Laidoudi et al. Parasites Vectors (2020) 13:319 https://doi.org/10.1186/s13071-020-04185-0 Parasites & Vectors

RESEARCH Open Access Development of a multiplex qPCR‑based approach for the diagnosis of Diroflaria immitis, D. repens and Acanthocheilonema reconditum Younes Laidoudi1,2, Bernard Davoust1,2, Marie Varloud3, El Hadji Amadou Niang1,2, Florence Fenollar2,4 and Oleg Mediannikov1,2*

Abstract Background: Diroflaria immitis, D. repens and Acanthocheilonema reconditum are the main causative agents of zoonotic canine flariosis. Methods: We developed a combined multiplex approach for flaria and Wolbachia detection using the 28S-based pan-flarial and 16S-based pan-Wolbachia qPCRs, respectively, involving a fast typing method of positive samples using triplex qPCR targeting A. reconditum, D. immitis and D. repens, and a duplex qPCR targeting Wolbachia of D. immitis and D. repens. The approach was complemented by a duplex qPCR for the diferential diagnosis of heart- worms (D. immitis and Angiostrongylus vasorum) and pan-flarial cox1 and pan-Wolbachia ftsZ PCRs to identify other flarial parasites and their Wolbachia, respectively. A total of 168 canine blood and sera samples were used to validate the approach. Spearmanʼs correlation was used to assess the association between flarial species and the strain of Wolbachia. Positive samples for both the heartworm antigen-test after heating sera and at least one DNA-positive for D. immitis and its Wolbachia were considered true positive for heartworm infection. Indeed, the presence of D. repens DNA or that of its Wolbachia as well as A. reconditum DNA indicates true positive infections. 1 4 Results: The detection limit for Wolbachia and flariae qPCRs ranged from 5 10− to 1.5 10− mf/ml of blood. When tested on clinical samples, 29.2% (49/168) tested positive for flariae or× Wolbachia DNA.× Filarial species and Wolbachia genotypes were identifed by the combined multiplex approach from all positive samples. Each species of Diroflaria was signifcantly associated with a specifc genotype of Wolbachia. Compared to the true positives, the approach showed excellent agreement (k 0.98–1). Unlike D. immitis DNA, no A. vasorum DNA was detected by the duplex qPCR. The immunochromatographic= test for heartworm antigen showed a substantial (k 0.6) and a weak (k 0.15) agreements before and after thermal pre-treatment of sera, respectively. = = Conclusions: The proposed approach is a reliable tool for the exploration and diagnosis of occult and non-occult canine flariosis. The current diagnosis of heartworm disease based on antigen detection should always be confrmed by qPCR essays. Sera heat pre-treatment is not efective and strongly discouraged. Keywords: Canine flariosis, Multiplex qPCRs, Diferential diagnosis, Heartworm disease, Diroflaria immitis

Background Canine flariosis includes diseases caused by parasitic *Correspondence: [email protected] nematodes called flariae, belonging to the order Spiru- 1 Microbes, Evolution, Phylogeny and Infection (MEPHI), UMR Aix- Marseille University, IRD, APHM, IHU Méditerranée Infection, 19–21, Bd rida. Tere are several species of veterinary and human Jean Moulin, 13005 Marseille, France importance. Dogs seem to be the natural hosts for sev- Full list of author information is available at the end of the article eral species, such as Diroflaria immitis, D. repens,

© The Author(s) 2020. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativeco​ ​ mmons.org/licen​ ses/by/4.0/​ . The Creative Commons Public Domain Dedication waiver (http://creativeco​ mmons​ .org/publi​ cdoma​ in/​ zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Laidoudi et al. Parasites Vectors (2020) 13:319 Page 2 of 15

