Enhanced Reliability and Accuracy for Field Deployable Bioforensic Detection and Discrimination of Xylella fastidiosa subsp. pauca, Causal Agent of Citrus Variegated Chlorosis Using Razor Ex Technology and TaqMan Quantitative PCR

A reliable, accurate and rapid multigene-based assay combining real time quantitative PCR (qPCR) and a Razor Ex BioDetection System (Razor Ex) was validated for detection of Xylella fastidiosa subsp. pauca (Xfp, a xylem-limited bacterium that causes citrus variegated chlorosis [CVC]). CVC, which is exotic to the United States, has spread through South and Central America and could significantly impact U.S. citrus if it arrives. A method for early, accurate and sensitive detection of Xfp in plant tissues is needed by plant health officials for inspection of products from quarantined locations, and by extension specialists for detection, identification and management of disease outbreaks and reservoir hosts. Two sets of specific PCR primers and probes, targeting Xfp genes for fimbrillin and the periplasmic iron-binding protein were designed. A third pair of primers targeting the conserved cobalamin synthesis protein gene was designed to detect all possible X. fastidiosa (Xf) strains. All three primer sets detected as little as 1 fg of plasmid DNA carrying X. fastidiosa target sequences and genomic DNA of Xfp at as little as 1 - 10 fg. The use of Razor Ex facilitates a rapid (about 30 min) in-field assay capability for detection of all Xf strains, and for specific detection of Xfp. Combined use of three primer sets targeting different genes increased the assay accuracy and broadened the range of detection. To our knowledge, this is the first report of a field-deployable rapid and reliable bioforensic detection and discrimination method for a bacterial phytopathogen based on multigene targets.


Introduction
Xylella fastidiosa (Xf), a xylem limited plant pathogen, causes a large number of diseases including plum leaf scald, phony peach, pear leaf scald, alfalfa dwarf, and leaf scorch of coffee, almond, elm, sycamore, oak, maple, mulberry, and oleander, but the two most economically important are Pierce's disease of grapevines and citrus variegated chlorosis (CVC) [1,2]. Of four subspecies, only X. fastidiosa subsp. pauca (Xfp) does not occur in the United States [3]. Xfp, categorized as a select agent until 2012, causes CVC and coffee leaf scorch (CLS). In the early 1990s, the world's largest citrus producer, Brazil, endured an outbreak of CVC that caused serious crop losses. After this outbreak, Brazilian researchers determined the complete genomic sequence of the CVC strain (9a5c) of Xfp [2,4], making it the first plant pathogenic bacterium to be completely sequenced [5,6]. According to Mansfield et al. [7] the pathogen ranked eighth among the 10 most important plant pathogenic bacteria, based on scientific and/or economic importance. Leaves of Xfp-infected citrus trees develop chlorotic spots on leaves and produce small, hard and juiceless fruits that lack commercial value, probably due to blockage of delivery of water and nutrients by aggregation of the bacteria as well as by the xanthan-like gum that the bacteria produce [8]. Citrus is produced in tropical and subtropical climates where the relatively high temperature and moisture are favorable for production. These same climatic conditions are also very advantageous for xylophagous sharpshooter leafhoppers and spittlebugs, which are important vectors of Xfp [9][10][11][12][13][14][15]. Xfp is considered a threat to the citrus industry in the U. S., and the U.S. Department of Agriculture (USDA) listed it as a quarantine select agent and considered it a high consequence pathogen. Timely diagnosis of CVC in the field is a challenge since it takes twelve months to develop the symptoms after the infection [16]. In vitro culture of all strains of X. fastidiosa is labor intensive and time consuming [17]. Thus, rapid discrimination of Xfp from other X. fastidiosa strains is essential for protecting the citrus industry.
As an exotic microorganism with a high risk profile, we chose Xfp for the development of an enhanced detection method. Whether this pathogen were to be introduced naturally (weather, insect vector, birds etc.), unintentionally (trade, travel, etc.), or intentionally, rapid pathogen detection and disease diagnostic assays will be critical during the initial outbreak delimitation, as well as during follow-on implementation of management activities, when decision making will require specific, accurate and rapid identification of the pathogen.
PCR based techniques are generally more sensitive than immunological methods and have high specificity and powerful discriminatory capabilities. Real-time qPCR offers greater sensitivity and speed compared to endpoint PCR for the detection of target DNA [18,19]. In field settings, however, plant pathogen detection can be challenging, since thermocyclers have limited sample capacity and require electrical power. The use of a portable, battery-operated real-time qPCR platform for in-field molecular testing allows minimally trained operators to test plant and environmental samples in the absence of laboratory facilities and conditions normally required, including electricity, centrifuges, liquid nitrogen, water baths, incubators and hazardous chemicals. Several portable instruments developed for this purpose include the Smart Cycler (Cepheid, Sunnyvale, CA), the LightCycler (Roche Applied Science, Indianapolis, IN), the Razor Ex Biodetection System (Razor Ex; Idaho Technology Inc.), and the Bio-Seeq instrument (Smiths Detection, Edgewood, MD).
In 2002, Schaad et al. used the Smart Cycler system to detect X. fastidiosa in sap from asymptomatic grapevines in two hours [20]. The Smart Cycler also has been applied in the identification of Phytophthora ramorum, which causes sudden oak death [21], and the Aphthovirus that causes foot-andmouth disease [22]. The Razor Ex was designed originally to allow first responders and front line military operations to identify biological threat organisms on-site. The Razor Ex system offers ready-to-use, freeze-dried reagent pouches, barcode-based PCR cycling program upload and Bluetooth capabilities for wireless data transmission. Due to fast cycling parameters, Razor Ex takes about only 30-40 minutes compared to a traditional PCR using the ABI 7300/7500 thermocycler (Applied Biosystems, Foster City, CA) that takes about 100 minutes. A Razor Ex based method also detected influenza A viruses near the patient's location and with sensitivity and specificity similar to those of the ABI 7300 [23][24][25].
Developing an assay for a select agent presents further challenges. We here report the development of such an assay for a pathogen that was on the select agent list during the course of development of the assay and should thus serve as a model for developing such future field detection procedures for regulated organisms. Specifically, field deployable, rapid TaqMan qPCR and Razor Ex protocols for reliable, sensitive, and accurate detection of X. fastidiosa and Xfp based on three discriminatory genome segments. This detection system will enhance investigative capability for ecological, agriculture and/ or biosecurity and microbial forensics.

