RPAcan3990: an Ultrasensitive Recombinase Polymerase Assay To Detect Angiostrongylus cantonensis DNA

ABSTRACT

Angiostrongylus cantonensis is one of the leading causes of eosinophilic meningitis worldwide. A field-deployable molecular detection method could enhance both environmental surveillance and clinical diagnosis of this emerging pathogen. Accordingly, RPAcan3990, a recombinase polymerase assay (RPA), was developed to target a region predicted to be highly repeated in the A. cantonensis genome. The assay was then adapted to produce a visually interpretable fluorescent readout using an orange camera lens filter and a blue light. Using A. cantonensis genomic DNA, the limit of detection was found to be 1 fg/μl by both fluorometer measurement and visual reading. All clinical samples known to be positive for A. cantonensis from various areas of the globe were positive by RPAcan3990. Cerebrospinal fluid samples from other etiologies of eosinophilic meningitis (i.e., Toxocara sp. and Gnathostoma sp.) were negative in the RPAcan3990 assay. The optimal incubation temperature range for the reaction was between 35°C and 40°C. The assay successfully detected 1 fg/μl of A. cantonensis genomic DNA after incubation at human body temperature (in a shirt pocket). In conclusion, these data suggest RPAcan3990 is potentially a point-of-contact molecular assay capable of sensitively detecting A. cantonensis by producing visually interpretable results with minimal instrumentation.

INTRODUCTION

Angiostrongylus cantonensis is a globally distributed nematode of medical significance due to the neurotropism of 3rd-stage larvae leading to eosinophilic meningitis in both human and animal incidental hosts (1–4). Accurate identification of the presence of this organism is of paramount importance for understanding the epidemiology and clinical decision making. Over the past decade, in an effort to meet this need, quantitative PCR (qPCR) assays have been developed to target A. cantonensis (4, 5). The most sensitive qPCR reported to date, AcanR3990, was designed to target a highly repetitive region in the satellite DNA of A. cantonensis (5).

Although qPCR is the most widely used method of molecular diagnosis, the inherent chemistry of the reaction necessitates thermal cycling, complicating its use as a field-applicable diagnostic test (6). As such, a number of isothermal nucleic acid amplification techniques have been developed for field settings. Among these, the recombinase polymerase assay (RPA), utilizing a bacteriophage recombinase for deannealing of nucleic acid, is adaptable to detect various organisms when maintained at near human body temperatures (7–14). Furthermore, RPA has recently been shown to produce fluorescence visible to the human eye when excited by the correct wavelength of light and viewed through the appropriate filter (15). Recently, an RPA has been developed to target the internal transcribed spacer 1 (ITS1) region of A. cantonensis (16).

To create a more sensitive RPA against A. cantonensis, RPAcan3990 was designed to target the same repetitive region as the AcanR3990 qPCR assay, which in real-time PCR resulted in a 100- to 1,000-times improved detection limit compared to that for the ITS1 target (5). Here, we report the development and initial validation of this RPA using visualized fluorescence as a photometric readout.

MATERIALS AND METHODS

Initial RPA design and evaluation.

RPA primer/probe combinations were designed according to manufacturer’s instructions (TwistDx, Cambridge, UK) to target differing areas of the repetitive AcanR3990 sequence as previously described (5). The assays were initially compared using the previously described A. cantonensis genomic DNA dilution series (5). The best performing assay, RPAcan3990, was carried forward for further evaluation. The sequences are as follows: RPAcan3990F primer, 5′-GTT GCT TTC GAA GCT ATG TAC ATG AAA CAC CTC AA-3′; RPAcan3990R primer, 5′-GCA ACA GTT TCA GCG CAA ATC TGA CGT TC-3′; RPAcan3990 probe, 5′-CAT GAA ACA CCT CAA ATG TGC TTC GAA GTC/iFluorT//idSp//iBHQ-1dT/A AAA TTA GCG CGT AAT/3SpC3/-3′.

RPA reaction.