Acanthocheilonema reconditum, A. dracunculoides, even more challenging. Te association of the heartworm Cercopithiflaria grassii, Brugia ceylonensis, B. patei, B. with D. repens infection may result in an unexplained malayi, B. pahangi, Onchocerca lupi and Telazia calli- suppressive efect on the production of microflariae of paeda [1–4]. Tese arthropod-borne flarioids produce D. immitis [18, 19]. In such cases, the cross-reactivity blood, cutaneous or mucous microflariae, where they are between D. immitis and D. repens may result in misdi- available to arthropod vectors [5]. Te most common and agnosis. Terefore, there is an urgent need for a more medically important species afecting dogs are D. immi- sensitive diagnostic method to detect occult as well as tis, D. repens and A. reconditum [6]. In addition to their non-occult canine flariosis and to identify the pathogen. veterinary importance, they can also afect human health. Detection of D. immitis has gained more and more atten- Diroflaria immitis causes pulmonary and cardiopul- tion; many trials have been performed for improving the monary diroflariosis in humans and dogs, respectively. quality of heartworm diagnostic tools, such as the detec- Cardiopulmonary diroflariosis usually called heartworm tion of a specifc antigen released by these worms [20], or disease has recently been considered as an emerging the use of a recombined antigen of D. immitis for specifc disease in Europe. In the USA, D. immitis is the most antibody detection [21]. important life-threatening parasitic infection in dogs Te endosymbiotic intracellular bacteria of the genus [7]. Elsewhere in the world, particularly in eastern Euro- Wolbachia are associated with some flarial species of two pean countries, D. repens is the most endemic parasitic subfamilies of the Onchocercidae: Onchocercinae and nematode causing subcutaneous infection, which is less Diroflariinae [22]. Tese bacteria are host-specifc, and virulent but more zoonotic than that caused by D. immi- each species of flarial worm is associated with a specifc tis [8]. Acanthocheilonema reconditum is an occasional bacterial genotype. Wolbachia spp. have been targeted zoonotic agent that afects the subcutaneous tissue and for the indirect diagnosis of D. immitis infection in dead- the perirenal fat [9, 10] causing a common but clinically end hosts such as humans and cats, whereby the strong less important infection in dogs [11]. reaction of the host against the parasite prevents them to Once mature, these flarioids can produce microflariae achieving their maturation, and, therefore, the produc- circulating in the bloodstream. Tis larval stage (L1) is tion of microflariae may not be achieved [23]. In such also a target for the diagnosis by microscopic detection of cases, the detection of flaria-specifc Wolbachia may the larvae or by detection of their DNA in the host blood indicate a flarial infection and can serve as an alterna- [10]. Diroflaria immitis, the agent of heartworm dis- tive diagnostic tool in endemic areas [23, 24]. In around ease, is distributed worldwide and is responsible for heart 40–60% of canine heartworm cases, both Wolbachia and failure in dogs after colonization of pulmonary arteries parasite DNA may be detected using conventional PCR and the right ventricle, where it can be fatal if untreated. [25–27]. Indeed, the combined detection of Wolbachia Due to the gravity of the disease, it remains the most and Diroflaria DNA was suggested to improve heart- commonly diagnosed flariosis in dogs due to the detec- worm detection [27]. tion of antigen circulating in the blood [11–13]. Several In the present study, we developed a multiplex real- problems with the current diagnostic methods have been time PCR-based approach allowing a specifc, rapid and raised, such as morphological confusion between the simultaneous detection of D. immitis, D. repens and A. microflariae of D. immitis, A. reconditum and D. repens. reconditum as well as the occult Diroflaria spp. infec- Commercially available diagnostic kits for the detec- tions in dogs. In addition, it was completed by a duplex tion of D. immitis antigens may also cross-react with real-time PCR-based assay for the simultaneous detec- flarial and non-flarial nematodes, such as D. repens, A. tion of D. immitis and A. vasorum as a diferential diag- reconditum and Onchocerca spp. [14], Spirocerca lupi and nostic for canine heartworms. Te approach can be used Angiostrongylus vasorum, especially the latter which can in routinely in a diagnostic laboratory. We also evaluated cause a cross-reaction without prior heat pre-treatment the efectiveness of a novel molecular approach to con- of the sera [15–17]. Angiostrongylus vasorum, the agent ventional serological diagnosis and assessed the impor- of French heartworm disease, should also be taken into tance of serum heating. account in the diferential diagnosis of pulmonary disease [17]. Tis so-called occult heartworm is characterized Methods by the absence of microflaremia or an amicroflaremia. Probes, primers design and PCR amplifcation protocol Tis may result from the hostʼs immune response, low Custom protocol and in silico validation parasite load and infertility or, incidentally, the microfla- First, for each PCR assay, the target gene was chosen to ricidal efect observed in dogs receiving macrocyclic lac- meet the objective of each system. Fasta fles were con- tone prevention [14]. When the occult heartworm occurs structed from the sequences of the representative mem- in a co-infection with another flariosis, the diagnosis is bers of the family Onchocercidae or Wolbachia genotypes Laidoudi et al. Parasites Vectors (2020) 13:319 Page 3 of 15

