Non‐lethal loop‐mediated isothermal amplification assay as a point‐of‐care diagnostics tool for Neoparamoeba perurans, the causative agent of amoebic gill disease

Abstract Neoparamoeba perurans is the causative agent of amoebic gill disease (AGD). Two loop‐mediated isothermal amplification (LAMP) assays targeting the parasite 18S rRNA and the Atlantic salmon EF1α, used as internal control, were designed. The N. perurans LAMP assay did not amplify close relatives N. pemaquidensis and N. branchiphila, or the host DNA. This assay detected 106 copies of the parasite 18S rRNA gene under 13 min and 103 copies under 35 min. Five “fast‐and‐dirty” DNA extraction methods were compared with a reference method and further validated by TaqMan™ qPCR. Of those, the QuickExtract buffer was selected for field tests. Seventy‐one non‐lethal gill swabs were analysed from AGD‐clinically infected Atlantic salmon. The pathogen was detected under 23 min in fish of gill score >2 and under 39 min for lower gill scores. About 1.6% of the tests were invalid (no amplification of the internal control). 100% of positives were obtained from swabs taken from fish showing gill score ˃3, but only ~50% of positives for lower gill scores. The present LAMP assay could be implemented as a point‐of‐care test for the on‐site identification of N. perurans; however, further work is required to improve its performance for lower scores.

AGD can be proactively managed by regular inspection of the gills of anaesthetized fish for gross AGD signs. A "gill index" is used internationally, with a scale from 0 = no lesions to 5 = extensive lesions, by examining all the hemibranch surfaces (Noguera et al., 2019). The gill index allows the farmer to plan treatments in a cost-effective manner (Taylor, Muller, Cook, Kube, & Elliott, 2009).
However, other pathogens can cause gill disorders in marine-farmed salmon, and can present as bleeding gills, pale/thickened patches on the gills, focal lesions and necrosis of the gill lamellae and rakers. Those gill disorders may be multifactorial and might include phytoplankton, parasites, jellyfish, algal blooms, bacteria and viruses (Baxter, Rodger, McAllen, & Doyle, 2011;Rodger, Henry, & Mitchell, 2011;Rodger & Mitchell, 2007). The gill score is therefore presumptive, a confirmatory test of the presence of N. Perurans is desirable before treatments are conducted. For the confirmation of N. perurans, a wet mount of gill smear can be done at the farm site for trained staff for the presence of amoebic cells; however, other morphologically similar Neoparamoeba species, such as N. pemaquidensis and N. branchiphila, which are not pathogenic, can be isolated from gill smears and confounded as the aetiological agent of AGD (Young et al., 2014). Histology assessments can also confirm the presence of amoebas in association with gill lesions, but it is time-consuming and difficult to carry out on-site.
The targeted tissue for the AGD infection allows for non-destructive sampling based on gill swabs. A strong correlation of the pathogen identification by molecular methods, gill score and histopathology scores has been demonstrated (Downes et al., 2017). Available PCR-based methods, all of them targeting the parasite 18S rRNA gene, can be used to confirm the presence of N. perurans (Bridle, Crosbie, Cadoret, & Nowak, 2010;Downes et al., 2017;Fringuelli, Gordon, Rodger, Welsh, & Graham, 2012;Young, Dyková, Nowak, & Morrison, 2008); however, those tests are time-consuming and require a laboratory and trained staff.
Loop-mediated isothermal amplification (LAMP) amplifies nucleic acids with high specificity, sensitivity and rapidity under isothermal conditions. The assay uses DNA polymerase with high strand displacement activity and a set of six specific primers on the target DNA to achieve highly selective nucleic acid amplification (Notomi et al., 2000). DNA can be amplified 10 9 -to 10 10 -fold in 15-60 min.
There is no requirement for the temperature cycling of conventional polymerase chain reaction amplification; therefore, assay times are reduced. Additionally, due to its ability to amplify nucleic acid under isothermal conditions simple and low-cost equipment can be used (Sahoo, Sethy, Mohapatra, & Panda, 2016). Several LAMP assays have been evaluated for the identification of aquaculture pathogens, including bacterial, viral and parasitic pathogens causing serious diseases in aquaculture (Biswas & Sakai, 2014). These tests present the potential of using LAMP for point-of-care (POC) tests; however, effective POC DNA extraction methods, based on "fast-and-dirty" DNA extractions, are also essential to develop rapid and user-friendly molecular diagnostic assays for field sampling (Lau & Botella, 2017).
In this study, a LAMP assay, using real-time fluorescence equipment, was evaluated for the detection of N. perurans on swabs taken non-lethally from Atlantic salmon gills. Furthermore, several POC DNA extraction methods were compared for field applications.
Finally, the chosen protocol underwent preliminary field testing, and the results were compared with visual gill scores.

