Update:
Objective 1: Identify pathogens found in foliar diseased plant samples from the field in year 1.

In year 1 of this project, we extracted RNA from 49 diseased soybean tissue samples that were collected during the 2017 growing season. These samples comprised leaves and in some cases stems from diseased plants found growing in farmer and ISU fields from late summer through fall of 2017. A total of 49 samples were collected from different IA counties and stored at minus 80 degrees Celsius. The samples represented a wide spectrum of different symptoms that included diseases that could be tentatively identified and those that could not be identified. Diseases that were tentatively identified were: Phyllosticta, midge gall, tan spot, Septoria brown spot, SVNV, downy mildew, Cercospora leaf blight, and frog eye leaf spot. Unknown diseases had an array of symptoms including: virus-like symptoms, different sizes and shapes of necrotic spots (some surrounded by yellow tissue), top die back, necrosis with an orange/yellow halo, and bud proliferation.

Following extraction of RNA from these samples, we made libraries for sequencing, and then performed the RNA sequencing using the Illumina HiSeq 3000 at The ISU DNA Facility. A total of approximately 769 million paired-end (PE, 150 base + 150 base) reads per lane was generated. After quality control checks, the paired-end reads were trimmed to remove bad quality reads. Trimmed reads were mapped and filtered against soybean genome using a mapper. The resulting reads that did not map to the soybean genome were considered to be non-soybean and possibly from pathogens. These unmapped, non-soybean reads were then mapped against known viral and viroid genomes (30,640 viral sequences) to identify viral reads. We then used the viral reads in a de novo assembly to assemble larger sequences known as contigs that represented full or partial viral genomes. The assembled contigs were identified for viruses using BLASTn searches followed by BLASTx searches of the GenBank non-redundant database. Using this pipeline, we have identified several known and unknown viruses that were present in Iowa soybean fields during the 2017 growing season. Of particular interest is Clover yellow vein virus (ClYVV) that was identified in Washington County, which we had also found for the first time in Story County, Iowa in the 2016 growing season (see File 1 for images of symptoms). Other viruses that were identified during the 2017 growing season in IA county were: Alfalfa mosaic virus (AMV) in Hancock county, Bean pod mottle virus (BPMV), Soybean dwarf virus (SbDV), Soybean vein necrosis virus (SVNV), Tobacco ringspot virus (TRSV) in Hancock; Story county, and Tobacco streak virus (TSV) in Story county. We validated the presence of ClYVV and TRSV by performing infection assays on Nicotiana benthamiana plants. Infected plants for both ClYVV and TRSV showed severe viral symptoms 10 days post inoculation. Infected leaves tested positive for ClYVV and TRSV using the Potyvirus ImmunoStrip and TRSV ImmunoStrip tests from Agdia, respectively.

We have performed multiple attempts to infect soybean with the ClYVV isolate that we obtained from 2017. Curiously, we have not been able to infect Williams 82 or any of the 41 parents of the soybean nested association mapping (NAM) lines. This suggests to us that there may be specific soybean genotypes that are especially susceptible to ClYVV, and this will be useful to determine, so that these may be avoided in the future. We are in the process of trying to identify and obtain seeds from soybean varieties that were infected by ClYVV in 2016 and 2017. We have initiated efforts to obtain images of the ClYVV virus particles (virions) by electron micrcoscopy.

The assembled genome sequences for the viruses that were found have been submitted to the GenBank database maintained by the National Center for Biotechnology Information, and we are in the process of submitting the raw sequencing data to GenBank as well. The data will become available upon publication of our initial results. Archiving these data in GenBank facilitates broad access and dissemination of our results to soybean researchers and the broader research community.

As described above, we have developed an efficient process by which to identify viruses from RNA samples derived from soybean fields. We are continuing to work on bioinformatics strategies to efficiently identify unknown fungal and bacterial pathogens. This is a much more difficult problem, because fungal and bacterial genomes are larger than viral genomes by orders of magnitude, meaning that the computational approaches require much more time and produce more complex results that require equally more complicated approaches to interpret.

During the 2018 growing season, we collected 43 samples corresponding to diseased plants found in soybean fields from across the state of Iowa. None of these diseased samples could be diagnosed as being caused by a specific pathogen. However, many appeared to have virus-like symptoms. Based on our experience in 2017, we improved our sample collection protocol by getting samples from the field or Plant Disease and Diagnostic Clinic more rapidly and taking steps to better preserve the samples for later processing. We expect that this will help us to obtain better quality RNA from the samples, which will improve the quality of the sequencing analyses.

View uploaded report

We completed bioinformatic analysis of samples collected from 2017. We confirmed identification of a soybean virus that is new to the state of Iowa, Clover yellow vein virus, and we validated this virus as being viable by using the tissue that we had sequenced to inoculate plants. We have established an efficient bioinformatics pipleline for identifying viruses from diseased samples using the RNA sequencing approach. We continue to work on the bioinformatics pipeline to identify fungal and bacterial pathogens.

We completed collection of samples from the 2018 growing season. A total of 42 samples representing a variety of different foliar symptoms that could not be diagnosed were collected. Most of these symptoms we would categorize as virus-like. These samples were collected and saved using an improved collection procedure.