available in the GenBank database. Te sequences were diferent dyes were used for specifc detection: FAM and aligned using BioEdit v 7.0.5.3 software [28] to reveal the VIC with a non-fuorescent quencher-TAMRA confned highly conserved inter- and intra-species regions as tar- to D. immitis and D. repens probes, respectively; Cyanine get regions for primers and probes. Tis region was sub- 5 (Cy5) with a non-fuorescent quencher-BHQ-3 for the mitted to Primer3 online software v. 0.4.0 (http://prime​ A. reconditum probe (Table 1). r3.ut.ee), in order to determine valuable candidate prim- ers and probes; the selection was based on the criteria for TaqMan duplex qPCR targeting D. immitis and A. vasorum primer and probe design. Te duplex cox1-based qPCR was designed (Table 1) with Physicochemical characteristics, annealing temperature primers Hw.COI.723-F and Hw.COI.950-R to amplify and the possibility for hairpin, self- and hetero-dimers partial cox1 gene (227 bp) of both flarial and non-flarial were tested using free online software Oligo-Analyzer nematodes, including D. immitis and A. vasorum. Te 3.1 [29]. Primer sets and probes were also checked within primers were chosen to fank the probe P.imm.COI.777, DNA databases of metazoans (taxid:33208), vertebrates previously designed for D. immitis. In addition, we (taxid:7742), bacteria (taxid:2), Canidae (taxid:9608), designed a new probe named A.vas.COI.813-P specifc to Felidae (taxid:9682) and humans (taxid:9605) using A. vasorum. Te TaqMan probes were labelled with FAM primer-BLAST [30]. Tis was completed for all possible and VIC, respectively, with a non-fuorescent quencher forward-reverse and probe-reverse combinations of each TAMRA. PCR system. Primers were synthesized by Eurogentec (Liège, Belgium) and the hydrolysis probe was synthe- TaqMan simplex qPCR targeting Wolbachia sized by Applied ­BiosystemsTM (Foster City, CA, USA). Te 16S rDNA gene has been reported as the most com- monly used gene for Wolbachia phylogeny [35]. Te TaqMan simplex qPCR targeting flarial nematodes simplex-qPCR was developed and validated in silico for Te choice of the large subunit rRNA (LSU) gene, also the conserved region of the frst third of the 16S rDNA called 28S gene, was based on several criteria such as: gene. Te qPCR system (Table 1) is composed of prim- the tandem repetition of about 150 times in the flarial ers Wol.16S.301f and Wol.16S.478r with the probe nematode genome, which improves the PCR detectabil- Wol.16S.347p targeting all Wolbachia lineages. ity [31]; availability on GenBank for representatives of all nematode families; and sharing a highly conserved region TaqMan duplex qPCR targeting flarial Wolbachia within the Onchocercidae. Te primers qFil-28S-F, qFil- Wolbachia ftsZ gene, the homologue of the eukary- 28S-R and a TaqMan® hydrolysis probe (qFil-28S-P) were otic protein tubulin, provides sufcient discrimination designed to amplify 28S gene for most flarial species between Wolbachia spp. of supergroups C and D found (Table 1). in flarial nematodes, and those of supergroups A and B found in arthropods with a higher divergence between TaqMan triplex qPCR targeting D. immitis, D. repens and A. flarial Wolbachia of supergroups C and D [36]. Te reconditum ftsZ-based duplex-qPCR was designed with primers Te gene encoding for the cytochrome c oxidase subu- WDiro.ftsZ.490f and wDiro.ftsZ.600r targeting flarial nit 1 gene (cox1) was selected for the development of Wolbachia belonging to supergroup C, which includes the triplex TaqMan qPCR system targeting D. immitis, those found in Diroflaria sp. However, the specifcity D. repens and A. reconditum (Table 1). Tis choice was of the duplex-qPCR was confned to probes wDimm. based on the availability of cox1 for the three species ftsZ.523p and wDrep.ftsZ.525p specifc to Wolbachia on GenBank. Indeed, the cox1 gene is recognized for sp. of D. immitis and that of D. repens, respectively its high sensitivity (a high copy number relative to the (Table 1). nuclear gene in each cell) [32]. Te cox1 gene has been described as a “barcode gene” for flarial nematodes [33]. Te primers Fil.COI.749 and dg.Fil.COI.914 (Table1) Run protocols were designed to amplify a 166 bp-long cox1 fragment Te simplex, duplex and triplex qPCR reactions were for most members of the Onchocercidae. Te system’s carried out in a fnal volume of 20 µl, containing 5 µl specifcity was confned to the TaqMan probes, namely of DNA template, 10 µl (2×) of Master Mix Roche P.imm.COI.777 specifc to D. immitis and P.rep.COI.871 (Eurogentec, Liège, Belgium). Volume of each primer specifc to both D. repens and “Candidatus Diroflaria per reaction was 0.5 µl (50 µM) for the simplex qPCR (Nochtiella) honkongensis” afecting dogs and humans and 0.75 µl (50 µM) for both the duplex and triplex in Japan [34]. Finally, the probe P.rec.COI.866 is specifc qPCR, with 0.5 µl of both UDG (1 U/µl) and each to A. reconditum. In the triplex TaqMan system, three probe (20 µM). Te fnal volume was made up to 20 µl Laidoudi et al. Parasites Vectors (2020) 13:319 Page 4 of 15

