Banner graphic that reads, 'Introducing H. H. Flor Distinguished Reviews. A landmark series from M P M I advancing the science of molecular plant-microbe interactions'
APS Online Publications
Skip main navigation
Previous
Next
RESEARCHFree Access icon

Mitochondrial Loci Enable Specific Quantitative Real-Time PCR Detection of the Pathogen Causing Contemporary Impatiens Downy Mildew Epidemics

    Affiliations
    Authors and Affiliations
    • Nicholas LeBlanc1 2
    • Frank Martin3
    • Vanina Castroagudín1 2
    • Jo Anne Crouch1 †
    1. 1United States Department of Agriculture, Agricultural Research Service (USDA-ARS), Mycology and Nematology Genetic Diversity and Biology Laboratory, Beltsville, MD 20705
    2. 2Oak Ridge Institute for Science and Education, Agricultural Research Service’s Research Participation Program, Oak Ridge, TN 37830
    3. 3USDA-ARS, Crop Improvement and Protection Research Unit, Salinas, CA 93905
    Published Online:21 Jan 2022https://doi.org/10.1094/PDIS-05-21-0933-RE

    Abstract

    Impatiens downy mildew (IDM) disease is a primary constraint on the production of Impatiens walleriana, a popular and economically important floriculture plant. IDM is caused by the biotrophic. oomycete Plasmopara destructor that emerged as a pathogen of I. walleriana in the 2000s. To enable P. destructor detection and quantification, a hydrolysis-probe-based quantitative PCR diagnostic assay was developed based on unique orientation and order of the mitochondrial cytochrome c oxidase subunit1 (cox1) and ATP synthase subunit alpha (atp1) genes in the genus Plasmopara. Nucleotide sequences and analysis of the cox1/atp1 region distinguished P. destructor and its sister-species P. obducens, consistent with prior phylogenetic analyses using cox2 and rDNA markers. Specificity for P. destructor was incorporated into a hydrolysis probe targeting the cox1 gene and flanking primers that amplified across the cox1/atp1 intergenic region. The limit of detection was 0.5 fg/μl of P. destructor DNA (∼100 plasmid copies/μl), with amplification efficiency = 0.95. The assay was validated against a panel of target and nontarget oomycetes, which showed that the primers were specific for Plasmopara spp., while the probe was specific for P. destructor infecting both I. walleriana and I. balsamina. Testing of Impatiens tissue collected from 23 locations across 13 states indicated all samples with IDM symptoms tested positive for P. destructor. Asymptomatic plants from two locations also tested positive for P. destructor.

    Impatiens walleriana (common names: impatiens, busy lizzy) are widely cultivated herbaceous ornamental plants, valued for their vibrant flowers and used as bedding plants in shady locations (Morgan 2007). Since the beginning of the 21st century, impatiens downy mildew (IDM) disease has emerged as a primary constraint on the production and sale of these plants (Daughtrey 2017; Daughtrey and Palmer 2014; Lane et al. 2005; Wegulo et al. 2004). Coinciding with the emergence of IDM there has been a significant decrease in the economic value of impatiens produced in the U.S. nursery and floriculture industry (USDA NASS 1998, 2009, 2014). With limited understanding of disease epidemiology or pathogen biology, efforts to control IDM have been limited to costly fungicide application. However, as shown by Harlan et al. (2017), managing the abiotic environment in combination with fungicide application may also help reduce disease. In 2020, IDM-resistant cultivars of I. walleriana have been released for greenhouse production, and are now undergoing widespread release for the 2020 annual bedding plant season (Daughtrey et al. 2020).