| DNA extractions from amoebic culture
Genomic DNA was extracted either from Isohelix DNA Buccal Swabs (Sigma-Aldrich, UK) soaked in the in vitro culture of N. perurans, or from aliquots of N. pemaquidensis, and aliquots of serial dilutions of N. perurans cells were counted in a TC20 automated cell counter (Bio-Rad, UK). Cells were pelleted by centrifugation at 18,000 ×g for 10 min, resuspended in the digestion buffer G2 and incubated at 56°C with proteinase K (600 mAU/ml) for at least 1 hr. Soaked swabs were placed in 200 µl of the G2 buffer and incubated with proteinase K as described before. DNA was then extracted using the EZ1 DNA Tissue Kit and an EZ1 extraction robot (Qiagen) following the manufacturer's protocol. DNA from N. branchiphila was sourced from the University of Tasmania (Australia).
In addition, DNA from Atlantic salmon gill homogenates (1:10 weight/volume in G2 buffer) was extracted and used as a negative control.
A second LAMP assay to amplify a region of 295 bp on the Atlantic salmon elongation factor 1 alpha (EF1α) (NM_001123629.1) was designed as described above and used as an internal control (Table 1).

| Recombinant plasmid
A fragment of 1632 bp of the N. perurans 18S rRNA gene containing the LAMP probing region was amplified using the set of primers Generic 1F (5'-TATGGTGAATCATGATAACTTWAC-3') and B3-Z (5'-GGAATTCCTCGTTCACGATAA-3') and cloned into the pGem-T Easy plasmid vector (Promega). The template (dsDNA) copy number was calculated using a QuantiFluor dsDNA kit in a Quantus fluorimeter (Promega), and a plasmid dilution series, from 10 6 to 1 copy, was generated to obtain a standard curve.

| Assay optimization
The reaction temperature was optimized using a block gradient from 60 to 68°C at 0.1°C intervals followed by an annealing step of 98-80°C, ramping at 0.05°C per second. LAMP reactions contained 15 μL of the fast isothermal master mix (ISO-004, OptiGene), 5 pmol of each primer F3 and B3, 10 pmol of each Loop-F and Loop-R, 20 pmol of FIP and BIP, either 10 5 copies of the recombinant plasmid or 5 μl of the extracted DNA and nuclease-free water to a final volume of 25 μl.
When DNA was extracted using KOH (as described below), the isothermal Lyse'n'LAMP master mix (ISO-001LNl, OptiGene) was used instead.
Isothermal amplification was performed either in a Genie® II or a Genie® III system (OptiGene) for real-time monitoring of the LAMP amplification. The amplification ratio measured as the change of fluorescence over time and expressed as the time of positivity (Tp), and the amplicon annealing temperature were analysed using a Genie® II or a Genie® III software (OptiGene).

| Specificity and sensitivity of the LAMP assay
The specificity of the N. perurans LAMP assay was assessed against close relatives N. pemaquidensis and N. branchiphila, as well as host (Atlantic salmon) DNA.
A total of 10-fold serial dilutions of the recombinant plasmid, ranging from 10 6 to 1 copy, were used to determine the limit of detection (LOD) of the assay. Linear regression analysis between the number of copies and Tp was performed from three different independent assays.

