Environmental detection of Fasciola hepatica by loop-mediated isothermal amplification

Fasciola hepatica genomic DNA extraction

Adult F. hepatica were sourced from Victorian abattoirs. Briefly, cattle and/or sheep livers were collected and cut into ~10 mm sections following Reichel (2002) with minor modifications: livers were palpated to remove adult flukes from the liver sections, flukes were then washed three times in sterile phosphate buffered saline (100 mM Na₂HPO₄, 20 mM KH₂PO₄, 1.37 M NaCl, 30 mM KCl, pH 7.4), placed into nucleic acid preservation buffer containing 19 mM EDTA, 18 mM Na₃C₆H₅O₇, 3.8 M (NH₄)₂SO₄, pH 5.2 (Camacho-Sanchez et al., 2013) and frozen at −20 °C until required for either genomic DNA (gDNA) extraction or egg collection. Genomic DNA was prepared following manufacturer instructions using a Bioneer AccuPrep® Genomic DNA Extraction Kit (Bioneer, Daejeon, South Korea) with minor modifications as follows: 20–25 mg of tissue was cut from the posterior end, minced using a scalpel blade on a petri dish and transferred to a microfuge tube containing tissue lysis buffer pre-heated to 60 °C for further homogenisation with a pellet pestle (Merck, Darmstadt, Germany). Elution was performed in 100 µL volumes using TE buffer (10 mM Tris HCl, 0.1 mM EDTA, pH 8.4) pre-heated to 65 °C and eluted twice to improve recovery. Eluted DNA was assessed using a NanoDrop™ 2000 Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), with absorbance readings for 280/260 and 260/230 ratios between 1.8–2.0 considered acceptable. Samples were stored at −20 °C until required.

Helminth genomic DNA extraction for specificity panel

Common livestock helminths (Table 1), kindly provided by Dr Tim Elliot (Invetus, Armidale, NSW, Australia), were washed three times with 70% EtOH. Adult helminths were weighed, and 20–25 mg of tissue excised from the anterior end where possible, minced using scalpel blades, and transferred to microfuge tubes. Larvae provided in absolute EtOH were sedimented by centrifugation at 13,000g for 15 min and assessed for visible pellet formation. Centrifugation was repeated if no pellet was observed and ethanol was then removed. Total gDNA was extracted following Brindley et al. (1989) with minor modifications as follows: 25 µL grinding buffer containing 80% (v/v) homogenisation buffer (100 mM NaCl, 200 mM sucrose, 10 mM EDTA, 30 mM Tris-HCl pH 8.0) and 20% (v/v) lysis buffer (250 mM EDTA, pH 8.0, 2.5% (w/v) SDS, 500 mM Tris-HCl, pH 9.2), pre-heated to 60 °C was added to each tube containing each helminth species and were mechanically disrupted using pellet pestles (Merck, Kenilworth, NJ, USA). Following incubation at 65 °C for 30 min, 14 µL of 8 M ammonium acetate was added to achieve a final concentration of 1 M, then placed on ice for 30 min. Tubes were centrifuged, 13,000g at 4 °C, for 10 min to pellet precipitated proteins and the supernatant containing DNA was transferred to a fresh microfuge tube for ethanol precipitation.

Briefly, 2.5 volumes of chilled absolute ethanol and 1:10 ratio 3 M sodium acetate was added to the solution. Samples were inverted and incubated at −20 °C for 60 min, or overnight and pelleted by centrifugation, 13,000g at 4 °C, for 15 min. Samples were checked for visible pellets before removing the ethanol. Pellets were subsequently washed with chilled 70% ethanol, re-pelleted by centrifugation, ethanol removed, and the process repeated at least three times. After the final wash, pellets were inverted and air dried before eluting in 50 µL TE buffer pre-heated to 65 °C. Sample quality and concentration was assessed using a NanoDrop™ 2000 Spectrophotometer as previously described, standardised to a final concentration of 5 ng/µL in TE buffer and stored at −20 °C.

