2013 Annual Report
1a.Objectives (from AD-416):
1. Use virus induced gene silencing (VIGS) to turn off candidate resistance genes in the Brown Stem Rot resistant genotypes. Test silenced plants for susceptibility to Brown Stem Rot.
2. Use Solexa transcript sequencing to determine which resistance genes in the Rbs3 locus are expressed preferentially in the resistant parent and correlate with Brown Stem Rot resistance.
3. Use the Williams 82 genome sequence to fine map Rbs3 in a population segregating for Brown Stem Rot resistance (BSR101 x PI437654, developed by Lewers et al. 1999). Develop SSR markers to be used for marker-assisted selection of BSR resistance.
1b.Approach (from AD-416):
Our approach will identify candidate Brown Stem Rot resistance genes for Rbs3. We will leverage the Williams 82 genome sequences and use it to identify candidate genes from the resistant parents. We have compared sequence from the Williams 82 soybean genome to markers that have been used to map Rbs3. The region corresponding to Rbs3 has been bioinformatically screened for the presence of genes with similiarity to known resistance genes. The sequences of these R-genes has been used to develop Virus Induced Gene Silencing Constructs. These constructs will be tested on Brown Stem Rot resistant genotypes to determine if we can turn off resistance to Brown Stem Rot. This information will identify a cluster of genes responsible for resistance. To identify the actual genes, we will use solexa transcript sequences to compare Brown Stem Rot resistant and susceptible genotypes before and after infection with Brown Stem Rot. Significantly differentially expressed genes in the Rbs3 cluster will be candidates for Rbs3. In addition, we will use fine mapping to further define the candidate genes.
Plant mechanisms for controlling infection by Phialophora gregata, the causal agent of Brown Stem Rot (BSR) in soybeans, are poorly understood. Unlike most soybean-pathogen systems, scoring resistance or susceptibility to BSR takes six to seven weeks. It is unclear if resistance is not induced until that time or if we are unable to detect early resistance by visual inspection. Therefore we are using Virus Induced Gene Silencing (VIGS), Quantitative Trait Loci (QTL) mapping and expression analyses to characterize resistance.
In order to identify the gene(s) corresponding to Rbs3, we took advantage of the soybean whole genome sequence. Markers used map Rbs3 were screened against the genome and identified a 470,000 base pair region, containing a cluster of candidate resistance genes. However, when we examined the regions next to this gene cluster, we found it was part of an even larger cluster containing >80 resistance genes (R-genes). All of the genes were similar to an apple’s scab fungal resistance gene. Using sequence differences we divided the genes into three distinct classes. Members of each of the three classes were present in the Rbs3 locus. Sequence alignments of the three gene classes were then used to develop VIGS constructs specifically targeting each class. The VIGS constructs were sent to our collaborator to see if they could silence BSR resistance.
In order to perform microarray analyses, we need to generate tissue from infected and mock infect plants, resistant and susceptible to BSR. Our group focused on early responses to BSR. One hundred seeds each of a BSR-resistant (BSR 101) and susceptible genotype were grown in individual containers in a randomized plot design in a growth chamber. Approximately two weeks later, 50 plants of each genotype were infected with P. gregata using the standard stem injection method. This left 50 plants of each genotype to be used as mock-infected controls. At time points 24, 48, 72 and 288 hours after infection, six plants from each genotype, both infected and mock-infected were sampled. Samples included the stem (from infection site to cotyledons) and first trifoliates. Samples from each plant were stored and frozen separately. After the 288-hour time point, the remaining plants were maintained in the growth chamber to allow BSR infection to proceed. At six weeks post infection, these plants were phenotyped for resistance and susceptibility. As expected, all infected plants from the resistant genotype were classified as resistant while all infected plants from the susceptible genotype were classified as susceptible. Non-infected plants showed no disease symptoms.
Once all the plant material was generated, we harvested mRNA from all of the leaf plant samples. RNA samples from the 24, 72 and 288 time points were analyzed by microarray to determine which genes changed their expression pattern in response to infection. In an initial test, we focused on the 24 and 288 hour time points including RNA from P. gregata-infected and mock-infected BSR101 (resistant) and PI43764 (susceptible) leaves. Using this approach, we identified 1,969 genes that were differentially expressed in BSR101 at the 24-hour time point. No differentially expressed genes were identified at the 288-hour time point. In comparison, no differentially expressed genes were identified in the susceptible line at 24 hours after infection and only 105 genes were affected in the 288-hour time point. Bioinformatic analyses were used to determine the function of the genes and compare their expression across the two parents and time points. Another approach to identify genes responding to P. gregata infection is to identify genes whose expression changes over time. Using this method, we identified 6,289 genes whose expression changed over time in the resistant line BSR101 and 8,108 genes in PI43764. We are currently mining this data to identify candidate defense genes that could be tested using VIGS. Data analysis on the other time points is ongoing.