Table 1 Primers and probes developed in this study

System name Target gene Primer and probe name Sequence (5′-3′) Assay specifcity

Pan-fl 28S qPCR-based system 28S rRNA qFil-28S-F TTG​TTT​GAG​ATT​GCA​GCC​CA Filariae qFil-28S-P 6FAM-CAA​GTA​CCG​TGA​GGG​AAA​GT-TAMRA qFil-28S-R GTT​TCC​ATC​TCA​GCG​GTT​TC All-Wol 16S qPCR-based system 16S rRNA all.Wol.16S.301-F TGG​AAC​TGA​GAT​ACG​GTC​CAG​ Wolbachia all.Wol.16S.347-P 6FAM-AAT​ATT​GGA​CAA​TGG​GCG​AA-TAMRA all.Wol.16S.478-R GCA​CGG​AGT​TAG​CCA​GGA​CT Triplex TaqMan cox1 qPCR-based cox1 Fil.COI.749-F CATCCT​ GAG​ GTT​ TAT​ GTT​ ATT​ ATT​ TT​ D. immitis, D. repens and system D.imm.COI.777-P 6FAM-CGG​TGT​TTG​GGA​TTG​TTA​GTG-TAMRA A.reconditum D.rep.COI.871-P 6VIC-TGC​TGT​TTT​AGG​TAC​TTC​TGT​TTG​AG- TAMRA A.rec.COI.866-P Cy5-TGA​ATT​GCT​GTA​CTG​GGA​ACT-BHQ3 Fil.COI.914-R CWG​TAT​ACA​TAT​GAT​GRC​CYCA​ Duplex Wol-Diro ftsZ qPCR-based ftsZ WDiro.ftsZ.490-F AAG​CCA​TTT​RGC​TTY​GAA​GGTG​ Endosymbiotic Wol- system WDimm.ftsZ.523-P 6FAM-CGT​ATT​GCA​GAG​CTC​GGA​TTA-TAMRA bachia of D. immitis and D. repens WDrep.ftsZ.525-P 6VIC-CAT​TGC​AGA​ACT​GGG​ACT​GG-TAMRA WDiro.ftsZ.600-R AAA​CAA​GTT​TTG​RTT​TGG​AAT​AAC​AAT​ Duplex HWs cox1 qPCR-based cox1 Hw.COI.723-F TCAGCA​ TTT​ GTT​ TTG​ GTT​ TTT​ ​ D. immitis and A. vasorum system D.imm.COI.777-P 6FAM-CGG​TGT​TTG​GGA​TTG​TTA​GTG-TAMRA A.vas.COI.813-P 6VIC-TGA​CTG​GGA​AGA​AGG​AGG​TG-TAMRA Hw.COI.950-R GCASTAA​AAT​AAG​YAC​GAG​WAT​C

using DNAse-RNAse free UltraPure water (Eurogen- of Wolbachia lineages that may be associated with flari- tec, Liège, Belgium). Te TaqMan cycling conditions oids. All PCR reactions were carried out in a total vol- included two hold steps at 50 °C for 2 min followed by ume of 50 µl, consisting of 25 µl (2×) of AmpliTaq Gold 15 min at 95 °C, and 39 cycles of two steps at 95 °C for master mix, 18 µl of DNAse-RNAse free UltraPure water 30 s and 60 °C for 30 s. Tese reactions were performed (Eurogentec, Liège, Belgium), 1 µl of each primer (20 µM) in a CFX96 Touch Real-time PCR Detection System and 5 µl of DNA template (except no-template controls). (Bio-Rad, Marnes-la-Coquette, France) after activating Te thermal cycling conditions were: incubation step at the appropriate absorption channels for the dyes used 95 °C for 15 min, followed by 40 cycles at 95 °C for 1 min, in each qPCR system. 30 s at the annealing temperature (with a diferent melt- Te accumulation of the relative fuorescence units ing temperature for each PCR assay, see Table 2), 72 °C (RFUs) was recorded during the extension step of each for 45 s for elongation, followed by a fnal extension step qPCR and was used to set-up the cut-of value for each at 72 °C for 5 min (Table 1). PCR amplifcation was per- TaqMan system, according to the formula described formed in a Peltier PTC-200 thermal cycler (MJ Research [37]. Te tolerance value was fxed at 5% for all systems. Inc., Watertown, MA, USA). Amplicons were purifed Te qPCR reaction was considered positive only if the using the flter plate Millipore NucleoFast 96 PCR kit fol- RFU value was higher than the cut-of value. qPCR data lowing the manufacturer’s recommendations (Macherey analysis was performed using the CFX Manager Soft- Nagel, Düren, Germany). Purifed DNA was sequenced ware Version 3 [37]. using the BigDye® terminator v3.3 cycle sequencing kit DNA in line with the manufacturer’s instructions Design of conventional PCR primers, amplifcation (Applied Biosystems). Sequencing was performed using and sequencing protocols 3500xL Genetic Analyzer (Applied Biosystems, Foster In order to complete the molecular identifcation of City, CA, USA) and a capillary electrophoresis fragment flariae and their Wolbachia spp., we designed two sets analyzer (Applied Biosystems). Te nucleotide sequences of degenerate primers (Table 2): (i) Fspec.COI.957f and were assembled and edited using ChromasPro 2.0.0 and Fspec.COI.1465r targeting a 509-bp fragment of the were then checked using the Basic Local Alignment cox1 gene of flarioids; and (ii) Wol.ftsZ.363.f and Wol. Search Tool (BLAST) [38]. ftsZ.958.r targeting a 595-bp fragment of the ftsZ gene Laidoudi et al. Parasites Vectors (2020) 13:319 Page 5 of 15