    Downy mildew of impatiens has been known since the 1880s, when Plasmopara obducens was described from wild I. noli-tangere seedlings growing as understory plants in German forests (Schröeter 1877). Since then, IDM has been reported from at least 12 species of Impatiens, with the majority of reports occurring only after the emergence of the disease on cultivated I. walleriana in the 2000s (Farr and Rossman 2021). Molecular analysis of Plasmopara infecting Impatiens showed that collections of Plasmopara spp. made before the onset of disease epidemics in the 21st century on noncultivated impatiens like I. noli-tangere are genetically distinct from those collected during contemporary epidemics on cultivated I. walleriana and I. balsamina (Görg et al. 2017; Salgado-Salazar et al. 2018b). The species infecting I. noli-tangere has retained the name P. obducens whereas the primary species responsible for modern IDM outbreaks on I. walleriana is now named Peronospora destructor (Görg et al. 2017; Salgado-Salazar et al. 2018b). Consequently, P. destructor represents the greatest level of risk for production of I. walleriana in the horticulture industry.

    Molecular diagnostic assays for downy mildew pathogens traditionally target high-copy-number, low-resolution rRNA loci in the nuclear genome (reviewed in Crandall et al. 2018). However, from the limitations of these loci for species-level pathogen identification, there has been an increased effort to identify novel high-resolution loci or additional genomic polymorphisms that can improve molecular diagnostics of downy mildew and closely related oomycete plant pathogens. Examples from 2016, 2017, and 2019 include comparison of mitochondrial and nuclear sequences to identify loci that are found only in the genome of the target pathogen (Kunjeti et al. 2016; Rahman et al. 2017, 2019). Another successful approach assesses variation in gene order or presence of unique putative open reading frames for targeted development of diagnostic assays (F. Martin, unpublished), which is used to design genus- and species-specific primers and probes (Bilodeau et al. 2014; Crandall et al. 2021; Kunjeti et al. 2016; Miles et al. 2017). These newer next-generation sequencing and genomic approaches show strong potential for improving molecular diagnostics of downy mildew pathogens.

    Despite the economic costs associated with IDM, options for detecting P. destructor infections are limited, particularly in the absence of symptoms or physical signs of the pathogen. The most straightforward method for IDM diagnosis is visual observation of disease symptoms, which includes leaf curling and chlorosis and/or the observation of pathogen sporulation on the abaxial leaf surface. This approach can be used to identify visibly diseased plants, but has limited application for detecting or quantifying the pathogen in asymptomatic plants or the environment. The only alternative tool that has been developed is a fluorescence in situ hybridization assay that enables species-specific visualization of P. destructor in planta (Salgado-Salazar et al. 2018a). This assay has clear potential for understanding the infection process of the biotrophic pathogen P. destructor, but is not likely to be useful for routine pathogen detection or quantification in a large number of samples or asymptomatic plants. Therefore, the goal of this work was to develop a molecular diagnostic assay capable of detecting and quantifying P. destructor responsible for ongoing outbreaks of IDM. To reach this goal, this work describes: assessment of variation in mitochondrial gene order and orientation within the target genus Plasmopara and nontarget oomycetes; the design and validation of a hydrolysis probe and primers for real-time quantitative PCR (qPCR)-based detection of P. destructor; and application of the resultant real-time qPCR assay for diagnostic testing of plant samples collected from residential landscapes and regional IDM sentinel sites.

    Materials and Methods

    Diagnostic marker identification and assay design.

    As part of an ongoing research project in the Martin lab (Crop Improvement and Protection Research, Agricultural Research Service, U.S. Department of Agriculture, Salinas, CA), mitochondrial genomes representing a wide range of oomycete taxa representing 158 species of 21 genera have been assembled from Illumina sequence data, including P. destructor PA1-1 and H12-14-11, and P. halstedii RDM20 (GCA_003640625.1, GCA_003640485.1, and GCA_003640505.1, respectively; F. N. Martin, unpublished). Comparative analysis of assembled genomes (including 26 species representing nine genera of downy mildews) identified the gene order of cytochrome c oxidase subunit 1 (cox1) and ATP synthase subunit alpha (atp1; hereafter referred to as cox1/atp1 locus) as unique for the genus Plasmopara, and was targeted for development of a species-specific diagnostic assay.