Ethics statement
All samples included in the exclusivity and inclusivity panels of this research were used with permission from concerned persons, scientists and diagnosticians who provided these samples. This research did not involve endangered or protected species.

Sources of inclusivity and exclusivity panels
Infected plants from which genomic DNA was extracted for use in the inclusivity and/or exclusivity panels are shown in Table 1. Microbes included in the exclusivity panel are presented in Table 2

DNA isolation from plants and microbes
Genomic DNAs of X. fastidiosa and infected plant/insect samples were obtained from the American Type Culture Collection (Manassas, VA) and from governmental and university laboratories (Table 1). Genomic DNA from plants was extracted using a DNeasy Plant Mini Kit (Qiagen, Valencia, CA) and bacterial DNA ( Table 2) was extracted using the Qiagen DNeasy Blood & Tissue Kit (Qiagen) following the manufacturer's instructions. Crude DNA from sharpshooters (Homalodisca vitripennis) was isolated using prepGEM™ (ZyGEM Corporation Ltd, Hamilton, New Zealand) following the manufacturer's protocol. The concentrations of total genomic DNAs were determined using a NanoDrop v.2000 spectrophotometer (Thermo Fisher Scientific Inc., Worcester, MA). For field application, isolation of X. fastidiosa DNA from infected plant samples (grape and oak; two samples each) to be tested with the Razor Ex was done using the Dynabeadsbased modified method developed for the fungus P. omnivora [25] using Dynabeads DNA Direct Universal Kit (Invitrogen, Carlsbad, CA). Briefly, 10 to 30 mg infected foliar tissues were macerated in 100 to 150 µl Tris-EDTA (TE) buffer (Promega, Madison, WI). The mixture of 40 µl of macerated supernatant and 200 µl of Dynabeads was incubated for 5 min at RT. Tubes containing this mix were placed in a magnetic rack to retain the beads while the liquid was discarded. The beads were rinsed twice with wash buffer. Manufacturer provided suspension buffer was added to suspend DNA.