Specifcity, sensitivity and system validation respectively. Finally, a serial 10-fold dilution of DNA DNA samples summarized in Additional fle 1: Table S1 extracted from microflaremic blood (Corsica, 2018) were used for the in vitro validation of all PCR systems as containing 4033 microflariae of D. immitis was used to follows. Nineteen samples of genomic DNA from flarial assess the analytical sensitivity of both the triplex and parasites were used to validate the pan-flarial 28S qPCR. duplex (Wol-Diro ftsZ) qPCRs in the direct and the indi- DNA from eight strains of Wolbachia endosymbionts of rect detection of single infection with D. immitis. Aedes albopictus, Anopheles gambiae, Cimex lectularius, C. hemipterus (PL13 strain), D. immitis microflariae, D. PCR tools validation by sample screening repens, Onchocerca lupi, Wuchereria bancrofti and Bru- and identifcation of flarial infection on biological samples gia sp. were used for the Wolbachia 16S-based qPCR A pre-existing collection of canine blood and serum sam- system. Diroflaria immitis, D. repens and A. reconditum ples was used in this study. Tis included: (i) 8 samples DNA were used to validate the triplex qPCR. Diroflaria composed of nematode-free laboratory Beagles from the immitis and A. vasorum DNA were used to validate the biobank of the Veterinary Research Center of the IHU duplex qPCR targeting heartworms. Finally, DNA of Wol- Méditerranée Infection were used as a negative control bachia endosymbiont of D. immitis and that of D. repens group; and (ii) 136 dogs enrolled in March 2017 from were used for the duplex ftsZ-based qPCR system. Corsica where heartworms are endemic; 7 military work- All PCR systems were tested for their specifcity using ing dogs from France recruited on October 2018; and 17 several nematodes, arthropods, laboratory-maintained dogs enrolled in April 2018 from Côte d’Ivoire in which colonies as well as human, monkey, donkey, horse, cattle, blood microflariae were recorded. Canine blood samples mouse and dog DNA. DNA samples used to test the sen- were collected by a veterinarian using cephalic venipunc- sitivity and specifcity of PCR systems are summarized in ture into a citrate and serum separator tube. Te serum Additional fle 1: Table S1 and Additional fle 2: Table S2. collected and citrate blood were then stored at − 20 °C. Te analytical sensitivity was assessed using a 10-fold Tese 168 samples were subsequently processed for dilution of DNA templates, then standard curves and molecular and serological analysis. Genomic DNA was derived parameters (PCR efciency, slope, Y-inter- extracted from the blood and the microflaria-containing cept and correlation coefcient) were generated using tissues using the Qiagen DNA tissue extraction kit (Qia- CFX Manager Software Version 3 [37]. Te triplex and gen, Hilden, Germany) following the manufacturer’s rec- pan-flarial qPCR systems were challenged in detect- ommendations. Te extracted DNA was eluted in a total ing the related numbers of microflariae of D. immitis, volume of 100 µl and stored at − 30 °C. D. repens and A. reconditum. Te DNA of each species First, all DNA samples were screened for both flarial was obtained from naturally infected canine blood, D. and Wolbachia DNA using the pan-flarial and pan-Wol- immitis (Corsica, 2018), D. repens (France, 2018) and bachia qPCRs, respectively. Ten, partial cox1 and ftsZ A. reconditum (Côte d’Ivoire, 2018). First, 1 ml of each genes were amplifed and sequenced according to the blood sample was examined by the modifed Knottʼs test previous protocol from all positive samples for flarioids [9] to identify the microflariae species and their num- and Wolbachia, respectively. Secondly, the fast typing ber (Fig. 1). Ten, the microflariae concentration was method based on the direct identifcation of flarial and adjusted to 1500 mf/ml by adding Hank’s balanced salt Wolbachia genotypes used the approach combining the solution (HBSS; Gibco-BRL, Life Technologies, Eragny, triplex cox1 and duplex Wol-Diro ftsZ qPCR-based sys- France). Tereafter, two extractions were performed tems. Finally, all samples were screened for heartworms from 200 µl: (i) from each separately calibrated sample; using the duplex cox1-based qPCR in order to diferenti- and (ii) after mixing an equal volume of each of them to ate between D. immitis and A. vasorum DNA. generate a concentration of 500 mf/ml per species. Tese Te serological analysis was performed on all sera were used to evaluate the pan-flarial and triplex qPCRs, using the DiroCHEK® heartworm antigen test kit

Table 2 PCR/sequencing primers developed in this study, their characteristics and conditions

System name Target gene Primer name Sequence (5′-3′) Amplicon size Tm (°C) Specifcity (bp)

Pan-fl cox1 PCR cox1 Fwd.957 ATRGTT​ TAT​ CAG​ TCT​ TTT​ TTT​ ATT​ GG​ 509 52.0 Filariae Rwd.1465 GCA​ATY​CAA​ATA​GAA​GCA​AAAGT​ Wol ftsZ PCR ftsZ Wol.ftsZ.363.f GGR​ATG​GGT​GGT​GGY​ACT​GG 560 59.5 Wolbachia Wol.ftsZ.958.r GCA​TCA​ACC​TCA​AAY​ARA​GTCAT​

Abbreviation: Tm, melting temperature Laidoudi et al. Parasites Vectors (2020) 13:319 Page 6 of 15