    To evaluate intraspecific genetic diversity at the cox1/atp1 locus, Sanger sequence data were generated from P. destructor and P. obducens samples. The samples included one representative of P. obducens collected from I. noli-tangere (Salgado-Salazar et al. 2018b) and six samples of P. destructor, all identified using 18 polymorphic simple sequence repeat loci, cox2, and rDNA markers (Salgado-Salazar et al. 2015; unpublished data). The cox1/atp1 locus was also sequenced from two samples of P. halstedii to serve as outgroups within the genus Plasmopara. PCR amplicons were generated, cloned, and sequenced with the Plas-F4/Plas-R3 primers (Table 1) designed to target the cox1/atp1 locus using protocols described in Salgado-Salazar et al. (2018b). Sequences are available from NCBI-GenBank under accession numbers MW275297 to MW275305. In addition to the mitochondrial genome data used in the initial identification of the diagnostic locus (see above), sequence data for the cox1 and atp1 genes were obtained from the following published mitochondrial genomes in NCBI GenBank: Hyaloperonospora arabidopsidis Emoy2 (GCA_000173235.2), Phytophthora infestans T30-4 (GCA_000142945.1), Plasmopara halstedii pathotype 710 (GCA_003724065.1), Podarcis muralis INRA-PM001 (GCA_003676415.1), Pseudoperonospora cubensis MSU-1 (GCA_000252605.1), Pythium ultimum DAOM:BR114 (NC_014280), and Sclerospora graminicola UoMm-SG-Pathotype1 (GCA_001887855.2). Data from the plant Vaccinium macrocarpon (GCA_000775335.2) were also included to ensure the assay did not generate false-positive results from plant DNA coextracted from downy mildew samples. Sequences were aligned using the program MUSCLE (Edgar 2004) and visualized using the program AliView v1.18.1 (Larsson 2014) with coding sequences from the 3′ cox1 and 5′ atp1. This alignment was used to design an additional reverse primer and internal hydrolysis probe, specific for P. destructor sequences. The reverse primer was designed to pair with the Plas-F4 forward primer and increase specificity for P. destructor, while amplifying across the cox1/atp1 intergenic region. A 40-nt hydrolysis probe was designed to target the 3′ region of the cox1 gene that was invariant across P. destructor samples and variant at multiple positions in nontarget organisms. The probe was synthesized to include a 5′ 6-FAM reporter dye, a ZEN internal quencher (Integrated DNA Technologies, Coralville, IA), and a 3′ Iowa Black FQ quencher (Integrated DNA Technologies).

    Table 1. Primers and probes used for DNA sequencing and quantitative PCR assay

    Sample collections, DNA extraction, and control synthesis.

    Fresh I. walleriana and I. balsamina samples were collected from 23 US sites. Twelve sites were residential landscapes where symptomatic Impatiens spp. were grown and one site represented a sampling of the I. walleriana ‘Imara’ (Syngenta Flowers, Gilroy, CA) that were symptomatic for IDM in a commercial greenhouse in California (Table 2). The remaining 10 sites were sentinel sites established in botanical and residential gardens in the Mid-Atlantic and Great Lakes regions (Supplementary Fig. S1). Sites were selected by contacting volunteer participants in the Sentinel Plant Network (https://www.publicgardens.org/programs/sentinel-plant-network/about-spn) and coordinating with collaborators in the Mid-Atlantic region. It is unknown if P. destructor was present at individual sites before establishment. Each sentinel site was established using four healthy I. walleriana ‘Super Elfin Violet’ plants (Ball Horticultural Company, West Chicago, IL) originating from a single nursery in Baltimore, MD. Plants were shipped overnight to individual locations and after installation were monitored on a biweekly basis. Plants were also monitored for IDM under growth chamber conditions at the U.S. Department of Agriculture’s Agricultural Research Service (USDA-ARS) laboratory in Beltsville, MD and remained healthy through the course of their life-cycle. When symptoms developed on sentinel plants, plant tissue was shipped overnight to the USDA-ARS Beltsville laboratory for processing. If plants remained healthy through the growing season at the sentinel sites, at the end of the growing season asymptomatic plant material was sent to the USDA-ARS Beltsville laboratory for diagnostic testing. Upon receipt of all plant material, individual leaves were frozen in microfuge tubes at −80°C until DNA extraction. Genomic DNA was extracted from individual leaves using the OmniPrep for Fungi (G-Biosciences, St. Louis, MO) extraction kit. Two leaves from different plants at a given location were processed. Extracted DNA was quantified using a Qubit 2.0 fluorometer (Invitrogen, Carlsbad, CA) and diluted to 10 ng/μl before use in qPCR reactions, as described below.