Primer and probe design
The genes for fimbrillin, periplasmic iron-binding protein, and cobalamin were targeted. The first two were used for specific detection of Xfp and the third for specific detection of all strains of X. fastidiosa. Three optimal primer and probe sets were designed following the parameters described by Arif and Ochoa-Corona [26] as shown in Table 3. The complete genome of Xfp 9a5c (accession number AE003849), retrieved from the NCBI GenBank database (http:// www.ncbi.nlm.nih.gov/), was used to subtract the Xfp specific sequences using MUMmer software [27]. Two primer and probe sets, Xf.CVC.fim and Xf.CVC.pib, specific to Xfp were designed using Primer3 [28]. The cobalamin synthesis protein gene sequence (accession number CP002165) was retrieved from the GenBank database and a primer set, Xf.csp6, was designed from it to detect all strains of X. fastidiosa. The primer pair Xf.csp.6F/Xf.csp.6R and the Xf.csp.6 probe were aligned with whole genome sequences of X. fastidiosa subsp. fastidiosa GB514 (accession number CP002165), X. fastidiosa M23 (accession number CP001011), X. fastidiosa M12 (accession number CP000941), X. fastidiosa Temecula1 (accession number AE009442), and Xfp 9a5c (accession number AE0003849), all available in GenBank. Primer thermodynamics, internal structures, and self-dimer formation were examined in silico with mFold [29]. The specificity was confirmed in silico by screening the primer and probe sequences with BLASTn, available from the GenBank nucleotide database [30] (Table 3). Primers and double quencher probes-linked 5' 6-carboxyfluorescein/ZEN TM /3' Iowa Black FQ (5' 6-FAM TM /ZEN TM /3' IB ® FQ) were synthesized by IDT (Integrated DNA Technologies, Inc., Coralville, IA).

PCR and qPCR amplification
Preliminary PCR assays were carried out in an Eppendorf thermal cycler (Eppendorf, Hauppauge, NY) using 20 μl reaction mixtures containing 10 μl GoTaq Green Master Mix (Promega), 1 μl of each forward and reverse primer from working stock of 5 µM, 1 μl of template DNA, and 7 μl nuclease free water. The cycling parameters consisted of 35 cycles as follows: Initial denaturation for 3 min at 94 °C followed by denaturation at 94 °C for 20 s, annealing at 56 °C for 30 s, extension 72 °C for 30 s, and final extension at 72 °C for 3 min. Plasmid DNA containing the target fragment and nuclease free water (non-template) were used as positive and negative controls, respectively, in each PCR amplification. Amplified PCR products were electrophoresed in a 1.5 % agarose gel in 1X TAE buffer, and amplicon sizes were estimated using 1kb plus ladders (Invitrogen).
The qPCR assays were performed in a Rotor-Gene 6000 thermocycler (Corbett Research, Sydney, Australia) and the results were analyzed using the Rotor-Gene 6000 series software 1.7 (Built 87). qPCR assays were carried out in 20 μl reaction mixtures containing 10 μl of Platinum qPCR SuperMix-UDG (Invitrogen), 0.8 μl from working stock of 5 µM of each forward and reverse primer, 0.8 μl probe from working stock of 5 µM, 0.3 mg per ml (0.12 μl) BSA (Invitrogen), 1 μl of template DNA, and 6.48 μl of nuclease free water. Rotor-Gene qPCR cycling conditions were: two initial holds, each for 2 min at 50°C and 95 °C, followed by 40 cycles at 95 °C for 15 s and 60°C for 1 min. A minimum of three dilutions of plasmid DNA (positive control; carrying the target gene sequence) were used to generate a standard curve and negative (non-template; water) controls were included in each round of qPCR amplification. Each reaction with each member of inclusivity and exclusivity panel was performed in three replicates (same reaction mixture in three tubes) using each primer and probe set.