(Zoetis, Lyon, France). Te test consisting of an enzyme- least one DNA marker of D. repens or DNA of its Wol- linked immunosorbent assay sandwich ELISA, targeting bachia were considered to be true positives. the antigen secreted by adult female heartworms [14]. Each serum sample was tested using two diferent proto- Statistical analysis cols: (i) 200 µl of serum was heated at 104 °C for 10 min Results generated through laboratory analysis were followed by centrifugation at 16,000×g; (ii) the second recorded in Microsoft Excel (Microsoft Corp., Redmont, protocol was performed without heat-treatment of the USA). In order to assess how Wolbachia sp. strains cor- sera following the recommendations by Beall et al. [39] related with flarial species, Spearmanʼs correlation coef- regarding the immune complex dissociation to detect any fcient was calculated. In order to evaluate the relevance heartworm antigen if present. of each diagnostic approach, the prevalence, correct clas- In order to evaluate the performance of molecular and sifcation, misclassifcation, sensitivity, specifcity, false serological assays in the absence of the gold standard positive rate, false negative rate, positive and negative test (necropsy), we developed the following approach to predictive value and Cohen’s Kappa (k) measure agree- determine true positive samples. Te sample was consid- ment was calculated. According to the scale of Landis ered a true positive for heartworm if it was positive for: & Koch [40], the agreement quality of Kappa values was (i) at least one of the molecular markers of heartworm interpreted as follows: < 0, no agreement; 0–0.2, slight (DNA of D. immitis or its Wolbachia); and (ii) a positive agreement; 0.2–0.4, fair agreement, 0.4–0.6, moderate antigen test after immune complex dissociation by heat- agreement; 0.6–0.8, substantial agreement; 0.8–1, almost ing sera. Tis approach eliminates false-positive sero- perfect agreement. Statistical analyses were performed logical results that may be obtained by increasing the using Addinsoft 2018 (XLSTAT 2018: Data Analysis and detection threshold (sensitivity) after heat-treatment of Statistical Solution for Microsoft Excel, Paris, France). the sera before use, which is subsequently confrmed by the molecular markers specifc to D. immitis. Once the Results DNA of A. reconditum was identifed, the sample was Validation of the PCR systems considered a true positive. Finally, samples positive for at Te in silico validation revealed that the pan-flarial sys- tems (28S qPCR and cox1 PCR) were specifc for flarial

Fig. 1 Microflariae in canine blood by modifed Knottʼs test. a Diroflaria immitis microflaria. b Diroflaria repens microflaria. c Diroflaria immitis (black arrow) and D. repens (blue arrow) co-infection. d Acanthocheilonema reconditum microflaria Laidoudi et al. Parasites Vectors (2020) 13:319 Page 7 of 15