Plasmid DNA preparation for specificity panel

Recombinant plasmids were needed, owing to the difficulty in obtaining snail individuals of species which either serve as intermediate host, or share the same habitat as F. hepatica transmitting snails in South-East Australia. Such synthetic DNA constructs containing the internal transcribed spacer (ITS-2) (Table 2) were cloned into pBHA vectors by Bioneer and transformed into chemically competent XL1 Blue E. coli cells (New England BioLabs, Ipswich, MA, USA) following manufacturer instructions. Total plasmid DNA was extracted using the AccuPrep^® Plasmid Mini Extraction Kit (Bioneer, Oakland, CA) as per manufacturer instructions, eluting twice with TE buffer pre-heated to 65 °C. Samples were assessed for quality and quantity using a NanoDrop™, standardised to 5 ng/µL, and stored at −20 °C until needed.

Fasciola hepatica quantitative PCR

Quantitative PCR was used to validate and compare initial F. hepatica LAMP (FhLAMP) results. A specific quantitative PCR (qFH) assay for F. hepatica detection was modified from Rathinasamy et al. (2018) and performed as a singleplex. Reactions were run in 25 µL volumes containing 1× SensiMix II Probe mastermix (Bioline, Alexandra, NSW, Australia), 0.3 µM each of forward and reverse primers, 0.1 µM probe (Table 3), 1 mM MgCl₂ to achieve a final concentration of 4 mM MgCl₂, and 2.5 µL template (diluted 1/50 for faecal samples). Amplification was performed using a MIC qPCR cycler (BioMolecular Systems, Upper Coomera, Qld, Australia) under the following cycling conditions: 10 min at 90 °C, followed by 40 cycles of 10 s at 95 °C and 20 s at 60 °C. All PCR runs contained a positive control to monitor intra-assay variation, and a no template control (NTC) containing nuclease-free water (NFW) to monitor contamination. All qPCR runs were assessed using the MIC PCR program (V2.6.4) applying a dynamic normalisation method, with a set threshold of 0.5 normalised fluorescence units, excluding the first give amplification cycles.

Further optimisation of a Fasciola hepatica LAMP

F. hepatica LAMP (FhLAMP) primers (Table 4) were sourced from Martínez-Valladares & Rojo-Vázquez (2016), and two loop primers, LF and LB (Table 4) were manually designed using CLC Genomics (Qiagen, Hilden, Germany) following loop primer design requirements (Nagamine, Hase & Notomi, 2002) Amplification was optimised in final volumes of 25 µL consisting of 15 µL isothermal mastermix (ISO-DR004; OptiGene, Horsham, United Kingdom), 5 µL primer mix with optimised final concentrations of 2 µM inner primers (FIP & BIP), 0.2 µM outer primers (F3 & B3), 2 µM loop primers (LF & LB), and 5 µL template following manufacturer instructions. All reactions contained a positive control to monitor inter-assay variation, and NTC reactions containing NFW to monitor contamination. Reactions were performed in either a Genie II or Genie III fluorometer (OptiGene) with an initial pre-heat of 40 °C for 1 min, followed by amplification at 65 °C for 30 min, and anneal from 94–84 °C at 0.5 °C/s. Results were reported as time to positive (T_p) in minutes and seconds (mm.ss), with anneal derivative melting temperature (T_a) reported in degrees Celsius (°C). Performance was determined by assessing the analytical sensitivity and specificity, with inter-assay variation of FhLAMP assessed by performing 10 replicate runs using a 10-fold serial dilution of F. hepatica gDNA with starting concentrations of 5 × 10⁰ ng/µL − 5 × 10⁻⁵ ng/µL in duplicate replicates. The limit of detection was determined where at least 95% of samples recorded T_p values with standard deviations below a value of ten, and analytical specificity assessed using prepared helminth and snail DNA as described earlier.