Razor Ex amplification
Amplification with each primer set was carried out in 150 μl reaction mixtures containing 75 μl of Platinum® Quantitative PCR SuperMix-UDG, 6.0 μl of each forward (biotinylated) and reverse primer from working stock of 5 µM, 6.0 μl probe from working stock of 5 µM, 4 μl (infected plant DNA) or 1 μl (pathogen genomic DNA) of template and nuclease free water to make up the volume. Positive (plasmid DNA; carrying the target gene sequence) and negative (non-template; water) controls were included in each Razor Ex amplification. Short cycling parameters included one initial hold for 2 min at 50°C, a first cycle at 94°C for 4 min and 60°C for 15 sec followed by 54 cycles at 91°C for 3 sec and 60°C for 15 sec. The PCR cycling program was uploaded using a barcode (Figure 1). The assays were performed in a Razor Ex BioDetection System.

Real time qPCR sensitivity assays
The detection limits of all three primer and probe sets was determine by performing four sensitivity assays with each set of primer and probe in the Rotor-Gene 6000 thermocycler. Tenfold serial dilution of plasmid or genomic DNA (Xfp) was used at 10 ng to 1 fg per reaction. For each spiked or mixed assay, 1 µl (per reaction) of rind extract (1 ml of TE buffer was used to macerate healthy orange rind and clarified by a 2 min centrifugation at 14,000 rpm; the supernatant was used) was added into serially diluted Xfp genomic DNA. A sensitivity assay spiked with sharpshooter DNA was performed by adding 1 µl (10 ng/µl) crude DNA of sharpshooter into serially diluted Xfp genomic DNA. Each reaction was performed in three replicates.

Positive controls
Positive controls carrying target gene segments of Xfp and Xf were generated for each primer set targeting three different genes. The amplicons generated using endpoint PCR were eluted from the agarose gel using Quantum Prep Freeze 'N Squeeze Spin Columns (Bio-Rad, Hercules, CA) and inserted into a plasmid pCR2.

Primer and probe design
The two primer sets specific for Xfp and the primer set for general detection of X. fastidiosa as well as all respective probes met the desired 100% query coverage and 100% identity after an alignment using BLASTn in the GenBank nucleotide database (Table 3). All primers and probes had ΔG ≤ 1.0 at 60°C (Table 3). To maximize signal and minimize background, the double-quenched probes contained a 5' FAM fluorophore, a 3' IBFQ quencher, and an internal ZEN quencher.

Real time qPCR protocol validation
All three primer and probe sets (Table 3) were designed to perform in endpoint PCR, qPCR, and the field-deployable Razor Ex. Primer and probe specificity was tested against a plant exclusivity panel (described above) and near-neighbor microbial panel (Table 2), and broad range detection of primer/ probe set Xf.csp6 was tested against an inclusivity panel (Table 1) of X. fastidiosa genomic DNA from purified Xf isolates and infected plants and sharpshooters. Genomic DNA of only two Xfp isolates was available, for use in the inclusivity panel for primer and probe sets, Xf.CVC.fim1 and Xf.CVC.pib4, because Xfp was classified as a select agent at the time of these experiments. The primer and probe sets Xf.CVC.fim1, Xf.CVC.pib4 and Xf.csp6 showed no cross reactivity with any member of the exclusivity panel, and generated the expected 109, 81 and 93 bp PCR amplicons, respectively. The primer/ probe sets Xf.CVC.fim1, and Xf.CVC.pib4 amplified only from Xfp while set Xf.csp6 amplified all X. fastidiosa (Table 1). To further confirm the specificity, the amplified products were cloned, sequenced, and assessed using BLASTn against the GenBank database. All the sequences showed highest similarity with the corresponding pathogen. Twenty symptomatic plants, three genomic DNAs from Xfp and X. fastidiosa Temecula, and four sharpshooters infected with X. fastidiosa subsp. multiplex and X. fastidiosa subsp. fastidiosa tested positive with primer/probe set Xf.csp6, while only Xfp isolates were positive using primer/probe sets Xf.CVC.fim1 and Xf.CVC.pib4 (Table 1).  Figure 2). Almost identical Ct values, ranging from 8.37 to 9.73, were obtained from 10 ng of plasmid DNA. The obtained linear graphs and standard curve values for each primer-probe set used to amplify corresponding positive control, suggest optimal reaction efficiency ( Table 4; Figure 2). The detection limit of primer-probe set Xf.CVC.fim1 reached as little as 1 fg; (Table 4; Figure 3) compared to those of primer and probe sets Xf.CVC.pib4 and Xf.csp6 that detected as little as 10 fg with Xfp genomic DNA (Table 4; Figure 3). A small difference in sensitivity and a variation among the replicates at lower concentration (especially at 10 fg and below) of genomic DNA was observed when a Xfp genomic DNA sample was mixed separately with extracts of orange rind and a crude sharpshooter DNA preparation (Table  4; Figure 3). Primer and probe set Xf.CVC.fim1 was able to detect down to 1 fg of genomic DNA of Xfp (CVC50031) mixed with extracts of orange rind and crude sharpshooter DNA but showed a variation in Ct values among the replicates ( Figure  3). To comparing standard graphs generated for all the three primer and probe sets, a manual normalized fluorescence value of 0.2 was used. Generated standard graph values suggested that orange rind extract and crude sharpshooter DNA have little or no inhibitory effect on qPCR sensitivity when the spiked and non-spiked sensitivity assays were performed using all three primer/probe sets (Table 4).