parasites belonging to the subfamilies Diroflariinae, systems (Additional fle 9: Table S6, Additional fle 10: Onchocercinae, Setariinae, Oswaldoflariinae, Icosiellinae Figure S4). and Waltonellinae. Te 16S qPCR targeting Wolbachia strains was specifc for all the lineages known so far. Molecular diagnostic approaches However, the ftsZ PCR showed specifcity for Wolbachia Results of molecular screening followed by the sequence strains belonging to supergroups C, D, F and J, that may typing approach are detailed in Fig. 2a, b and Additional be associated with flarioids. Likewise, the multiplex fle 11: Table S7. Of the 168 samples tested, 49 (29.17%) qPCRs were also specifc for the target species without were positive for DNA of at least one flaria species or its failure. For each qPCR system, primer melting tempera- Wolbachia genotype. All positive results were grouped tures were closely identical and were lower than that of in: (i) 19 blood samples positive only for flarioid DNA; the probe. Indeed, the absence of primer-dimer forma- (ii) 9 samples positive only for Wolbachia DNA; and (iii) tion and hairpin structures was also confrmed. Further- 21 samples positive for both flarial and Wolbachia DNA. more, the specifcity was confrmed again by an in vitro Although partial cox1 and ftsZ amplicons were amplifed validation, as shown in Additional fle 1: Table S1, where from all positive samples for flariae and Wolbachia, cox1 positive reaction was obtained only from the target DNA gene sequence-based identifcation allowed the identif- and no negative control was amplifed. Despite using cation of the causative agent of flariosis in 35 (87.5%) out single-species or pooled DNAs, the specifc fuorescence of 40 samples amplifed by PCR (Fig. 2a). We here report signals generated through the multiplex qPCR systems 12 (30%) cases of D. immitis, 7 (17.5%) cases of D. repens, were successfully related to the target DNA (Additional 15 (37.5%) cases of A. reconditum and one case of both D. fle 3: Table S3, Additional fle 4: Figure S1). repens and A. reconditum DNA. Noteworthy, the ampli- con sequences were obtained separately from this latter Determining assay performance characteristics case after serial dilution of blood before DNA extraction. Te assay characteristics were assessed for the pan- However, the ftsZ gene sequence-based identifcation flarial, the triplex and the duplex qPCR targeting allowed the identifcation of Wolbachia sp. genotype in Wolbachia sp. endosymbiont of Diroflaria spp. Te 25 (83.33%) out of 30 samples amplifed by PCR (Fig. 2b). analytical sensitivity of the pan-flarial qPCR was con- Twenty-two (73.33%) of the Wolbachia sequences were frmed three times using D. immitis, D. repens and closely related to the strain identifed in D. immitis (Gen- A. reconditum DNA sharing the same microflariae Bank: AJ010272, 99.58%) and 3 (10%) were similar to concentration. Tis assay was able to detect up to the strain identifed in D. repens (GenBank: AJ010273, −4 1.5 × 10 microflariae per ml (mf/ml) (correspond- 99.80%). However, sequence-based identifcation failed to −6 ing to 0.75 × 10 mf/5 μl). Efciency ranged from 99.1 yield sequences from amplifed DNA in 5 cases for each to 100.7%, with a slope from − 3.34 to − 3.30, Y-inter- system, which corresponds to 12.5% and 16.7% of flaria cept values from 21.71 to 21.72 and an R2 from 0.996 and Wolbachia DNA samples, respectively. Te com- to 0.999 for microflariae of all species tested (Addi- bined multiplex approach based on the triplex cox1 qPCR tional fle 5: Table S4, Additional fle 6: Figure S2). targeting flariae and the duplex qPCR targeting Wol- However, the analytical sensitivity of the triplex qPCR, bachia, allowed the detection of 49 samples previously using pooled DNA of three species, was confrmed by considered positive based on flariae markers (Fig. 3a −1 the detection of up to 5 × 10 mf/ml (corresponding and Additional fle 12: Table S8). Te triplex cox1 qPCR −3 to 2.5 × 10 mf/5 μl) of each species simultaneously identifed the corresponding species from all the positive (Additional fle 7: Table S5, Additional fle 8: Figure samples for flariae (n = 40, 100%). Of these, 34 (34.85%) S3). qPCR efciency ranged from 100.4 to 103.7%, with samples had DNA from a single flarial species. Dirof- a slope from − 3.30 to − 3.24, Y-intercept values from laria immitis was identifed in 12 (30%), D. repens in 7 32.89 to 33.19 and an R2 from 0.993 to 0.999. Finally, (17.5%) and A. reconditum in 15 (37.5%) of these samples. the analytical sensitivity was confrmed for both the Five samples (12.5%) were positive for DNA of two flarial cox1-triplex and the ftsZ duplex qPCRs in detecting the species, of which 4 were positive for D. immitis and D. infection by D. immitis (Additional fle 9: Table S6).Te repens and one was positive for D. repens and A. recondi- −2 −1 detection limit was up to 4.03 × 10 and 4.03 × 10 tum. Positivity for three flarial species was detected for −3 mf/ml, respectively (corresponding to 2.01 × 10 and only one sample (2.5%). Te duplex Wol-Diro-ftsZ qPCR −2 2.01 × 10 , respectively) mf/5 μl, qPCR efciency was allowed the identifcation of Wolbachia genotype from 104.8 and 100.5%, respectively, slopes were − 3.212 and all samples positive for Wolbachia DNA. Twenty-one − 3.309, respectively. Y-intercept values were 31.17 and samples (21.70%) were positive for Wolbachia sp. endo- 35.98, respectively, and R2 was above 0.995 for both symbiont of D. immitis, three samples were positive for Wolbachia sp. endosymbiont of D. repens and fve Laidoudi et al. Parasites Vectors (2020) 13:319 Page 8 of 15

samples (16.67%) were positive for both strains. Among P < 0.002) and also with the presence of both Wolbachia the 168 samples screened by duplex qPCR for heart- strains together (r = 0.175, P = 0.023). On the other hand, worms (D. immitis and A. vasorum), 17 (10.12%) were 29.63% (8/27) of the samples harboring the Wolbachia positive for D. immitis DNA and no A. vasorum DNA endosymbiont of D. immitis were free of flarioid DNA was detected. (Fig. 3a) and 25% (2/8) of the samples positive for Wol- bachia endosymbiont of D. repens DNA were also free Link between Wolbachia genotype and flarial species for flarioid DNA. No correlation was observed between within the infected host A. reconditum and Wolbachia strains. Te results of the distribution of flarioid markers obtained by multiplex qPCRs are shown in Fig. 3a and Heartworm antigen detection and infection status Additional fle 12: Table S8. Interestingly, most samples Of the 168 dog sera tested for heartworm antigen, 16.67% which were positive for flarioid DNA were also positive (28) were positive before pre-treatment of the sera and for Wolbachia (21/40, 52.5%). Indeed, Wolbachia DNA were grouped into three groups (Fig. 3b): (i) 9 (5.36%) was associated with at least one Diroflaria spp. in 80% were mono-infected by heartworm and were positive (20/25) of the samples having diroflarial DNA. Analysis for both D. immitis and its Wolbachia sp. DNA except of the correlation between Wolbachia strains and Diro- one, which was positive for D. immitis DNA only; (ii) 10 flaria species are shown in Table 3. Seventy-fve percent (5.95%) were samples co-infected with at least one other (9/12) of the samples positive for D. immitis DNA alone flarioid detected by PCR, comprising 8 (4.76%) posi- were also positive for Wolbachia genotype known to be tive for D. repens and D. immitis and 2 (1.19%) positive associated with this flarioid, which corresponds to a sig- for D. immitis, D. repens and A. reconditum; and (iii) 9 nifcant correlation (r = 0.509, P < 0.0001). In addition, (5.39%) were positive for flariae other than D. immitis, there was a signifcant correlation (r = 0.181, P < 0.019) with 7 (4.17%) samples positive for A. reconditum DNA between the presence of D. repens DNA alone and that only, 1 (0.6%) positive for DNA of Wolbachia endosymbi- of Wolbachia strain commonly associated with this flari- ont of D. repens and 1 (0.6%) positive for DNA of both A. oid. In general, the presence of D. repens DNA was cor- reconditum and Wolbachia endosymbiont of D. repens. related with the presence of DNA of both Wolbachia Once the heat pre-treatment of sera was performed, the strains (r 0.454, P < 0.0001). Te presence of both D. = rate of positive samples increased up to 71.43% (n = 120). immitis and D. repens DNA was associated with the pres- Of these, 39.17% (n = 47) harbored at least one DNA ence Wolbachia endosymbiont of D. immitis (r = 0.244, marker of flarial parasites or their Wolbachia. However,

