A real-time PCR tool for the surveillance of zoonotic Onchocerca lupi in dogs, cats and potential vectors

The ocular onchocercosis is caused by the zoonotic parasite Onchocerca lupi (Spirurida: Onchocercidae). A major hindrance to scientific progress is the absence of a reliable diagnostic test in affected individuals. Microscopic examination of skin snip sediments and the identification of adults embedded in ocular nodules are seldom performed and labour-intensive. A quantitative real-time PCR (qPCR) assay was herein standardized for the detection of O. lupi DNA and the results compared with microscopic examination and conventional PCR (cPCR). The specificity of qPCR and cPCR was assessed by processing the most common filarial nematodes infecting dogs, skin samples from O. lupi infected (n = 35 dogs) or uninfected animals (n = 21 dogs; n = 152 cats) and specimens of potential insect vector (n = 93 blackflies; n = 59 mosquitoes/midges). The analytical sensitivity of both assays was assessed using 10-fold serial dilutions of DNA from adult specimen and from a pool of microfilariae. The qPCR on skin samples revealed an analytical specificity of 100% and a sensitivity up to 8 x 10−1 fg/2μl O. lupi adult-DNA and up to 3.6 x 10−1 pg/2μl of mfs-DNA (corresponding to 1 x 10−2 mfs/2μl). Only 9.5% O. lupi-infected skin samples were positive for cPCR with a sensitivity of 8 x 10−1 pg/2μl of DNA. Out of 152 blackflies and mosquitoes/midges, eight specimens experimentally infected (n = 1 S. erythrocephalum; n = 1 S. ornatum; n = 6 Simulium sp.) were positive by qPCR. The qPCR assay herein standardized represents an important step forward in the diagnosis of zoonotic onchocercosis caused by O. lupi, especially for the detection and quantification of low number of mfs. This assay provides a fundamental contribution for the establishment of surveillance strategies aiming at assessing the presence of O. lupi in carnivores and in insect species acting as potential intermediate hosts. The O. lupi qPCR assay will enable disease progress monitoring as well as the diagnosis of apparently clinical healthy dogs and cats.


Introduction
Within the genus Onchocerca (Spirurida: Onchocercidae), Onchocerca volvulus and Onchocerca lupi parasitize humans and carnivores, respectively [1][2][3][4][5], the latter being a zoonotic agent [6,7]. While O. volvulus is a well-known parasite of humans transmitted by blackflies (Simulium spp.) [8,9], the epidemiology of O. lupi is far from being understood, particularly because the information about insect species acting as vectors is lacking. Only Simulium tribulatum was suggested as the putative vector of this filarial worm in California (USA), but proof of its intermediate host competence is currently absent [10]. Onchocerca lupi belongs to the spirurids in the Nematode clade III [11] was first detected from a Caucasian wolf (Canis lupus) in Georgia [12], and, only recently, diagnosed in dogs and cats from Europe (Greece, Portugal, Spain, Germany, Hungary) and USA [13][14][15][16][17][18][19][20]. The reports of O. lupi infection are mainly based on the presence of ocular nodules on the eyelids, conjunctiva, and sclera [3,21,22], though the localization of adult worms in the retrobulbar area of the canine patients may impair the assessment of its distribution in endemic areas [23]. The detection of microfilariae (mfs) in skin snip sediments is the only available tool for the diagnosis of the infection when nodules are not apparent in the eyes. The retrieval and identification of mfs in skin snip samples is a rather invasive and time-consuming method, highly dependent on the anatomical location of skin biopsy and mfs density [24]. Again, the detection of mfs may depend upon the prepatent period, previous microfilaricidal treatments, and on the operator's skills in examining skin sediments, as described for O. volvulus [25,26].
Conventional PCR (cPCR) amplification and sequencing of mitochondrial NADH dehydrogenase subunit 5 (ND5) and cytochrome c oxidase subunit 1 (cox1) genes are available for the molecular identification of O. lupi adults and mfs [7,27,28]. The cPCR, however, may be relatively labour-intensive and exhibit low sensitivity, mainly for mfs detection, limiting the establishment of large-scale epidemiological studies in vertebrate hosts and putative vectors.
Here, we developed a quantitative real-time PCR (qPCR) assay based on the hybridization probe to detect O. lupi DNA in host and putative vector samples. The diagnostic validity of qPCR assay was compared with microscopic examination and cPCR methods.