Razor Ex BioDetection System
Empty pouches filled with TaqMan qPCR reagents were used in place of the freeze-dried reagent pouches provided by the manufacturer. Only primer/probe set Xf.csp6 was used to detect X. fastidiosa from infected plant samples with the Razor Ex system. The Primer sets, Xf.CVC.fim1 and Xf.CVC.pib4 were not tested with samples infected with Xfp due to the categorization of Xfp as select agent in the USA at the time this research was conducted. However, all three primer and probe sets were tested using genomic Xfp DNA (isolate CVC50031) using the Razor Ex system and all were positive ( Figure 4) with estimated Ct values of 24 (Xf.CVC.fim1), 24 (Xf.CVC.pib4) and 20 (Xf.csp6). All four infected plant samples (from grape and oak) were positive for X. fastidiosa using primer and probe set Xf.Csp6 ( Figure 5) with estimated Ct values from 33-35. The reactions were performed in only one or two replicates due to the limited number of well slots (only 12) in the Razor Ex pouch. Razor Ex amplification and template DNA preparation results were reproducible. The entire protocol, from DNA extraction to final detection, takes approximately 30 min with no need for laboratory equipment.

Discussion
Xfp CVC is a threat to the U.S. citrus industry. We developed and validated a field deployable, reliable and sensitive Razor Ex and qPCR assays for detection between Xfp and X. fastidiosa, using three primer and probe sets targeting the Table 4. Average Ct values of sensitivity and spiked assays using primer/probe sets Xf.CVC.fim1, Xf.CVC.pib4 and Xf.csp6.  genes encoding fimbrillin, periplasmic iron-binding protein, and cobalamin.
In biosecurity, quarantine and microbial forensics, assay specificity, accuracy and reliability are critical. The use of a multigene format maximizes reliability, specificity and broad range detection and minimizes the chances of false negative and positive results because each targeted gene-segment serves as an internal control for the other targeted genesegments [25]. Two primer and probe sets specific for Xfp, selected after in silico evaluation, targeted genes encoding fimbrillin and a periplasmic iron-binding protein, and one set specific to all X. fastidiosa strains targeted the cobalamin synthesis gene. The three primer and probe sets were highly specific for their targets, and there was no cross reactivity with any other species in the exclusivity panels, which included important crops, vegetables, flowers, grasses, fruit trees and near neighbor microorganisms (Table 2). Primer and probe set Xf.csp6 detected twenty symptomatic X. fastidiosa-infected grape, oak, elm and mulberry plant samples collected from Texas and Oklahoma that did not carry the Xfp strain but were presumptively infected with Xfp closely related species, X. fastidiosa subsp. multiplex (causing agent of scorch disease in peach, almond and oaks) and X. fastidiosa subsp. fastidiosa (causing almond leaf scorch and Pierce's disease of grapes) [3], while primer/probe sets Xf.CVC.fim1 and Xf.CVC.pib4 showed no reaction with these samples ( Table 1). The primerprobe sets Xf.CVC.fim1 and Xf.CVC.pib4 detected the Xfp isolates as expected. Xylella fatidiosa Temecula was also not amplified using the Xf.CVC.fim1 and Xf.CVC.pib4 primer and probe set but amplified using primer and probe set Xf.csp6 (Table 1).