Each assay tested was prepared by combining 30 μl of TwistDx mastermix buffer (TwistDX, Cambridge, UK), 10 μl of molecular-grade water, 2 μl of 10 μM forward primer, 2 μl of 10 μM reverse primer, 0.6 μl of 10 μM probe. Forty μl of this mixture was then used to suspend the contents of one TwistDx lyophilized EXO reaction tube. The suspension was then halved by transferring 20 μl of this suspension into a second empty microcentrifuge tube. Two microliters of sample and 2.5 μl of 280 mM Mg acetate were then added to each reaction tube. Prior to incubation, the reaction was vigorously mixed by pipetting the entire reaction volume 5 times.

The limit of detection was determined by making serial one-tenth dilutions of A. cantonensis genomic DNA from 100 pg/μl to 1 fg/μl and monitoring 6-carboxyfluorescein (FAM)-gated fluorescence on a ViiA7 qPCR apparatus (Applied Biosystems) for 30 min while incubating the reactions at 40°C. The clinical samples were also incubated at 40°C for 30 min. The effect of temperature and time on assay performance was assessed by using serially diluted genomic DNA as described above and incubating reactions contemporaneously at 30°C, 35°C, 40°C, and in an investigator’s shirt pocket (i.e., near human body temperature) while monitoring every 10 min for a total of 60 min. Both the clinical sample reactions and the time/temperature assessment reactions utilized a visual readout according to a previously published method. In brief, the reaction tubes are placed on a standard blue light gel reader (470-nm wavelength) and visualized by digital photography through a transparent orange photographic filter (15). Positivity was determined by visually comparing the degree of green fluorescence of each reaction mixture to that of the negative control.

DNA extraction.

A. cantonensis genomic DNA used was extracted using a DNA Wizard kit (Promega, Madison, WI) per the manufacturer’s instructions, as described previously (5). DNA was extracted from erebrospinal fluid (CSF) samples (a subset of the samples submitted to the CDC for diagnostic testing due to clinical suspicion for meningitis of parasitic etiology and analyzed as previously described [5]) by using a QIAsymphony instrument (Qiagen, Gaithersburg, MD). The clinical samples were used in accordance with the CDC Institutional Review Board (IRB) protocol entitled “Use of residual diagnostic specimens from humans for laboratory methods research in parasitology.”

RESULTS

Following testing of several potential RPAs against A. cantonensis, we identified (see Materials and Methods) the most analytically sensitive. This assay targeting the AcanR3990 repeat had a limit of detection of 1 fg/μl of genomic DNA when analyzed by fluorometry (Fig. 1). Equal analytical sensitivity was obtained by using a visual readout (Fig. 1, inset). Next, the assay was performed on DNA extracted from CSF samples previously used in the validation of a recently published qPCR assay (5). The visualized readout of each tested sample was in concordance with that obtained by the above-mentioned qPCR. Namely, all samples qPCR positive for Angiostrongylus from diverse geographic locations were positive by RPAcan3990. Furthermore, CSF samples positive for non-Angiostrongylus causes of eosinophilic meningitis (i.e., Toxocara and Gnathostoma) were negative in the RPAcan3990 assay (Fig. 2).

FIG 1.

The visual- and fluorometer-derived limits of detection of RPAcan3990 are equivalent. The limit of detection of the RPAcan3990 was estimated using a one-tenth dilution series of Angiostrongylus cantonensis genomic DNA by fluorescence detection in a fluorometer. The same dilution series was then examined photographically using blue light illumination through an orange transparent filter (inset). The visual and fluorometer readouts of the RPA showed an equivalent limit of detection, 1 fg/μl.

FIG 2.

RPAcan3990 performance on CSF samples from patients with eosinophilic meningitis of known parasitic causes and from uninfected control CSF. (A) RPAcan3990 was performed on archived CSF samples from cases of encephalitis caused by Angiostrongylus cantonensis, Toxocara, Gnathostoma, or by nonparasitic causes (negative controls). Products were visualized through digital photography using a blue light at a 470-nm wavelength (electrophoresis gel reader) and a transparent orange filter. (B) Fluorimetry of the corresponding reaction mixtures was performed and compared with the visualized fluorescence; the fluorimetry data are numbered according to the corresponding reaction in panel A (total time is equal to cycle time × 40 s). (C) RPAcan3990 did not produce visualizable fluorescence when applied to DNA extracted from a CSF sample from a patient with known gnathostomiasis (G. sp.). Also included are a positive control (A. cantonensis genomic DNA; Ac+) and a negative control (Neg).

