Performing Department
(N/A)
Non Technical Summary
Public interest in gluten-free foods and beverages has increased dramatically in recent years. Grain sorghum [Sorghum bicolor (L.) Moench] is a common ingredient in many gluten-free products. As a low-input, heat- and drought-tolerant crop, sorghum is playing an important role in climate adaptation. However, improving sorghum for human consumption requires clarifying the gene networks underlying the accumulation of macromolecules that influence end-use quality and nutrition, especially starch, protein, and oil. Here, we propose to perform the first systems study of the networks regulating the production of starch, protein, and oil in sorghum grain by dissecting endosperm and embryo development in a grain quality mapping population. The goals are to clarify the pathways related to sorghum grain starch, protein, and oil biosynthesis and to identify potential gene targets (alleles) for sorghum grain quality improvement. Specifically, a grain quality Multi-Parent Advanced Generation Intercross (MAGIC) population will be used in multiomic network analyses (transcriptomes, metabolomes) at five stages of endosperm and embryo development. Network analyses of transcriptomes of seed development in this population will identify the interaction of genes underlying the accumulation of these three high-value biomolecules. Field studies will be used to validate the analysis and to ensure that we are able to relate variation in gene expression with variation in phenotypic outcomes relevant to plant breeding. Target alleles (those that are identified in the network analysis and confirmed in the field to confer high vs. low contents of starch, protein, and oil) will be identified to assist plant breeding.
Animal Health Component
(N/A)
Research Effort Categories
Basic
100%
Applied
(N/A)
Developmental
(N/A)
Goals / Objectives
The goal of this project is to elucidate the physiological and regulatory mechanisms underlying the production of starch, protein, and oil in sorghum grain to assist breeders. Three specific aims are revised as below:Aim 1. Characterization of endosperm and embryo structure and physiochemical properties in the sorghum grain quality mapping population. Six MAGIC parental lines with different grain quality features and two bi-parental recombinant inbred lines (RILs) will be selected for the metabolism analysis and SEM visualization. The plants will be planted at the research farm of Texas Tech University during the summer of 2023. GC-MS and SEM grain structure imaging will be performed at five developmental stages.Aim 2. Reveal the dynamic transcriptome landscapes of sorghum endosperm and embryo development.Matching RNA-seq data will be generated using the same tissue at the five development stages from Aim 1. By comparing the transcriptome profiles of seed development in the eight lines with their grain chemical composition, we will be able to answer the questions: 1) How do the transcriptome profiles differ among the genotypes during grain development? 2) How many genes are specifically involved in seed development in this population?Aim 3. Identify conserved genes controlling grain quality in thePoaceaeusing reverse genetics.The functions of sorghum orthologs of 30 maize and rice grain quality genes will be validated using EMS mutants as the backbones of the network analysis in Aim 4.Aim 4. Identify potential grain quality breeding targets by comparing gene networks resulting in high or low production of starch and protein.Networks of starch, protein and oil biosynthesis will be built and compared among different genotypes using the metabolism and transcriptome data. The key question to answer is which genes are primarily responsible for tradeoffs between carbon allocation pathways influencing starch, protein, and oil content in mature grain. A selection of 20 genes will be validated from the network using the mutant population
Project Methods
Aim 1:A graduate student (TBD) will be responsible for growing and evaluating the parents and selected 6 RILs at the Texas Tech University research farm. Plants will be produced in the field so that gene networks can be assembled that are representative of typical farm environments. Plants will be grown in two-row plots (experimental units; 6-m rows with 72-cm row spacing) in a completely randomized design using conventional agronomic practices for drip irrigation. Three randomly-selected, representative plants (biological replicates) from each MAGIC parent and RIL (F4generation) will be used for total RNA extraction at five grain developmental time points (Aim 2), scanning electron microscopy (SEM), and wet chemistry.Endosperm structure imaging will be conducted through the TTU Microscopy Core Facility as described in published protocols for crop grain SEM imaging(Yu et al., 2016, Liu et al., 2019). Training for the graduate student will be provided by the Core Facility. A microtome will be used to dissect immature (14-DAP) and mature (35-DAP) grain bilaterally to view embryo and endosperm structures. Sections will be placed in fixative (5mL 38% formalin, 5mL glacial acetic, 90mL 70% ethyl alcohol) for more than 12h. A dehydration step will follow by soaking the samples in 70% ethanol solution (20min), 80% ethanol solution (20min), 90% ethanol solution (overnight), and 100% ethanol solution (20min). The samples will then be treated stepwise for 20min each in mixtures of ethanol andisoamyl acetatewith ratios 3:1, 1:1, and 1:3 before soaking in isoamyl acetate. Finally, critical-point drying will be performed for SEM observation on a Hitachi S-4700 II Scanning Electron Microscope (SEM) at the University of Nevada, Reno. MIRS, wet chemistry and fatty acid analysis will be conducted on whole, mature grain in triplicate 20-g samples at the USDA-ARS Grain Quality and Structure Research Unit in Manhattan, KS.Aim 2:Transcriptomes will be characterized at five seed developmental stages. The sorghum embryo sac is filled with cellular endosperm by 3-DPA(Paulson, 1969), so this will be the earliest stage that we consider. Since the endosperm and embryo cannot be separated at 3-DPA, the samples will be collected as whole seeds. 