Primer and probe sets Xf.CVC.fim1, Xf.CVC.pib4 and Xf.csp6 showed high sensitivity and efficiency, detecting as little as 1 fg of plasmid DNA and 1 fg (Xf.CVC.fim1) to 10 fg (Xf.CVC.pib4 and Xf.csp6) of Xfp genomic DNA. When genomic DNA from Xfp was mixed with extracts of orange rind and sharpshooter crude DNA, primer and probe set Xf.CVC.fim1 showed slight variation in their replicates, but only at ≤10 fg, indicating that orange rind and sharpshooter crude DNA cause little or no inhibition of qPCR sensitivity. Arif et al. [25], working with cotton leaf and soil extracts in PCR reactions containing genomic DNA of Phymatotrichopsis omnivora, also observed small differences among Ct values. However, they also indicated low reaction efficiency of 0.69 and 0.76 when primer set PoRPB2-2 was tested against P. omnivora genomic DNA spiked with cotton and soil extracts, respectively.
Due to restrictions in the availability of Xfp infected plant samples, only the primer and probe set Xf.csp6, was tested with infected plant samples for pathogen detection using the Razor Ex. However, all three primer-probe sets were tested and validated with Xfp genomic DNA using the Razor Ex for on-field application. Compared to other on-site PCR instruments, the Razor Ex can be more easily transported because of its compact size and light weight (11 lb compared to the SmartCycler 74 lb) and was specifically designed for very rapid thermocycling. To perform the assays with these rapid cycling conditions and regular TaqMan reagents is not possible using traditional qPCR machines. The PCR cycling parameters can be loaded through barcodes (Figure 1) to operate the Razor EX means that the assay can also be used by other end users through direct scan from this publication. Commercially available Razor Ex pouches contain lyophilized PCR reagents to minimize contamination and circumvent cold storage. Because one aim of this research was to develop primers useful in different formats, we introduced commercially available TaqMan PCR components into empty pouches with disposable syringes. The entire assay required about 30 minutes, including approximately 10 minutes for sample preparation (DNA extraction) and 20 to 25 minutes for final detection. For further confirmation of the results for microbial forensics application, the biotinylated forward primer was used in Razor Ex to capture the amplified fragment using streptavidin magnetic beads, if required. Arif et al. [25] has demonstrated that biotinylated primer has no adverse effect on PCR amplification and sensitivity. The detection performances of the Razor Ex system and standard real-time qPCR technology (ABI 7300) were compared for specific detection of the causal agents of anthrax, brucellosis, tularemia, and plague . Standard curves and graphs generated using 10-fold diluted genomic DNA and genomic DNA mixed with orange rind extract or insect crude DNA. A1/A2, A3/A4 & A5/A6: Graphs/standard curve generated using primer and probe set Xf.CVC.fim1 with genomic DNA, genomic DNA mixed with orange rind extract and genomic DNA mixed with insect crude DNA, respectively; B1/B2, B3/B4 & B5/B6: Graphs/standard curve generated using primer/ probe set Xf.CVC.pib4 with genomic DNA, genomic DNA mixed with orange rind extract and genomic DNA mixed with insect crude DNA, respectively; C1/C2, C3/C4 & C5/C6: Graphs/standard curve generated using primer/ probe set Xf.csp6 with genomic DNA, genomic DNA mixed with orange rind extract and genomic DNA mixed with insect crude DNA, respectively. doi: 10.1371/journal.pone.0081647.g003 [23] as well as influenza A viruses [24]. In our hands, the Razor Ex detected P. omnivora [25], High plains virus [31], Xfp and Xf with high assay specificity. The system generates reliable results and can be applied to phyto-sanitary diagnosis, in-field pathogen detection, and other applications in biosecurity and microbial forensics. Our results provide the framework for future development and validation of similar assays for other bacterial plant pathogens of high consequence.