A recombinase polymerase amplification-lateral flow dipstick assay for rapid detection of the quarantine citrus pathogen in China, Phytophthora hibernalis

RPA primers and probe design

Sequences of the Ypt1 gene of P. hibernalis (GenBank Accession No. KJ755160) and other Phytophthora species were obtained from GenBank (Benson et al., 2018). Alignment of these sequences (Fig. S1) was carried out using Clustal W (Larkin et al., 2007). Six pairs of primers (Table S1) targeting the polymorphic regions of the Ypt1 gene were designed for P. hibernalis according to the TwistAmp^® DNA Amplification Kits Assay Design Manual (https://www.twistdx.co.uk/docs/default-source/RPA-assay-design/twistamp-assay-design-manual-v2-5.pdf?sfvrsn=29). Their specificity to P. hibernalis was preliminarily evaluated using a collection of Phytophthora species (Table 1) in conventional PCR. A primer pair, including a forward primer PhRPA-F and a reverse primer PhRPA-R was found exclusively and consistently amplified the Ypt1 gene of P. hibernalis isolates. The reverse primer was then adjusted by conjugating biotin at its 5′-end (Table 2). A probe PhProbe (Table 2) was designed according to the TwistAmp^® DNA Amplification Kits Assay Design Manual. This set of primers and probe was tested in RPA and subsequently used for optimizing the reaction condition and evaluating assay specificity and sensitivity. Primers and probe (Table 2) were synthesized by Nanjing GenScript Co. Ltd. (Nanjing, China).

Isolate selection

Isolates of oomycete and fungal species used in this study are listed in Table 1. P. hibernalis isolate 9099 was recovered from an imported pomelo fruit (Citrus grandis) at Shanghai Export and Import Inspection and Quarantine Bureau (Jiao et al., 2017). All isolates were part of collections at Department of Plant Pathology at Nanjing Agricultural University and Department of Forest Protection at Nanjing Forestry University in Nanjing, China.

Culture conditions and DNA extraction

Isolates of Phytophthora species were cultured in 10% clarified V8 juice agar (cV8A) at 18 to 25 °C in the dark. Mycelia of oomycete and fungal isolates were produced by growing individual isolates in 20% clarified V8 juice and potato dextrose broth, respectively, at 18–25 °C for 3–5 days. Mycelia were harvested by filtration and then frozen at −20 °C. Genomic DNAs (gDNA) were extracted using a DNAsecure Plant Kit (Tiangen Biotech, Beijing, China). DNA concentrations were measured by a NanoDrop 2000c spectrophotometer (NanoDrop Technologies, Thermo Fisher Scientific, Wilmington, DE, USA). Nuclease-free water (nfH₂O; Thermo Fisher Scientific, Wilmington, DE, USA) was used to dilute gDNA extractions as needed. DNA samples were stored at −20 °C until use.

RPA–LFD assay

RPA amplifications were performed in 50-μL reactions according to the quick guide of the TwistAmp^® nfo Kit (TwistDx Ltd., Cambridge, UK). After mixing 2.1 μL of each of PhRPA-F and PhRPA-R primers (10 µM), 0.6 μL of PhProbe probe (10 μM), 29.5 μL of rehydration buffer (supplied in the kit) and 11.2 μL of nfH₂O, the reagent mixture of each reaction was then added to the freeze–dried reaction pellet of the TwistAmp^® nfo Kit. DNA template (2 µL) and 2.5 µL of magnesium acetate (280 mM) were added into the mixture. Reactions were carried out at 38 °C. To determine the optimal amplification duration, reactions were performed for 5, 10, 15, 20, 25, and 30 min. Each duration was tested three times. Each reaction included 2 µL of the gDNA (10 ng per µL) of P. hibernalis isolate 9099. After amplification, 10 μL of RPA product was mixed with 90 μL of phosphate buffered saline with Tween 20 (PBST) running buffer then 10 μL of RPA product-PBST mixture was pipetted to the sample pad of a HybriDetect one LFD (Milenia Biotec, Giessen, Germany). The dipstick was dipped into a tube containing 200 μL of PBST for up to 5 min until a clear control line was visible. This visualization step using LFD was performed at room temperature (average 22 °C). When test and control lines were simultaneously visible, the detection result was positive. If only control line was visible, it indicated a negative detection result. After visualization, LFDs were air-dried and photographed using a Canon PowerShot SX730 HS camera.

Sensitivity of RPA–LFD and PCR assays

Ten-fold dilutions of gDNA of P. hibernalis isolate 9099 including 100, 10, 1, 0.1, 0.01, 0.001 and 0.0001 ng per µL were used as templates in the RPA–LFD assay (20 min RPA). nfH₂O was used in place of the DNA template as a no-template control (NTC) included in each set of reactions. This evaluation of sensitivity was performed three times.