| "Fast-and-dirty" DNA extraction protocols
Isohelix swabs were soaked in the N. perurans in vitro culture and DNA-extracted following either "fast-and-dirty" DNA extraction methods as described below or the reference laboratory method using the EZ1 DNA tissue Kit.
Five "fast-and-dirty" DNA extraction methods for POC testing were evaluated: sodium hydroxide (NaOH), QuickExtract™ DNA extraction TA B L E 1 Sequences of primers designed for the Neoparamoeba perurans 18S rRNA gene and the Atlantic salmon elongation factor 1 alpha LAMP assays and compared with the reference laboratory method. Five swabs per protocol were tested in duplicate by LAMP assay and TaqMan™ qPCR.
N. perurans-soaked swabs were placed in 475 μl of 50 mM NaOH (pH 12) lysis reagent and incubated for 10 min at 95ºC, and then, 25 μl of 100 mM of Tris-hydrochloric acid (pH 5) was added to neutralize the lysis reaction.
QuickExtract™ DNA extraction solution was used following the manufacturer's recommendations. Swabs were placed in 500 μl of the QuickExtract™ buffer and incubated for 6 min at 65°C followed by 2-min incubation at 98°C.
For the alkaline KOH protocol, swabs were placed in a tube containing 250 μl of filtered SW and 250 μl of 600 mM KOH (pH13) and incubated for 5 min at 95°C followed by a brief cooling period on ice (OptiGene, 2018).
For testing the KAPA extraction kit, swabs were placed in 472 μl of water, 25 μl of KAPA buffer and 3 μl of KAPA enzyme, and incubated for 10 min at 75°C, followed by 5-min incubation at 95°C as recommended by the manufacturers.
Finally, for the Buccalyse DNA kit, swabs were placed in 500 μL of Buccalyse and incubated for 15 min at 70°C followed by an incubation of 2 min at 95°C following the manufacturer's instructions.
The amount of the extracted DNA obtained with the different POC methods was measured using a NanoDrop ND-1000 spectrophotometer (Labtech).

| Gill swabs from naturally infected Atlantic salmon
To validate the selected POC DNA extraction method, 11 naturally AGD-infected Atlantic salmon, of approximately 200g, were collected from an open-water pen from Northern Scotland. Gill swabs were taken in duplicate from the anaesthetized fish by swabbing the gills. For each animal, one swab was DNA-extracted with the laboratory reference protocol using the EZ1 robot extraction and the second swab was DNA-extracted using the QuickExtract protocol. A visual gill index, ranging from 0 to 5, was recorded for each animal (unaffected gills: gill score 0; to severe lesions covering the majority of the gill area: gill score of 5) as the average of the 16 hemibranch surface (both sides of all 8-gill arches) scores (Taylor et al., 2009).
The identification of N. perurans from the swabs was analysed both by TaqMan™ qPCR (as described below) and by LAMP assay.

| Gill swabs from challenged Atlantic salmon
To generate AGD-positive gill swabs of Atlantic salmon with low visual gill score, an AGD bath challenge was carried out as described previously (Cano et al., 2019). Briefly, two tanks containing 45 Atlantic salmon reared in the biosecure stock aquarium areas of the Cefas Weymouth Lab from ova, weighing approximately 200g, were exposed to N. perurans by static bath immersion using either 2,500 or 500 trophozoites per L 1 for 4 hr, respectively. Then, the flow rate was restored to 5-7 L per minute and the water temperature was maintained at 12 ± 1°C. Fish were examined from the high dose at day 22 to the low dose at day 29 after challenge, and then weekly to follow the disease progression. The visual gill index was noted, and a single gill swab per animal was taken. DNA was extracted from a total of 60-gill swabs using the QuickExtract protocol, and the presence of N. perurans was analysed by LAMP assay. A positive control using either 10 5 copies of the recombinant plasmid or DNA extracted from positive AGD-infected fish was run alongside the tests. A LAMP assay to amplify the Atlantic salmon EF1α was carried out in parallel. Any test with no amplification of the host DNA was considered an invalid test.
An additional 12 swabs, taken from the specific-pathogen-free Atlantic salmon, were used as non-infected control samples. and 1 min at 60°C. Each sample was tested in duplicate.

| Result validation by TaqMan™ qPCR
In addition, the qPCR assay was used to estimate the number of copies of the 18S rRNA gene in a single amoebic cell. DNA was extracted from 10-fold serial dilutions of amoebic cells, containing 10 3 to 1 cell, and the CT values were correlated with the number of copies of the 18S rRNA gene in a standard curve.

| Ethics statement
Animal procedures were approved by the Animal Welfare and Ethical Review Body (AWERB) at the Cefas Weymouth Laboratory and conducted in compliance with the Animals (Scientific Procedures) Act 1986.

