Genetic characterization of Rpp1-mediated Asian soybean rust resistance using in-depth transcriptomic and genomic analyses

Soybean-rust interaction;Transcriptome profiling;Genome sequencing;Rpp1

Wu, Xing ; Hudson, Matthew, ; ,Walker, David ; Clougn ; Steven

关键词: Soybean-rust interaction;    Transcriptome profiling;    Genome sequencing;    Rpp1;   

Others  :  https://www.ideals.illinois.edu/bitstream/handle/2142/98297/WU-THESIS-2017.pdf?sequence=1&isAllowed=y

美国|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Asian Soybean Rust, caused by the fungus Phakopsora pachyrhizi, is one of the most destructive diseases affecting soybean yield worldwide. Rpp1 is one of the six known resistant loci that can confer qualitative resistance in soybean resulting in either red-brown lesions with no or limited sporulation, or immune phenotype of the incompatible interaction that show miniscule or no lesions and no sporulation. Due to the rapid changing nature of rust effector proteins to avoid being detected by the resistant genes, and to the spectrum of resistance-associated phenotypes observed on the same soybean genotype, successful discovery of the genes that can provide qualitative or quantitative resistance would benefit breeding programs, and would advance our understanding of the basic biology of the soybean-rust interaction. Two resistant soybean genotypes, PI 594760B and PI 561356, which both carry a resistant Rpp1 allele, and a dominant susceptible genotype, TMG06_0011, were used in the studies of this thesis. RNA sequencing was performed at six different treatments: mock, 24-hours, 48-hours, 72-hours, 96-hours and 120-hours post rust inoculation. Small RNA sequencing was conducted at four different treatments: mock, 24-hours, 48-hours and 72-hours post rust inoculation. Both transcriptional experiments were aimed to profile gene / small RNA expression and to identify candidate genes or small RNA loci that are involved in increasing either pathogen virulence or host resistance. To possibly identify and clone the Rpp1 allele(s) in PI 594760B and PI 561356, and to uncover the mechanism of the dominant susceptible phenotype, fosmid libraries were constructed, screened and sequenced, and a whole genome sequence assembly were conducted on TMG06_0011 using the new 10X Genomics technology.Due to the lack of biological replications, the statistical analysis of differential expressed genes was computed by combining the results from GFOLD and EdgeR by using time points within a treatment as replicates. By looking at the genes that were consistently down regulated after rust infection compared to mock treatment in all three genotypes, cell wall biosynthesis and lignin biosynthesis were found to be significantly enriched, suggesting these pathways could be potentially targeted by the rust effectors to attenuate defense responses. A clear negative correlation was observed between three miRNAs and their targets that are putatively involved in lignin biosynthesis, which indicated that the reduction of lignin biosynthesis gene expression might be due to the up regulation of the corresponding miRNAs. A total of 1110 soybean genes were defined as differentially expressed between resistant genotypes and the susceptible genotype. Among the genes that had different expression between genotypes, a total of 200, 312 and 367 genes that were also differentially expressed in the time course for PI 594760B, TMG06_0011, and PI 561356, respectively. A clear expression pattern was found in that the majority of the defense-related genes and putative signaling-related genes were up regulated in expression after rust inoculation for the two resistant genotypes. However, fewer genes in the same categories showed up-regulation in the susceptible genotype. By focusing on the difference between resistant and susceptible genotypes, and on whether the gene was rust responsive, a total of 14 genes were selected as candidate quantitative resistance genes. Small RNAs are also involved in the soybean rust interactions. Nine soybean-gene-hitting siRNA were considered as rust responsive. Especially one of them, overlapped with a MYB transcription factor gene, exhibited distinct expression patterns in resistant and susceptible genotypes after rust infection. Both the candidate quantitative resistant genes and siRNA loci have potential to increase the rust resistance and be utilized in breeding programs. Multiple evidences suggest the dominant susceptible locus in the TMG06_0011 genome is the same locus as Rpp1, or at least close to Rpp1 (Garcia, Calvo, Kiihl, & Souto, 2011). According to the gene annotation of the TMG06_0011 genome, no novel gene was found in the or close to Rpp1 locus. Therefore, the dominant susceptible is most likely caused by a deleterious version of the gene that is also present in the resistant genotypes, presumably, a deleterious version of Rpp1. In light of these observations, identifying the dominant susceptible allele will probably simultaneously identify the Rpp1 gene, and vice versa. The first hypothesis was that the dominant susceptible phenotype was due to RNA silencing that an antisense RNA was transcribed from a gene overlapping with Rpp1, or transcriptional read-through of a divergently facing gene, resulting in the formation of double stranded RNA and then small RNA silencing of Rpp1. However, the data from multiple small RNA and mRNA sequencing experiments did not support this hypothesis. Another small region was identified to become more promising as harboring the candidate Rpp1 gene, and is the same region suggested by the fine mapping of Kim et al (Kim et al., 2012). The reference genome suggests that this region contains three NBS-LRR genes, however, in the gene annotation and RNA sequencing of TMG06_0011, and in our RNA sequencing of the two resistant PIs, this region codes for only one gene, an NBS-LRR gene in all three genotypes. Interestingly, the alignment of the amino acid sequences of the gene from three genotypes showed consistently that TMG06_0011 misses 14 consecutive glutamic acids in the NBS domain area. Missing a string of glutamic acids could result in the disruption of protein-protein binding (such as a Rpp1 dimer or Rpp1 plus another unknown protein), or it could also lead to a non-functioning protein if this region is needed for function, such as in binding of calcium or other positively charged molecules. Also, the expression of this gene is consistently higher in the dominant susceptible genotype compared to both resistance genotypes. Above all, the gene becomes the best candidate of Rpp1 gene. The observations fit the hypothesis that Rpp1 works as a homo-dimer, and that the dominant susceptible is providing abundant defective Rpp1 protein that de-activate the functional Rpp1 when dimer together. Western blots would be necessary to confirm the transcriptional data. The excessing defective Rpp1 protein would also minimize the numbers of functional Rpp1 protein that could self-associate in the heterozygous plants, giving an approximate 3:1 phenotype, susceptible to resistant in the F2 population.