Determining optimal dilution factor for FhLAMP amplification from cattle faecal samples

Known negative cattle faecal samples confirmed by liver fluke faecal egg counts (LFEC) and coproantigen ELISA (cELISA) used in this study were kindly provided by Dr Jane Kelley, La Trobe University. During initial validation to mimic sample preparation under field conditions and to assess the effects of inhibitory substances on DNA amplification, samples were measured using disposable 5 µL inoculating loops. Two large loopfuls of faeces was transferred into a microfuge tube ranging between 190–350 mg (average 239 mg) faeces per sample. Faecal samples were spiked with decreasing quantities of F. hepatica gDNA; 1, 0.5, 0.25, and 0.125 µg equivalent to 1.96 × 10⁰, 9.9 × 10⁻¹, 4.98 × 10⁻¹, and 2.49 × 10⁻¹ ng/µL respectively when back calculated on the assumption of 500 µL faecal content volume. DNA extraction was performed following the method by Matsui et al. (2007) with minor modifications as follows: 500 µL 10% w/v Chelex resin (Bio-Rad, Hercules, CA, USA) prepared in NFW and 50 µL 10% w/v polyvinylpolypyrrolidone (average molecular weight 360,000) was added to samples described above then shaken by hand to form an even homogenate before boiling at 100 °C for 10 min. After boiling, samples were cooled to ambient temperature and to sediment any debris, including the Chelex resin. The supernatant was used for DNA amplification and assessed using a range of dilution factors from neat, 1/10, 1/20, 1/50, 1/100, 1/200, 1/500, and 1/1,000 for each sample to determine the optimal dilution factor required for diluting out faecal inhibitory substances known to interfere with DNA polymerases.

Acquisition of F. hepatica eggs for faecal spiking trials from adults

Eggs were collected from previously obtained adult liver flukes obtained (as described earlier), by placing a single adult in a 35 mm diameter petri dish under a stereo microscope with the specimen placed ventral side up to locate the ventral sucker. The specimen was then covered in MilliQ water to prevent desiccation. Fine forceps and an angled teasing needle were used to gently open the specimen immediately underneath the ventral sucker where the uterus, containing eggs was located and agitated to release eggs into solution. Specimens were then rinsed with MilliQ water using a wash bottle prior to removal to dislodge residual eggs. The contents of the dish were thoroughly washed with MilliQ water into a Falcon® 40 µm cell strainer (Corning, Corning, NY, USA) to ensure no residual eggs remained on the petri dish. The strainer was rinsed three times with MilliQ water before carefully being placed upside down over a 50 mL conical tube and washed with MilliQ water to dislodge any trapped eggs into the tube. Washed eggs were sedimented by centrifugation at 3,000g for 10 min and checked for pellet formation before transferring them into a microfuge tube containing 200 µL NFW, covered with foil and stored at 4 °C for up to 1 month, or until required. Prior to use, the egg solution was gently swirled and 50 µL of egg suspension transferred into a 35 mm diameter petri dish to count the required number of F. hepatica eggs. Eggs were checked for integrity, discarding damaged eggs, counted and then transferred by pipetting into known negative faecal samples for spiking trials.

Optimisation of a minimal processing DNA extraction method from cattle faecal samples

Different lysis buffers were trialled to assess which one yielded the most consistent amplification and lowest T_p values using minimal processing steps. Each lysis buffer was trialled using known F. hepatica negative cattle faecal samples and F. hepatica eggs (as described earlier), with lysis buffers assessed in this study described in Table 5 below. For each buffer analysed microfuge tubes containing faecal samples weighing between 190–350 mg (prepared as described earlier) were spiked with 1, 5, 10, 20, and 50 F. hepatica eggs each (as described earlier). Samples were vigorously shaken by hand to form an even homogenate before boiling at 95–100 °C for 10 min, then cooled to ambient temperature and allowed to sediment prior to LAMP amplification.