7-DPA is also considered early development, but the endosperm and embryo can be differentiated and will be harvested separately. The subsequent three stages include major grain filling stages: 14-DPA (milk stage; also a key time point for SEM analysis), 21-DPA (soft dough) and 28-DPA (hard dough). Sorghum seeds will reach physiological maturity by 35-DPA.Three representative plants (biological replicates) from each parent and RIL (14 genotypes total) will be selected for the destructive harvest of embryos and endosperms at each of the five time points. Five seeds per panicle (technical replicates) will be dissected under a microscope at each time point, pooled into a 2-mL Eppendorf tube, flash frozen in liquid nitrogen, and stored at -80 C until RNA extraction and sequencing. Each total RNA sample will be split into two tubes for Iso-seq and RNA-seq. The three replicates for full-length cDNA sequencing by Iso-seq will be pooled together as one sample. High-quality samples (RIN > 6.5) will be submitted to a commercial company for library prep followed by sequencing. Each RNA-seq sample will generate about 20million, 100bp pair-end reads.RNA-seq reads will be used to quantify the differential expression levels of genes during sorghum grain development. First, the RNA-seq reads will be mapped to the full-length cDNA sequences from Iso-seq with STAR. The expression level in FPKM and the differential expression among different genotypes and developmental stages will be calculated using the Cufflinks package. Endosperm and embryo developmental stage-specific genes will be identified by comparing them with other published sorghum transcriptome data that was generated from 11 tissues covering a range of sorghum growth(Wang et al., 2018).Aim 3:Validate the functions of sorghum orthologs of maize and rice grain quality genes using EMS mutants. Selection of the 30 genes as the candidate skeletons for sorghum pathways: As many starch synthesis genes and regulators have been identified in maize and rice(Huang et al., 2021), the genes overlapped with sorghum starch QTLs from Sorghum QTL Atlas(Mace et al., 2019)were first included. The protein and oil synthesis genes are from the summaries of thecited publications(Liu et al., 2016, Li and Song, 2020). The sorghum orthologs were identified using the Gramene database(Tello-Ruiz et al., 2022). To perform co-segregation analysis, the M3 generation of these lines will be planted at the Texas Tech research farm. We will genotype about 100 plants from each EMS line about 10-12 days after germination, using the first leaf to extract gDNA and Kompetitive allele specific PCR(KASP) markers to confirm expected mutations. Plants will be selfed to achieve homozygosity of the expected mutations, if necessary. Measurement of grain starch, protein, and fat content via wet chemistry will be conducted as in Aim 1 by Dr. Scott Bean at the USDA-ARS in Manhattan, KS (see Letter of Support). For each gene, we will assess the grain chemistry profile of 20 mutant plants. The grain from 20 wild type BTx623 plants grown in the same environment will serve as a control. These phenotypes will be compared between the mutants and the wild type. The mutants with statistically significant differences in starch, protein, and/or oil contents will be assessed in light of the functions of their orthologs in rice and/or maize. If similar changes in grain chemistry are present, this will provide evidence that the sorghum genes have similar functions. The expression level of these candidate genes will be studied in the transcriptomic data sets from the MAGIC population as an additional step to associate their relationship with starch, protein, and/or oil synthesis. The final datasets will include genes that are functionally validated by both EMS mutants and native mutants represented in the MAGIC population's transcriptomic data.Aim 4:The co-expression networks will be constructed using a Weighted correlation network analysis (WGCNA)approach(Zhang and Horvath, 2005). WGCNA defines gene clusters by hierarchically clustering the gene-gene proximity matrix built using a method called topological overlap measure (TOM). Co-expression networks will allow a modularized analysis of biological processes to discover regulatory genes or modules of vital traits. The package cFinder(Niklas et al., 2015)will be used to search and visualize the networks around the genes with functions related to grain quality from Aim 3 (gene-centered networks). The function of each module will be predicted by gene set analysis through GO and KEGG. This analysis will lay out the landscape of gene-gene interactions during sorghum seed development, and in particular, how they differ when higher or lower amounts of starch, protein, and oil result in the mature grain (as determined by wet chemistry).Network validation: As networks will be started from genes validated by reverse genetics analysis as skeletons, new genes will be identified in the starch, protein, and oil pathways through network analysis. To build evidence-based networks, a total of 30 new genes will be selected for validation. The genes that are not characterized in other crops will be prioritized for validation. The method of phenotyping and genotyping of mutants will be the same with Aim 3. For each grain trait, we expect that ten genes will be validated by the EMS mutants using our co-segregation approach for the evidence based network.