| LAMP assay optimization
Testing temperatures from 60 to 68°C resulted in the selection of the optimal amplification temperature (faster detection) of 62.9°C for the N. perurans LAMP assay. For standardization purposes, the internal control Atlantic salmon EF1α LAMP assay was run at the same temperature as the N. perurans LAMP assay. The annealing curves of the amplified products for the N. perurans LAMP assay, using either the recombinant plasmid or extracted DNA, showed a single peak in the range of 81-82°C, while the annealing temperature for the Atlantic salmon LAMP assay was 88°C (Figure 1).

| N. perurans LAMP assay specificity and analytical sensitivity
The N. perurans LAMP assay did not amplify close relatives N. pemaquidensis and N. branchiphila, or Atlantic salmon DNA from tissue homogenates (Figure 2).
Taking the average of 3 independent runs, the assay detected 10 6 copies of the recombinant plasmid under 13 min and 10 3 copies under 35 min. Dilutions containing 100 copies showed inconsistent results, with the generation of a second amplicon at a different annealing temperature than the expected at 81-82°C, probably due to the amplification of secondary structures as a result of low amounts of the target sequence or primer dimers. Either both 10 and 1 copies of the template failed to amplify the target or the amplicon showed a different annealing temperature (Figure 3). Therefore, 10 3 copies were accepted as LOD for the assay, and a test run of 40 min was established for later analysis. Linear regression analysis showed a strong correlation (Pearson's r = −.96) between the number of the recombinant plasmid copies and Tp for 10-fold dilutions between 10 6 and 100 copies (r 2 = .93) (Figure 4).
Linear regression analysis carried out on a TaqMan™ qPCR assay between the number of amoebic cells and a plasmid copy standard curve ( Figure S1) gave an estimated average number of 856 copies of the 18S rRNA gene per cell.

| Comparison of N. perurans detection in gill swabs using the reference laboratory method and POC protocol
Two swabs per animal were analysed from 11 naturally infected AGD Atlantic salmon showing a visual gross gill score between 2 and 5. The (1 out of 3) of the swabs taken from fish of gill score of 2 (Table 3).
From those two negative swabs extracted with the QuickExtract, of whom duplicate swab using the reference DNA extraction method tested positive by TaqMan™ qPCR, one of them also failed to detect the parasite by TaqMan qPCR, suggesting a poor DNA extraction.
Therefore, to allow for the identification of false negatives due to the failure of the POC DNA extraction, the Atlantic salmon EF1α LAMP assay was run in parallel as an internal control in subsequent analysis.
Although a trend of a negative correlation between the visual gill score and the LAMP assay detection was observed (shorter Tp for higher gill score), the correlation coefficient was not significant for either of the DNA extraction methods (Pearson's r = −.57 for swabs DNA-extracted with the EZ1 Biorobot and Pearson's r = −.22 with the QuickExtract method).

| N. perurans POC detection in Atlantic salmon gill swabs showing low gill score
Sixty further swabs were taken from challenged Atlantic salmon showing a visual gill score between 0.13 and 3.58. DNA was extracted solely with the QuickExtract method, and both the N. perurans 18S rRNA gene and the Atlantic salmon EF1α, used as internal control, were analysed by LAMP assay (summarized in Table 4, raw data in Table S1).
An invalid test, failure to detect the Atlantic salmon EF1α gene,