To compare the sensitivity of the RPA–LFD assay to conventional PCR, the same 10-fold dilutions of gDNA of isolate 9099 were used in a PCR assay. Each dilution was used as the DNA template (2 µL) in each 50-µL PCR reaction. Each reaction also included 25 µL of 2× Taq Master Mix (containing Taq DNA Polymerase, dNTP and an optimized buffer; Vazyme Biotech, Nanjing, China), 2 µL of each of PhRPA-F and PhRPA-R primers (10 µM) and 19 μL of nfH₂O. PCR reactions were completed in a SimpliAmp™ thermal cycler instrument (Model A24812, Thermo Fisher Scientific, Wilmington, DE, USA) following an initial denaturation step at 95 °C for 3 min, 35 cycles of 95 °C for 15 s, 56 °C (optimized) for 15 s, and 72 °C for 15 s, plus a final extension at 72 °C for 5 min. Each set of PCR reactions included an NTC. PCR products were examined in 1% agarose gel electrophoresis at 120 V for approximately 25 min. Agarose gel was stained by ethidium bromide and visualized on a transilluminator. The PCR assay was carried out three times.

Specificity of the RPA–LFD assay

The RPA–LFD assay was evaluated using gDNAs of 45 isolates of 39 species (Table 1). Each reaction included 2 µL of DNA template (10 ng per µL). The assay was performed three times against each isolate.

Detection of P. hibernalis in artificially inoculated orange fruits using RPA–LFD

Prior to inoculation, orange fruits (Citrus sinensis) were washed in distilled water, immersed in 70% ethanol for 10 sec and then rinsed with distilled water. Each fruit was wounded (approximately 2 cm depth) using a sterile inoculation needle. A 5-day-old cV8A plug (1 × 1 cm) of P. hibernalis isolate 9099 was placed onto the wound site of each of six replicate fruits and secured with parafilm. A sterile cV8A plug was used on a non-inoculated control fruit. All fruits were then placed on two layers of wet filter papers in a container and stored in a dark incubator set at 20 °C. After four days, symptoms resembling those of brown rot were observed around the wound sites of inoculated fruits. Fruits were rinsed with distilled water to remove the remaining agar plugs. Both flesh and peel tissues around the wound site were cut into approximately 1 × 1 cm pieces and placed onto PARP-cV8A selective media (Jeffers & Martin, 1986) to recover Phytophthora isolates at 20 °C. Total DNAs of these tissues were extracted using a previously described NaOH method (Wang, Qi & Cutler, 1993). Briefly, 20 mg of plant tissues collected from the wound site of each fruit were placed into a 1.5 mL microtube containing 200 µL of NaOH (0.5 N). They were grinded for approximately 1 min until no large pieces of plant tissues were visible using a sterile tissue grinder pestle. Then, 5 µL of grinded tissues in NaOH were transferred to a new microtube containing 495 µL of Tris buffer (100 mM, pH 8.0). Two µL of the mixture were used as the DNA template in each RPA reaction. Isolate 9099 gDNA (10 ng per µL) and nfH₂O were used as a positive control and NTC, respectively. This experiment was repeated once.

Optimal duration of RPA

RPA was performed using isolate 9099 gDNA (10 ng per µL) for different amplification durations ranging from 5 to 30 min with 5-min intervals. Both test and control lines were visible on all dipsticks in three repeats of the experiment. Whereas the test lines correlated to 5 min of RPA were weak, as the amplification extended to 10 min and longer, they increased in intensity (Fig. 1). Based on the finding, the amplification duration was set at 20 min for the subsequent RPA reactions.

Sensitivity and specificity of the RPA–LFD assay

In the sensitivity evaluation of the RPA–LFD assay, all dipsticks yielded visible control lines, indicating valid tests. Test lines were visible on all dipsticks correlating to RPA reactions using ≥0.2 ng (2 µL, 0.1 ng per µL) gDNA of P. hibernalis isolate 9099 (Fig. 2A). No test lines were observed on dipsticks using ≤0.02 ng of gDNA (Fig. 2A). In the comparative evaluation of the PCR assay, ≥2 ng of gDNA was required in each PCR reaction for the successful amplification using primers PhRPA-F and PhRPA-R (Fig. 2B). The above results were consistent among three repeats of each assay.

In the specificity evaluation of the RPA–LFD assay, all LFDs had visible control lines. Only the reactions using gDNAs of P. hibernalis isolates yielded positive detection results (Table 1). The results were consistent among three repeats of the experiment against each isolate.

Detection of P. hibernalis in artificially inoculated orange fruits using RPA–LFD

In the first repeat of the experiment, P. hibernalis isolates were recovered from all six inoculated orange fruits, which had shown typical brown rot symptoms at the wounded sites four days after inoculation, while not from the non-inoculated control fruit. In the RPA–LFD assay, P. hibernalis DNAs were detected in total DNAs extracted from six artificially inoculated fruits, while the non-inoculated control lacked detection of P. hibernalis (Fig. 3). In the second repeat experiment, only three of six inoculated fruits had shown brown rot symptoms after four days, while the other three as well as the non-inoculated control remained asymptomatic. Correspondingly, total DNAs of three symptomatic fruits yielded positive detection of P. hibernalis in the RPA–LFD assay, whereas the remaining fruits gave rise to negative results (Fig. 3).