a multi-faceted approach to determine profiles of message RNAs, small RNAs, DNA methylation, and major lipid species in soybean cotyledons over the course of seed maturation. We further identified the genes and biological pathways that are differentially regulated over the course of seed maturation, and inferred a set of transcription factor regulatory networks underlying soybean seed maturation.
To further discover DNA markers and genetic variation responsible for oil quality phenotypic differences among genotypes, we applied next generation DNA sequence technologies to determine the sequence and accumulation patterns of messenger RNAs produced in the soybean seeds from several genotypes varying in oil content and composition. We developed a range of bioinformatics tools to compare the sequence and expression data between the genotypes to identify variation in gene expression, gene sequences, RNA splicing, DNA insertion and deletions. DNA methylation plays a key role in gene regulatory programs. We constructed and sequenced bisulfate-modified DNA libraries for soybean cotyledon tissues at early-, mid- and late-maturation stages to determine genome-wide DNA methylation patterns. We also sequenced the entire genome of the Jack genotype. The data will be analyzed for integrating DNA methylation into the gene regulatory networks and assessing genome sequence variation between Jack and Williams 82. In an effort to validate the function of miRNAs, we analyzed a set of soybean seed degradome data available in the public domain, and showed that over 50% of the cotyledon miRNAs can be mapped to the cleavage region of their target transcripts. A novel and effectively bioinformatics algorithm was also developed to cluster biological processes with a similar molecular mode of action into the same network based on transcriptomic changes and to identify the genes that potentially participate in those biological processes.
A novel computer-based method for assigning gene activity to biological processes. ARS scientists in St. Louis, MO developed a novel and effective algorithm (a mathematical method) for genetic analysis to identify biological processes that share a similar mode of action, and to further predict the genes that are important to those processes. Each biological process is accompanied by changes in activities for hundreds to thousands of defined genes. Those so-called transcriptomic changes can be used as “fingerprints” to cluster biological processes with a similar mode of action into the same network. Using this approach, a total of 1993 distinct biological processes were clustered into 60 distinct networks and identified the associated potential genes and molecular pathways. The algorithm will enable plant researchers to accelerate the discovery of the genetic control of important biological processes such as seed maturation and oil production, and to apply this knowledge for crop improvements that will benefit producers and consumers.