| D ISCUSS I ON
An N. perurans LAMP assay has been evaluated for its application as a POC diagnostic test. Current industry practices rely on the visual gill score to plan cost-effective treatments; however, a confirmatory test for the presence of the parasite is recommended before any treatment is given to the animals (Taylor et al., 2009 that of a conventional N. perurans diagnostic PCR, which can detect a LOD of 0.5 amoebae, equivalent to 2 pg of extracted DNA (Young et al., 2008). The N. perurans LAMP assay can thus be used to confirm the presence of the parasite in clinically infected fish. However, the analytical sensitivity of this assay is lower than expected for a real-time fluorescence LAMP assay, which is typically comparable to a qPCR assay (McKenna et al., 2011). Published N. perurans qPCR assays reported a LOD within 3-100 copies of the 18S rRNA gene depending on the assay (Downes et al., 2017). In the present LAMP assay, the detection of a lower number of copies (˂100) of the N. perurans 16S rRNA  (Kuiper et al., 2006), Acanthamoeba has approximately 600 copies, and Naegleria species seem to have several thousand copies (Qvarnstrom, Visvesvara, Sriram, & Da Silva, 2006).
Thus, given the LOD of the LAMP assay, this test can potentially detect on-site the presence of one amoeba in the extracted DNA.
To develop the LAMP assay as a POC test, five "fast-and-dirty" DNA extraction methods, three commercial kits and two housemade buffers, were tested and compared with a reference laboratory DNA extraction method using positive N. perurans swabs.
The robustness, compatibility with TaqMan™ qPCR chemistry for laboratory validation, simplicity and rapidity in the LAMP assay detection were compared across the five POC methods. The QuickExtract™ DNA extraction solution produced better results.
This DNA extraction method has been used previously for the molecular characterization of amoebae of the genus Flamella and other aquatic microorganisms (Kotov & Taylor, 2010;Walthall, Tice, & Brown, 2016 (x)) and the time of positivity (Tp). Mean data of three independent assays future be implemented in commercial kits for the detection of N. perurans in the LAMP assay.
In developing a POC diagnostic kit, it is important to consider the presence of inhibitors during the DNA extraction method, which can then affect the test outcome (Ali et al., 2017). To detect invalid tests, an internal control based on the LAMP amplification of the host EF1α gene was run alongside the N. perurans test. Internal controls are a routine practice to identify invalid tests in LAMP assays (Nurul Najian, Engku Nur Syafirah, Ismail, Mohamed, & Yean, 2016).
From those gill swabs analysed, only 1.6% of the tests were considered invalid. In the present study, the LAMP detection of the internal control was analysed in a separate tube than the LAMP detection of the pathogen. Due to the low ratio of the parasite DNA versus host DNA in the samples, the LAMP multiplexing for those genes was not successful (data not shown). Therefore, a separate assay (but run in parallel) for the target gene and the internal control was used for the analysis of the field samples.
Gill swabs are a resourceful non-lethal sampling tool for AGD in Atlantic salmon. In the present study, a trend of negative correlation between the visual gill score and the Tp for the LAMP assay detection was observed. In previous studies, the gill score has shown a good correlation with histopathology scores and the molecular detection of N. perurans when using TaqMan™ qPCR assays (Downes et al., 2017 Note: N. perurans detection was assayed by TaqMan™ qPCR (expressed as cycle threshold Ct values) and LAMP assay (expressed as time of positivity Tp (mm:ss)). Visual gill index expressed as the average of the 16 hemibranch scores. Number of positive samples (+ve) expressed as: number of positive swabs/total number of swabs analysed per gill score.

TA B L E 3 Comparison of
Neoparamoeba perurans detection in gill swabs from naturally infected Atlantic salmon using a reference laboratory method (EZ1 Biorobot) and a point-of-care DNA extraction protocol (QuickExtract) its poor performance in the field could be explained due an inefficient POC DNA extraction method. Therefore, future work should focus on improving POC DNA extraction methods to allow for the field deployment of this AGD-POC test.

| CON CLUS IONS
Overall, the simplicity, performance and low cost of the present AGD-LAMP assay, with a gross estimated cost of £2.5 per sample (including the cost for DNA extraction and the isothermal amplification) versus £6.8 for a TaqMan™ qPCR assay, make this test a good candidate for the on-site confirmation of N. perurans in non-lethal gill swabs taken from Atlantic salmon clinically infected with AGD; however, further work is required to improve the parasite detection in low gill scores.

ACK N OWLED G M ENT
This project has received funding from the UK Department for This output reflects only the author's view, and the European Union cannot be held responsible for any use that may be made of the information contained therein.
The authors want to thank Dr Barbara Nowak (University of Tasmania) for kindly sourcing N. branchiphila DNA.

F I G U R E 6
Summary of the number of Atlantic salmon gill swabs analysed (bars) and the percentage of LAMP-positive tests (line) per gill score range. Bars include swabs taken both from naturally infected and from challenged fish 3.1-3.6 29:05 ± 4:20 6/6 16:45 ± 1:40 6/6 Note: DNA was extracted from non-lethal gill swabs using the QuickExtract method. The Atlantic salmon elongation factor 1a LAMP assay was used as an internal control. LAMP detection expressed as time of positivity Tp (mm:ss). The visual gill index expressed as the average of the 16 hemibranch scores. Number of positive samples (+ve) expressed as: number of positive swabs/total number of swabs analysed per gill score range.
TA B L E 4 Point-of-care identification of Neoparamoeba perurans by LAMP assay