    Table 2. Sample summary and quantitative PCR (qPCR) assay results using a Roche LightCycler 480 II

    DNA was extracted as described in Salgado-Salazar et al. (2020) from freshly collected field samples of symptomatic and asymptomatic Impatiens leaves and from symptomatic Cucurbita sp. and Ocimum sp. infected with P. cubensis and Peronospora belbahrii, respectively. In addition, DNA was extracted from pathogen-free Impatiens plant material obtained from Ball Horticulture to confirm host DNA would not generate false-positive signals. Plant stems, roots, or seeds were not included in the DNA extractions or assay validation. DNA from cultures of Phytopthora ramorum and Pythium splendens was extracted as described in Miles et al. (2017). DNA controls were constructed by preparing plasmid DNA containing the targeted region of the cox1-atp1 locus from P. destructor and total DNA from nontarget oomycetes (Table 2). Control plasmids from PCR amplicon generated from the Plas-F4/Plas-R3 primer were cloned into pCR 2.1-TOPO (Thermo Fisher Scientific, Waltham, MA) as described in Salgado-Salazar et al. (2018b), with Sanger sequencing to confirm the presence of the insert. The presence of intact mitochondrial DNA in extracted genomic DNA in pathogen and plant control samples was confirmed to ensure negative results from the assay were not a result of poor-quality mitochondrial DNA. This was determined by amplification of the cox1 gene with universal oomycete primers OomCox1-levup and OomCox1-levlo (Robideau et al. 2011) or universal plant primers FMP12b and FMP13b using PCR cycle parameters as described in Martin and Tooley (2004). To verify amplification, reactions were resolved using the QIAxcel capillary electrophoresis system (QIAGEN, Germantown, MD).

    Assay optimization.

    qPCR reactions were optimized using a Roche LightCycler 480 II (Roche, IA). Individual 20-μl reactions were prepared in master mixes, with each individual reaction including the following: 10 μl of Roche LightCycler Probes Master Mix, 0.625 μl each of the forward and reverse primers (10 μM), 0.2 μl of the probe (20 μM), 6.55 μl of PCR-grade water, and 2 μl of DNA template. Reactions were run with the following protocol: 95°C (10 min), 40 cycles of 95°C (10 s), annealing temperature (1 min 30 s), and 72°C (1 s), followed by 40°C (10 s). DNA concentration for plasmid and genomic DNA controls was 10 ng/μl. Three annealing temperatures (58, 59, and 60°C) were tested to optimize the assay. Data were analyzed using the absolute quantification/fit points method included in the LightCycler 480 software (v.1.5.0; Roche). Assay limit of detection was tested using a serial dilution of plasmid DNA ranging from 5 ng/μl to 5 × 10−7 ng/μl, with three replicates per dilution. Assay slope and efficiency was calculated from data within the limit of detection using the same serial dilution. Copy number was calculated based on plasmid template size of 3,931 bp using the program NEBioCalculator (v1.19.1, New England Biolabs; https://nebiocalculator.neb.com/#!/dsdnaamt). To verify assay results were reproducible across equipment types, standard curve dilutions were also amplified with the Mic qPCR Cycler (Bio Molecular Systems, Queensland, Australia) and used to calculate assay efficiencies as described above.