Ethics statement
All dogs' and cats' skin samples were collected in previous studies [17,29] and approved by the ethical committee of the Department of Veterinary Medicine of the University of Bari (Prot.

Samples
Genomic DNA of adult specimens of O. lupi (n = 3), as well as DNA from single (n = 7) or pooled mfs (n = 10), collected from dogs in different geographical locations (Table 1) were used as control. All specimens were previously identified based on morphological and molecular analyses [18,30].

Primers and probe design and qPCR run protocol
and a TaqMan-MGB hydrolysis probe (FAM-5'-CTTAGAGTAGAGGGTCAG CC-3'-non-fluorescent quencher-MGB; Applied Biosystems; Foster City, CA, USA), targeting partial cox1 gene (90bp), were designed by alignment of sequences from a wide range of closely related filarial nematodes available from GenBank database ( Table 2), using Primer Express 2.0 (Applied Biosystems, Foster City, CA). Specificity of the primers and probe for O. lupi were confirmed in silico using the basic local alignment search tool (BLAST, GenBank, NCBI).
The run protocol consisted of a hot-start at 95˚C for 3 min, and 40 cycles of denaturation (95˚C for 10 sec) and annealing-extension (64˚C for 30 sec). All assays were carried out in duplicate and a no-template control was included in each run. The qPCR was performed in a

Specificity and sensitivity of qPCR
To investigate the analytical specificity of the assay, genomic samples of Onchocerca spp. and of the most common filarial nematodes infesting dogs (Table 1) were used. The specificity of the assay was tested by using DNA from skin samples of naturally infected dogs, which were positive for O. lupi (n = 35) at microscopic examination [29]. Skin samples were divided in five groups (G1-G5) according to their mfs load (Table 3), being 14 also co-infected with Cercopithifilaria bainae and Cercopithifilaria sp. II. Skin samples (dogs n = 21; cats n = 152), which did not test positive to any mfs [17,29], were used as negative control. Specimens of blackflies (n = 66) and mosquitoes/midges (n = 39) collected from 2011 to 2014 in Greece [31], and 27 blackflies and 20 Aedes albopictus (colony specimens) experimentally infected by intrathoracic microinjection with mfs of O. lupi (parasitic load of 20mfs/μl) were analyzed after death (i.e., from one to 10 days post infection) ( Table 4).
The analytical sensitivity of the qPCR assay was assessed using 10-fold serial dilutions of DNA from adult specimen (i.e., ranging from 8 × 10 4 to 8 × 10 −3 fg/2μl of reaction) and from a pool of 10 mfs (i.e., ranging from 10 to 10 × 10 −3 microfilariae/2μl of reaction, corresponding to 3.6 ×10 −1 ng/2μl to 3.6 ×10 1 fg/2μl of DNA). Ten replicates of each serial dilution were submitted to the same run for assessment of intra-assay reproducibility. Genomic DNA was isolated from all skin samples and from O. lupi adults and mfs, blackflies, mosquitoes and midges specimens using the commercial kits DNeasy Blood & Tissue Kit (Qiagen, GmbH, Hilden, Germany), respectively, following the manufacturers' instructions. The amounts of purified DNA were determined spectrophotometrically using the Qubit (Applied Biosystems, Foster City, CA, USA).
All cPCR products were resolved in 0.5x GelRed stained (Biotium, CA, USA) agarose gels (2%), purified using enzymatic purification (Exo I-FastAP; Thermo Fisher Scientific, MA, USA) and sequenced in an automated sequencer (3130 Genetic Analyzer). All sequences generated were compared with those available in GenBank using Basic Local Alignment Search Tool (BLAST) [33].