We next examined the effect of the reaction temperature and incubation time during the RPA on the visual readout (Fig. 3A). Using genomic DNA (gDNA) at a range of concentrations, we saw that at a reaction temperature of 30°C, only gDNA at or above a concentration of 100 pg/μl was positive by 50 min (additional time did not improve the limit of detection). At 35°C, 100 pg/μl was positive by 20 min, and the 1-fg/μl sample was positive at 40 min. At 40°C, 100 pg/μl was positive at 10 min; the 1-fg/μl sample was positive at 20 min. A representative photo of the reactions visualized at 30 min is shown (Fig. 3B). Next assessed was the performance of the assay when incubated at near human body temperature by being carried in a shirt pocket. The reaction produced a visualizable readout at a limit of detection of 1 fg/μl at 30 min (Fig. 3C).

FIG 3.

Effects of temperature and incubation time on the limits of detection using RPAcan3990. (A) Schematic of the kinetics of RPAcan3990 performed on 100 pg/μl, 1 fg/μl, and negative control samples at 30°C (blue), 35°C (purple), and 40°C (red) with visualization of fluorescence at 10-min intervals for 60 min total. (B) Visualization at 30 min is shown in this representative set of products described in panel A. (C) Results obtained using RPAcan3990 performed in duplicate on 100 pg/μl genomic DNA, 1 pg/μl genomic DNA, 1 fg/μl genomic DNA, and negative control (NC) samples following a 30-min incubation in a shirt pocket.

DISCUSSION

Isothermal molecular pathogen detection methods hold promise as a way to leverage the benefits of the molecular revolution into field diagnostics. Recently, a qPCR assay targeting a region of the A. cantonensis genome predicted to be highly repetitive was reported to be the most sensitive test for the detection of A. cantonensis (5). Here, we report the initial development and preliminary validation of RPAcan3990, an RPA targeting the same highly repeated region of the A. cantonensis genome.

The initial evaluation showed that the RPA had an equivalent limit of detection to the previously published qPCR (Fig. 1). In addition, when applied to clinical samples previously tested using the qPCR, the results were concordant with all tested CSF samples known to be positive for neuroangiostrongylias by RPA, while the cases of known neurognathostomiasis, neurotoxocariasis, and parasite-negative samples remained negative (Fig. 2).

As a simplified incubation and detection method would facilitate point-of-contact utility, we streamlined the previously described reaction preparation protocol by replacing a centrifugation step with vigorous pipetting (15). Furthermore, we evaluated a range of temperatures felt to be easier to achieve in the field and demonstrated that the optimal temperature range was between 35°C and 40°C. The incubation was also successfully carried out in a shirt pocket, achieving limits of detection equivalent to those seen using a heating block or thermal cycler. This suggests one could rely on the human body as a constant supply of thermal energy for the RPA (Fig. 3C), making the assay useful in remote settings.

With the purpose of developing a field-friendly readout, we adapted a visualization technique previously described for a COVID-19 diagnostic assay (15). Fluorophores utilized frequently in molecular probes have both absorbance and emission wavelengths within the visual spectrum (e.g., FAM, 6-carboxy-2,4,4,5,7,7-hexachlorofluorescein [HEX], etc.). Taking advantage of this attribute, reaction illumination using a standard gel reader as a blue light source and visualized through an orange filter allowed an equal limit of detection to that with the fluorometer (Fig. 1, inset). When applied to clinical samples, this method found concordance with fluorometer readouts. We are actively attempting to develop methods to improve the visual discernibility of samples near the limit of detection (such as samples 1 and 7) (Fig. 2).

Although the present study is limited by the small number of samples, because it targets the exact same repeat (AcanR3990) known to be absent in all other viruses, bacteria, protozoa, and helminths except Angiostrongylus mackerrasae, this RPA would be expected to remain highly specific for A. cantonensis except in the narrow geographic area in Australia where it overlaps with Angiostrongylus mackerrasae.

In conclusion, RPAcan3990 is an RPA designed to target a DNA sequence predicted in silico to be highly repeated within the A. cantonensis genome. The data suggest that a limit of detection of 1 fg/μl of A. cantonensis genomic DNA can be achieved using a simplified reaction preparation, a visualizable fluorescent readout, and incubation using the human body as the only heat source.