    Results

    Diagnostic assay design and optimization.

    A hydrolysis probe was designed to anneal to the 3′ end of the cox1 gene and incorporated at least five SNPs that differentiated it from nontarget taxa (Fig. 1; Table 1). The probe also contained two SNPs that differentiated P. destructor from a sample of P. obducens originating from I. noli-tangere. Using the optimal annealing temperature of 60°C, the assay showed reproducible detection from 5 ng/μl to 5 × 10−7 ng/μl (124 plasmid copies/μl). Lower concentrations produced inconsistent results across technical replicates. Based on the serial dilution, the assay had an amplification efficiency of 0.95 and R2 of 0.995. Analysis of standard curve dilutions using the Mic qPCR Cycler showed similar results with an efficiency of 0.947 and R2 = 0.999, indicating the assay gave reproducible results across instrument types (Fig. 2). qPCR detection of pathogen-free Impatiens tissue, nontarget taxa, or plant DNA associated with other downy mildew samples was not observed, confirming the specificity of the assay.

    Fig. 1.

    Fig. 1. Schematic of diagnostic assay design for Plasmopara destructor using the cytochrome c oxidase subunit 1 (cox1) and ATP synthase subunit alpha (atp1) locus. Hydrolysis probe (Pod-P3) is shown at the 3′ end of the cox1 coding sequence. Forward (Plas-F4) and reverse (Pod-R9) primers that flank the probe are specific for the cox1 and atp1 coding sequences, respectively. The black bar indicates an intergenic region of ∼200 bp between the two genes found only in Plasmopara spp. The sequence alignment below the probe sequence shows variation across the probe site. Sequences from P. destructor samples are highlighted in bold text. See Table 2 for additional sample meta-data.

    Download as PowerPoint
    Fig. 2.

    Fig. 2. Plot of crossing points (Cp) for serial dilution of Plasmopara destructor plasmid DNA. Serial dilution ranged from 5 ng/μl to 5 × 10−7 ng/μl and each dilution has three technical replicates. Top plot shows results from a Roche LightCycler 480 II; bottom plot shows results from an Mic qPCR Cycler.

    Download as PowerPoint

    Visualization of qPCR reactions using capillary electrophoresis showed that only samples from Plasmopara spp. produced amplicons in the reactions (Fig. 3). In addition, distinct interspecific size polymorphisms between amplicons generated from P. destructor and P. halstedii were observed. Amplicons from the two P. halstedii samples were smaller (404 and 397 bp) than amplicons from samples of P. destructor (483 and 464 bp). Amplicons from P. obducens BPI871273 were larger (483 bp) than P. destructor 096-3A (464 bp). These results indicated that the primers were specific for Plasmopara species included in this study, and the hydrolysis probe was specific for P. destructor and did not detect P. obducens.

    Fig. 3.

    Fig. 3. Plasmopara-specific end-point PCR amplification across the intergenic region of the cytochrome c oxidase subunit 1 (cox1) and ATP synthase subunit alpha (atp1) locus. Taxonomic identities of samples are on the x-axis and amplicon size is on the y-axis. All of the Plasmopara samples produced a fragment of ∼400 bp. No amplification was observed for nontarget samples from oomycete genera commonly found in the environment (Pythium and Phytophthora) and nontarget genera of downy mildew pathogens.

    Download as PowerPoint

    Field sample collection and diagnostics.

    All field samples from residential landscapes showed symptoms of chlorosis and signs of pathogen sporulation from the leaves. Samples from eight of the 10 IDM sentinel site locations showed symptoms of chlorosis and pathogen sporulation. Plant material from the remaining two sentinel site locations was submitted at the end of the growing season without any visible symptoms of IDM or signs of the pathogen. In 2019, an additional sample was collected in CA from symptomatic ‘Imara’ impatiens, a new series of I. walleriana with reduced susceptibility to P. destructor (Daughtrey et al. 2020).