Environmental DNA concentration of water samples

Water known to be free from F. hepatica was collected from the La Trobe University moat (GPS coordinates: −37.718, 145.044) and spiked with F. hepatica gDNA to evaluate a field-appropriate sampling method of circulating free gDNA from water sources. It was necessary to validate the water sampling method this way due to COVID-19 travel restrictions limiting access to F. hepatica positive properties. Samples were aliquoted into 50 mL conical tubes and spiked with the same quantities of F. hepatica gDNA used for the faecal spiking trials, with starting DNA quantities of 1, 0.5, 0.25, and 0.125 µg added to each aliquot to final concentrations of 2 × 10⁻²_, 1 × 10⁻²_, 5 × 10⁻³ and 2.5 × 10⁻³ ng/µL per spiked sample. Additionally, 250 mL volumes with the same starting quantities of F. hepatica DNA were used (1. 0.5, 0.25, and 0.125 µg), to yield approximate final concentrations of 4 × 10⁻³, 2 × 10⁻³, 1 × 10⁻³, and 5 × 10⁻⁴ ng/µL each. A negative control of moat water with no spiked gDNA was also included to assess non-target gDNA amplification. Each sample was aspirated through a 50 mL Luer-Lok® syringe before attaching a 0.22 µm Millipore Express (PES) cartridge (Cat #: SVGP1050, Sigma). Contents were filtered until either the entire contents were filtered or the filter was clogged and the filtrate discarded. The syringe and filter were then reversed and the filter contents back drawn, with the resulting eluate transferred to a fresh microfuge tube. A dilution factor of 1/10 was chosen from previous studies for DNA amplification to reduce the presence of DNA amplification inhibitors present from environmental samples (Rathinasamy et al., 2018).

Analytical performance of the Fasciola hepatica LAMP (FhLAMP)

The analytical performance of FhLAMP was assessed using a ten-fold serial dilution of F. hepatica gDNA with starting concentrations ranging from 5 to 5 × 10⁻⁵ ng/µL. Reactions were performed in duplicate and repeated 10 times to assess analytical sensitivity and inter-assay variation (Table 6). Amplification times ranged from 07.15–22.25 (mm.ss), with T_a values ranging from 88.3–89.4 °C. At a starting concentration of 5 × 10⁻⁵ ng/µL, this resulted in some T_p values exceeding 20 min with high inter-assay variation assessed by percent coefficient of variation (CV%) as being above the 10% cut-off value, where amplification was deemed unacceptable due to high T_p variability. At this concentration (5 × 10⁻⁵ ng/µL) amplification was unreliable, with T_p values ranging from 16.30–22.30. Therefore 5 × 10⁻⁴ ng/µL was considered the assay limit of detection producing an average T_p value of 13.78, with all replicates consistently amplifying through all ten runs and low inter-assay variation of <5% CV.

The analytical specificity of FhLAMP was assessed using two measures: (i) gDNA from common livestock helminths (Figs. 1A and 1B) (Table 1) and synthetic constructs of snails (Fig. 1C) relevant to either the F. hepatica lifecycle or fluke ecology (Table 2), (ii) assessing non-target amplification from known F. hepatica negative cattle faecal samples (Fig. 1D), with all specificity panel results displayed in Fig. 1. Samples were considered negative for F. hepatica if no T_p values were recorded. Some intermittent ramping was visible in some of the Genie amplification curves after 25 min (Figs. 1B and 1D), however no time to positive detection was recorded for those samples when analysed using the Genie Explorer program or directly on the Genie fluorometer. This is likely due to fluorescence values from these samples failing to exceed the automatically defined detection threshold set by the manufacturer, and were subsequently considered negative. This may be explained by intermittent non-target amplification being occasionally observed in late-stage amplification, probably caused by primer-dimer formation and/or a deterioration in polymerase activity.

No T_p values were recorded for all non-target samples used in this study, suggesting that this assay has high specificity even from samples containing large quantities of non-target DNA from a broad range of samples.

