Location: Molecular Plant Pathology Laboratory
2021 Annual Report
Objectives
Objective 1: Evaluate patterns of gene expression in mixed infections between CTV and HLB in citrus, with the goal of discovering both cross-protecting strains and potential synergistic interactions.
Objective 2: Develop new diagnostic reagents for emerging citrus pathogens and evaluate them as research reagents for the citrus community.
Approach
Host gene expression with two phloem limited pathogens - We intend to define citrus genes expressed in response to infection by phloem limited pathogens Citrus tristeza virus (CTV) and 'Ca. Liberibacter asiaticus' (CaLas). RNA Seq: Asymptomatic, young leaf tissue will be ground to a powder in liquid nitrogen and RNA will be isolated using a Trizol protocol. RNA will be used for RNA sequencing (Illumina 2500). Paired-end reads will be mapped to Citrus sinensis ‘Valencia’ reference genome. Differentially expressed transcripts will be identified. P < 0.01 and log2 fold change (log2FC) =¦1¦will be set as cut-off values.
Curation of Exotic Pathogens of Citrus Collection: Each CTV isolate is composed of different genotypes, and we do not have detailed information on the CTV strain composition within each isolate. We will obtain this information by extracting RNA, preparing cDNA and performing multiplex PCR to identify mixed genotypes to characterize the CTV isolates. We will then extract dsRNA and use it as template for single read RNASeq and CTV genome assembly of CTV isolates of interest. ‘Ca. Liberibacter asiaticus’: We will determine the genome sequence of the 14 strains of ‘CaLas’ present in the EPPC. The parasitic plant dodder, Cuscuta indecora, is itself parasitized by ‘Ca. Liberibacter asiaticus’. The original genomic sequence data for ‘Ca. Liberibacter asiaticus’ was obtained from a single super-infected psyllid that provided a high ratio of CaLas to psyllid DNA required for shotgun sequencing. We can’t use psyllids under the conditions of our permit from USDA APHIS. We have shown that 2 cm segments of dodder stem infected with CaLas vary greatly in the concentration of CaLas, reaching concentrations of 109/g. We will infest citrus inoculated with CaLas from our collection with dodder and allow it to establish. We will harvest the dodder, cut them in 2 cm segments, and extract DNA. The extracts will be tested for the CaLas using our standard assay. We expect to find segments which have very high concentrations of CaLas (Cq<16). These will be sent to Jianchi Chen at Parlier for sequencing following amplification using his established protocol.
We have recombinant antibodies that recognize surface antigens of CaLas. We have MTRAs with USDA APHIS CPHST and PathSensors, Inc (Baltimore, MD) to use our antibodies to develop a ‘CANARY’ assay for ‘Ca. Liberibacter asiaticus’. CANARY enables serologically based detection of pathogens in three minutes starting from an environmental sample. Direct tissue blot immunoassay (DTBIA) – The DTBIA is a well established technique for localizing proteins in plant tissue. We have used a rabbit polyclonal antibody for DTBIA of CaLas. We therefore expressed and purified the same antigens used to generate scFv to immunize New Zealand white rabbits to produce conventional polyclonal antisera. The DTBIA format preserves the localized concentration of CaLas observed in phloem cells and works in leaf midribs as well as in fruit petioles and peduncles, seed, stem and root tissues. These new antibodies will be used in DTBIA and the results compared with those obtained with anti-OmpA antibodies.
Progress Report
The goal of this project,’Invasive Citrus Pathogens’ is to prevent the introduction or spread within the citrus industry of a number of graft-transmissible and invasive pathogens of citrus. Due to quarantine considerations, this work is carried out at Beltsville, Maryland.
The Molecular Plant Pathology Laboratory (MPPL) has continued their study of how several of the diseases of interest affect the expression of genes in infected citrus and has identified families of genes that are differentially expressed in sweet orange trees infected by Citrus tristeza virus and ‘Ca. Liberibacter asiaticus’. We have moved on from assays with trees infected by pathogens one by one and continued with ‘deep sequencing’ of RNA in trees simultaneously infected with Citrus tristeza virus and ‘Ca. Liberibacter asiaticus. We have analyzed the data to give a more precise understanding of how gene expression varies in trees infected by these two important citrus pathogens. Gene expression analysis of the doubly-infected plants will provide information on the specificity of gene expression in response to different pathogens and is also reflective of real-world conditions. Because these genes are plant genes, expression is likely to be more uniform than is the distribution of the pathogens themselves, and the resulting assays to reveal the pattern of plant gene expression in response to the pathogen may be more reliable. The data have been analyzed, and the manuscripts have been published. We have also documented a beneficial protective effect of prior inoculation with a mild strain of CTV against subsequent infection by ‘Ca. Liberibacter asiaticus’. A manuscript on this topic has been submitted for publication.