    All 46 plant samples tested positive for P. destructor by qPCR. Cycle crossing-points ranged from 18 to 36 (Table 2). Four asymptomatic plants from two locations where IDM was not reported also showed positive detection of P. destructor, as did samples of the ‘Imara’ impatiens (Table 2). Plasmid and genomic DNA from nontarget oomycetes were not detected when tested at the same time as the field samples. Overall, these results indicated the qPCR assay could detect P. destructor in symptomatic and asymptomatic I. walleriana and I. balsamina from samples obtained from the landscape, field, and production environment.

    Discussion

    The goal of this work was to develop a qPCR assay that enabled specific detection of P. destructor responsible for IDM outbreaks on I. walleriana in the ornamental industry. Initial comparison of mitochondrial genome assemblies showed the order and orientation of cox1 and atp1 were unique to the genus Plasmopara. Further sequencing of these genes in samples representing known genetic diversity in P. destructor enabled design of a hydrolysis probe to introduce specificity for P. destructor. Application of the qPCR assay to test field samples showed the widespread presence of P. destructor infecting I. walleriana, including plants that remained disease-free through the growing season and samples from released IDM-resistant plants.

    The limited number of genome sequences available as of this writing for the biotrophic organisms that cause downy mildew diseases is a primary hindrance for the development of molecular diagnostic assays. In addition, sequence data from the most widely accessible and published loci (i.e., rDNA ITS, cox2) often fail to discriminate closely related downy mildew species (Salgado-Salazar et al. 2018b, 2020; Voglmayr et al. 2014). To overcome these limitations, in this study we targeted mitochondrial genes, which often display greater genetic diversity than commonly used nuclear loci (Robideau et al. 2011). We also identified genes that showed unique order and orientation for the target genus Plasmopara. The focus on gene order and orientation was based on the rationale that PCR amplification across an intergenic region will be possible when genes are neighbors in the target genus but will be impossible when the genes are separated by a large physical distance in nontarget organisms (Bilodeau et al. 2014; Miles et al. 2017). Analysis of mitochondrial gene order in oomycetes representing 21 genera and 158 species (F. N. Martin, unpublished) revealed a unique orientation and order of the cox1 and atp1 genes in the genus Plasmopara. In Plasmopara, an ∼200-bp intergenic region separated 3′ ends of the cox1 and atp1 genes. In contrast, the cox1 and atp1 genes were separated by multiple predicted genes or displayed different orientations in mitochondrial genome assemblies from oomycetes outside this genus. Consistent with these data, we were not able to produce amplicons from nontarget oomycetes outside of the genus Plasmopara, highlighting the utility of this approach for identifying genus-specific diagnostic loci. Consistency in results when the assay was run on different qPCR thermal cyclers (LightCycler and Mic qPCR Cycler) demonstrates stability of the assay across equipment platforms.

    Multiple studies have shown Plasmopara spp. collected before the IDM epidemic on noncultivated Impatiens spp. like I. noli-tangere were P. obducens, which is genetically distinct from P. destructor, the pathogen responsible for today’s IDM epidemics on cultivated I. walleriana (Salgado-Salazar et al. 2018b). Recognizing that P. destructor is the primary threat to the ornamental industry motivated the development of a molecular diagnostic assay that can identify and quantify this species in symptomatic and asymptomatic plant material or the environment. To integrate specificity for P. destructor and account for potential intraspecific genetic diversity, we sequenced the Plasmopara-specific cox1 and atp1 target region from a panel of samples that represent known genetic diversity in the pathogen P. destructor (Salgado-Salazar et al. 2018b; unpublished data). As expected, these data showed clear differentiation between the target pathogen P. destructor and nontarget P. obducens as well as other Plasmopara spp. By identifying a region that was conserved across the samples representing P. destructor we were able to design additional primers and a probe to increase assay specificity. Application of this assay showed all samples of I. walleriana and I. balsamina from two locations tested positive for P. destructor, providing additional evidence that this species is the pathogen responsible for the emergence of IDM in the U.S. ornamental industry.