Determining the optimum dilution factor from faecal DNA sampling

The initial trial of known negative samples spiked with F. hepatica gDNA was assessed using both qFh and FhLAMP, with samples screened using FhLAMP across a range of dilutions from 1 to 1/1,000 (Fig. 2). A dilution factor of 1/50 was chosen as this dilution provided the most reliable T_p values ranging from 10–14 min (± ≤0.81 SD) across all starting DNA quantities (Fig. 2C). Although the fastest T_p values (between 10–14 min, ± ≤1.52 SD) were observed using a 1/100 dilution (Fig 2D), at lower starting quantities of DNA (from 0.25 µg) amplification was inconsistent or T_p values between replicates were highly variable. Dilution factors of 1/20 (Fig. 2B) and 1/200 (Fig. 2E) produced slower T_p values (between 11–17 min for 1/20, and 11–15 min 1/200) relative to the 1/50 and 1/100 dilutions. Higher standard deviations and mean T_p values of 15.75, ±0.60 SD and 16.30, ±1.52 SD for 1/20 and 1/200 respectively where also recorded at a starting DNA quantity of 0.125 µg in comparison to the 1/50 dilution factor (Fig. 2). This suggests that the amplification at these dilutions may have been affected by either amplification inhibition (1/20) or excess dilution of samples (1/200). All samples prepared using a 1/50 dilution were further analysed using qFh to assess Cq variation (SI 1) between intra-assay replicates to confirm that a 1/50 dilution factor is sufficient for both assays, as faecal samples are known to contain inhibitors that interfere with DNA polymerases. Both qFh and FhLAMP were able to reliably detect F. hepatica gDNA from spiked cattle faecal samples with minimal variation between replicates (± ≤0.5 SD qFh, ± ≤0.81 SD FhLAMP), using a 1/50 dilution factor with Cq values for qFh ranging from 26–28, and T_p values for FhLAMP from 10–14 min (Table 7). This suggests that a dilution of 1/50 is sufficient for DNA amplification and that subsequent clean-up steps are not necessary prior to DNA amplification.

Comparison of lysis buffers for use in cattle faecal samples spiked with F. hepatica eggs

Cattle faecal samples were spiked with known quantities of F. hepatica eggs containing: 1, 5, 10, 20, and 50 eggs per lysis buffer. Samples were boiled for 10 min at 95–100 °C for DNA extraction using the minimal processing method described earlier and assessed by FhLAMP. Various lysis buffers were used as detailed in Table 5, and a dilution factor of 1/50 chosen for all samples. Figure 3A represents the mean of all duplicate replicate T_p values using different lysis buffers, with extraction and alkaline lysis buffers failing to amplify any F. hepatica DNA. Additionally, Chelex and gram-positive buffers amplified inconsistently, with one of two duplicate replicates returning T_p values. Tissue lysis and SET buffers returned positive T_p values for both replicates from samples containing 20 and 50 eggs with an average T_p for samples prepared using tissue lysis buffer of 12.22 (±1.10 SD) and 14.88 (±0.81 SD) for 20 and 50 eggs respectively, with mean T_p values of 15.58 (±0.60 SD) and 13.73 (±0.60 SD) for 20 and 50 eggs prepared using SET buffer (Fig. 3B). As standard deviations were higher in samples prepared using tissue lysis buffer, SET buffer was therefore determined to be the optimal lysis buffer for use in F. hepatica egg DNA extraction from faecal samples, consistently amplifying F. hepatica DNA from as little as 20 eggs per sample.

Evaluation of an extraction free water DNA isolation method

Water known to be F. hepatica free was collected from the La Trobe University moat and used to assess an extraction-free DNA capture method suitable for field-use and LAMP detection. Each sample, containing either 50 or 250 mL of water was spiked with decreasing quantities of DNA (1, 0.5, 0.25 and 0.125 µg) and a negative extraction control of untreated moat water, was filtered and diluted 1/10 for use in FhLAMP assays. The average T_p values for each sample, with all spiked samples using 50 mL volumes amplifying, returned T_p values ranging from 10.30–20.15 (mm.ss) (± ≤4.21 SD) (Table 8). Increasing sample volume to 250 mL, whilst retaining the same starting DNA quantities, provided slower T_p values of 13.15–21.15 (mm.ss) (± ≤5.06 SD) and samples spiked with 0.125 µg F. hepatica gDNA failed to amplify. This is likely due to the Sterivex® filters clogging after ~100 mL and therefore were unable to filter the entire sample volume to sufficiently concentrate the free DNA in that sample. Regardless, this suggests that the extraction free method of DNA capture described here is suitable for field-use with all 50 mL spiked samples and three of four 250 mL spiked samples recording positive T_p values. Despite whole gDNA being spiked into moat water, samples containing high levels of competing non-target DNA and DNA amplification inhibitors with no DNA denaturation or clean-up steps used, produced detection levels as low as 1 × 10⁻³ ng/µL F. hepatica gDNA from partially filtered 250 mL spiked water amplifying in under 20 min.