We have also worked with an antibody against the critically important pathogen that causes citrus greening disease, ‘Ca. Liberibacter asiaticus’. We have raised and purified this antibody from rabbits and have used them in ‘tissue print’ assays to reveal the presence of the pathogen in infected citrus tissues. This has never been done before with this pathogen and allows us to both detect the pathogen and study its distribution in infected citrus trees. Several manuscripts using the ‘tissue print’ method have been published, including a manuscript in which the tissue print method is combined with PCR- and qPCR based methods.
In collaboration with ARS scientists in Florida, a canine olfactory detection method was shown to discriminate between citrus plants infected with the pathogen causing citrus greening, CLas, and other citrus pathogens and Liberibacter species in a 'blind' field test of the Exotic Citrus pathogen collection in Beltsville. The manuscript describing the detection method was published by the ARS scientists in Florida.
MPPL has also worked on projects to identify novel pathogens of citrus and establish the causes of diseases of citrus. In cooperation with researchers at the University of Florida and at Ft. Detrick, Maryland, we have characterized a group of novel pararetroviruses that are associated with citrus blight disease. Citrus blight has been a serious problem for Florida citrus growers since at least the 1880s, but the cause has always eluded researchers. We have developed evidence that an endogenous pararetrovirus is associated with the disease. This group of viruses may be integrated into the genome of citrus and, under certain environmental or stress conditions, may become active and cause disease. We continue to work on this important disease problem. We are also working with RNA library sequencing methodology to characterize gene expression in trees affected by citrus blight.
MPPL also gained important insights into the dissemination of the destructive huanglongbing citrus disease in China. Huanglongbing (HLB) is the most serious disease of citrus in the world today. HLB, also known as citrus greening, originated in China and has devastated the citrus industry in Florida since its introduction there in 2005 and currently threatens the California citrus industry. ARS scientists carried out an extensive review of the literature, and unpublished records in China related to the spread of HLB in China from its original description in Fujian province in 1919 to the present and performed molecular assays to support the literature review. The records and data show that although the disease had been known for at least 30 years by the time it was described in 1919, it had not spread widely from its possible sites of origin, the Guangdong, Yunnan, and Sichuan provinces.
Cilevirus and Dichorhavirus are two genera of plant viruses infecting citrus and cause leprosis disease. Hibiscus-infecting Cilevirus (HiCV), which is considered as a hibiscus strain of cytoplasmic Citrus leprosis virus 2 (CiLV-C2H), was detected in Hibiscus in Hawaii and Florida. A Ribozero High throughput sequencing protocol was developed and applied to identify CiLV-C2H in passion fruit (Passiflora sp.) in Hawaii. The full genome sequence was determined. The dichorhaviruses that infect citrus in Mexico, Colombia, and South Africa were identified as citrus and orchid strains of OFV (OFV-Cit and OFV-Orc). In 2020, we detected Brevipalpus transmitted dichorhaviruses in multiple hosts: (1) OFV-Orc2 strain, infecting rough lemon and mandarin orange in Hawaii, (2) OFV-Orc1 and OFV-Orc2 in monkey grass (Liriope spp.), greenbrier (Smilax auriculata), lilyturfs (Ophiopogon spp.) and cast-iron (Aspidistra elatior) plants in Florida and (3) OFV-Orc1 and OFV-Orc2 in cymbidium and dendrobium orchids in California.
Accomplishments
1. Developed tissue printing and machine learning technology for detection of the citrus greening pathogen and identified ten viruses associated with the citrus leprosis disease complex. Since 2005, Citrus greening or Huanglongbing (HLB) disease has caused unprecedented losses to the citrus industry in Florida and continuing the threat to California and Texas citrus industry. In Florida, Sweet orange and grapefruit production has been reduced by 80 and 96%, respectively. Detection of CLas for diagnosis of HLB is generally done by qPCR assay. An immune tissue print assay for CLas was developed to provide an alternative detection paradigm to qPCR. The antibodies were used in a tissue print format to visually detect CLas as purple spots in the phloem cells of infected citrus. Usually, this straightforward detection assay may provide a misleading qualitative yes/no decision. To avoid this misleading interpretation machine learning based approach was integrated to do the image analysis. A supervised representation learning based approach was developed to train convolutional neural networks (CNNs) and implemented to evaluate the images. Analysis provides either positive or negative for the presence of CLas confidence scores associated with the determination of a given image. Current model hypothesis worked very well, with 90% of curated samples identified in agreement with the expert curator. Another advantage of using the models to classify the tissue prints in an automated format is that human observer bias is eliminated, and consistency of classification has been improved. The computational models can be developed for classification of immune tissue prints for the detection of other phloem-limited pathogens like Spiroplasma citri, Citrus tristeza virus and Citrus chlorotic dwarf associated virus infecting citrus.