    Positive detection for P. destructor in susceptible asymptomatic plant material and symptomatic samples from released I. walleriana ‘Imara’, which have been shown to be less susceptible to IDM, suggests that this disease may continue to limit the commercial production of I. walleriana (Daughtrey et al. 2020). Presence of the pathogen in asymptomatic plants suggests that the sale and distribution of healthy-looking plants may contribute to the dispersal of IDM in the ornamental industry. Considering the infected asymptomatic plants were sampled at the end of the growing season (September), removal of these plants from the landscape at the end of the season may help reduce pathogen inoculum in the environment and limit formation of long-lived resting spores of P. destructor (Shishkoff 2019). In the likely scenario that loss of resistance is from genetic changes in pathogen populations, this diagnostic assay remains an effective tool for pathogen detection and quantification. Future studies on the epidemiology of IDM and resistance breeding would benefit from the ability to detect and quantify pathogen density in the absence of disease symptoms, as aided by this qPCR assay.

    Diagnostic markers based on mitochondrial loci have provided a sensitive approach for conducting epidemiological studies of other downy mildews. An assay targeting a unique putative open reading frame in Bremia lactucae was sensitive enough to detect a single sporangium well within the linear range of amplification sensitivity for the assay (Kunjeti et al. 2016). Using a similar approach for identifying a unique taxon-specific target, multiplexed assays for detection of Pseudoperonospora humuli and both phylogenetic clades of P. cubensis were able to consistently detect three sporangia of each pathogen (Crandall et al. 2021).

    In addition to the applications discussed above, this qPCR assay could be modified for diagnostics of other biotrophic downy mildews in the genus Plasmopara. Considering the unique gene order and high resolution of cox1-atp1 locus in sequenced Plasmopara spp., these genes may be an optimal starting place for development of diagnostics assays for this and other emerging pathogens in the genus. P. halstedii and P. viticola are important pathogens of sunflower and grape, respectively. As both of these species are composed of differentiated genetic clades (Rivera et al. 2016; Rouxel et al. 2014), the cox1-atp1 locus could be used to develop diagnostic assays for different groups and applied to understand pathogen ecology and disease epidemiology.

    Development of molecular diagnostic assays for obligate biotrophic downy mildew pathogens presents many challenges. Not only is there a limited amount of genomic data available from these organisms, many downy mildew pathogen species are species complexes, containing multiple genetically differentiated groups (Rivera et al. 2016; Rouxel et al. 2014; Salgado-Salazar et al. 2018b) that represent different levels of risk for plant health. The generation and analysis of genomic data described in this and other work (Rahman et al. 2019) highlights the potential for the use of next-generation sequencing to improve molecular diagnostics of closely related species or pathogen genotypes. Ultimately, the use of genomic data will not only improve diagnostics of emerging downy mildew pathogens, but also contribute to controlling the economically devastating diseases these pathogens cause.

    Acknowledgments

    The authors thank collaborators and volunteers that contributed to sample collection including cooperators in the Sentinel Plant Network. The Oak Ridge Institute for Science and Education, the U.S. Department of Energy, and the Oak Ridge Associated Universities did not have any role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture.

    The author(s) declare no conflict of interest.

    Literature Cited

    Current address of Nicholas LeBlanc: USDA-ARS, Crop Improvement and Protection Research Unit, Salinas, CA 93905.

    Funding: This work was supported by the Oak Ridge Institute for Science and Education under grant no. DEAC05-06OR23100, the U.S. Department of Agriculture’s Agricultural Research Service under grant no. 8042-22000-298-00-D, and the U.S. Department of Agriculture’s Animal and Plant Health Inspection Service under program no. Farm Bill 10007.

    The author(s) declare no conflict of interest.