Results
All O. lupi naturally-infected dog positive at skin samples examination by microscopy, considered the gold standard method as true positives, were positive by the O. lupi qPCR herein assessed (specificity of 100%). Out of 21 skin samples microscopically and qPCR positive for O. lupi, two were positive by cPCR (parasite load of 8 and 25 mfs), revealing a low analytical cPCR specificity (i.e., 9.5%). None of cat's skin samples were positive by qPCR.
A specific fluorescent signal was recorded for all O. lupi adult and mfs positive controls tested (Fig 1). No fluorescence was obtained for all other Onchocerca species or filarial nematodes examined as well as for skin samples used as negative control.
The analytical sensitivity of qPCR was confirmed by detection of up to 8 x 10 −1 fg/2μl and 3.6 x 10 −1 pg/2μl of DNA (i.e., corresponding to 1 x 10 −2 mfs/2μl) of O. lupi adult worm and mfs, respectively (Fig 2A and 2B). qPCR efficiencies ranged from 108.7 to 115.3% with an R 2 from 0.996 to 0.999 and Slope from -3.003 to -3.131, for both adult and mfs (Fig 2A and 2B). The mean parasite load detected for the positive skin samples, ranged from 1.9 to 96 mfs/2μl of reaction, corresponding to 6.1 x 10 −2 ng/2μl (mean cycle threshold of 33.49) and to 3.4 ng/2μl DNA (mean cycle threshold of 27.52), respectively ( Table 3). The results of mfs detection by qPCR overlapped the values obtained by the microscopic examination. The detection limit registered for cPCR was up to 8 x 10 −1 pg/2μl for adult worms and up to 3.6 x 10 1 pg/2μl for mfs DNA (i.e., corresponding to 1 mf/2μl), respectively (Fig 3).
Sequences derived from all amplicons of cPCR matched with 99-100% nucleotide identity appropriate reference sequences of O. lupi available from GenBank (accession numbers KC686702, KC686701).

Discussion
A qPCR assay has been developed for the detection of O. lupi in animal skin snip samples and potential vectors and proved to be a sensitive and specific tool for the diagnosis of this parasite, with a mean detection limit as low as 1.9 mfs per reaction. In addition, the high sensitivity of the qPCR protocol has been demonstrated by detecting a small amount of DNA (up to 8 x 10 −1 fg/2μl for adult and up to 3.6 x 10 −1 pg/2μl for mfs), by the slope value of standard curve (−3.131), the efficiency (115.3%) and the coefficient of determination (R 2 = 0.999). These features of the assay are due to the selection of a stable hydrolysis probe designed (100% specific for O. lupi DNA), as well as to the choice of the target gene used. Indeed, cox1 gene of the mitochondrial DNA has been well recognized as a "barcode" for filarial nematodes [34], with a high amplification efficiency, also due to the large copy numbers enabling the detection of minimum amounts of DNA [35][36][37]. Though few Onchocerca species DNA were herein tested, which may represent a limitation of the qPCR assay, this new tool provides an alternative to the labor intensive microscopic examination of skin snip samples and to cPCR for the diagnosis of O. lupi [38]. The qPCR assay was highly specific in revealing O. lupi DNA both in coinfected samples from dogs as well as in potential vector species, avoiding the sequencing confirmation needed using cPCR with filarioid generic primers [32]. Overall, the positive fluorescent signal from samples of O. lupi, from different geographical areas (i.e., Europe and USA), which displayed genetic intraspecific variability [18], indicates the usefulness of the qPCR also for the surveillance of O. lupi where the parasite has been reported [13,14,16,17,19,[39][40][41]. Similarly, even if the qPCR cannot discriminate between viable and nonviable parasites or immature and infective larvae, the assay could be useful for detecting O. lupi in blackfly, mosquito and/or midge species, potentially involved in the transmission of this parasite. Indeed, the specificity of the qPCR to amplify exclusively the DNA of the pathogen in potential insect vectors herein tested, may ultimately assist in the quest to identify the elusive vector of O. lupi.
The newly designed assay represents an improvement in the diagnosis of onchocercosis, by the detection and quantification of low mf densities from tissue samples and could provide a contribution to disease progress monitoring and to the surveillance of O. lupi-infected dogs, avoiding the introduction and/or spread of this life-threatening parasitic nematode, as well as to the identification of apparently healthy animals [29,42].
The qPCR may speed-up time of diagnosis and prompt treatments of infected animals, which may avoid the appearance of nodular lesions in the eyes or in other anatomical localizations [43].
A TaqMan-based specific and sensitive assay without sequencing is expected to assist highthroughput analysis of samples, eventually leading to improve disease monitoring under the frame of a Public Health perspective. This would be particularly relevant considering that, since its first description of its zoonotic potential [7], cases of zoonotic onchocercosis are being detected increasingly in people from Europe, Iran and the USA [44][45][46][47].