Fig. 2 Venn diagram depicting the distribution of positive samples for flariae and Wolbachia DNA. a Sequence typing method based on the cox1 gene of flariae. b Wolbachia genotyping based on the ftsZ gene Laidoudi et al. Parasites Vectors (2020) 13:319 Page 9 of 15

two samples (1.19%), one positive for Wolbachia endo- to a specifcity of 82.8% and a sensitivity of 99.3%, thus symbiont of D. immitis by qPCR and the other positive resulting in an almost perfect agreement with the true for D. repens by qPCR, remained serologically negative. positive rate (k = 0.87). Compared to the gold standard, For 73 samples (43.46%) no flarioid marker was detected; the approach combining the multiplex qPCR systems these were considered positive for unknown antigens detected one more positive sample for Wolbachia endo- (Fig. 3b and Additional fle 13: Table S9). No positive symbiont of D. immitis (Fig. 3b). Tis approach showed results in the negative control group for both serological a sensitivity of 100% and a specifcity of 99.3%, with an and molecular assays were obtained. almost perfect agreement (k = 0.98) (Additional fle 14: Table S10), whereas the detection of heartworm anti- Performance characteristics comparison of the diagnostic gen prior to heat pre-treatment of sera showed a sensi- tools tivity of 65.5% and a specifcity of 93.3%, corresponding Once the true positive samples for each flariosis were to moderate agreement (k = 0.6). Additionally, the heat determined, the diagnostic value was evaluated for each pre-treatment of sera allowed the detection of 71.4% test in the specifc detection of flariosis. Te sequence (n = 120) including 24.17% (n = 29) positive for D. immi- typing approach combining the identifcation of the flar- tis infection, 15% (n = 18) positive for flariae other than iae and Wolbachia allowed the diagnosis of heartworm D. immitis and 60.83% (n = 73) without molecular mark- infection in 86.21% (25/29) of cases, which corresponds ers of flariae. Te performance characteristics of this

Fig. 3 Venn diagram depicting the distribution of positive samples detected by molecular and serological assays. a Molecular identifcation of flariae and associated Wolbachia using the multiplex approach. b Comparison between molecular diagnosis and heartworm antigen detection before and after heat pre-treatment of sera

Table 3 Spearman correlation matrix depicting the strength of association between flaria and Wolbachia species within infected dogs Groups of Wolbachia DNA Filarial DNA D. immitis and D. Single-species Single-species A. reconditum DNA repens DNA DNA of D. immitis DNA of D. repens

Wolbachia DNA 0.506 (< 0.0001) 0.284 (0.0002) 0.565 (< 0.0001) 0.447 (< 0.0001) 0.053 (0.490) − Both Wolbachia of D. immitis and D. repens DNA 0.313 (< 0.0001) 0.175 (0.023) 0.087 (0. 259) 0.454 (< 0.0001) 0.059 (0.449) − Single-species DNA of Wolbachia ex D. immitis 0.478 (< 0.0001) 0.244 (< 0.002) 0.509 (< 0.0001) 0.079 (0.381) 0.072 (0.335) − Single-species DNA of Wolbachia ex D. repens 0.334 (< 0.0001) 0.024 (0.614) 0.037 (0.630) 0.181 (0.019) 0.104 (0.180) − − Note: The frst number represents the correlation coefcient. Values close to zero refect the absence of correlation. The associated P-values are in parentheses Laidoudi et al. Parasites Vectors (2020) 13:319 Page 10 of 15

tool in detecting heartworm infection when the sera were improvement and a tool for evaluating treatment proto- heated were 100% sensitivity and 34.5% specifcity with a cols targeting flariae and Wolbachia [44]. In the present slight agreement (k = 0.15) (Additional fle 14: Table S10). study, the analytical sensitivity of the new qPCR assays Taking the combined multiplex approach as the gold ranged from 99.3% to 107.6%, with the slope value of the standard, the sequence typing method had a specifc- standard curves ranging from − 3.34 to − 3.15 and coef- ity of 100% and a sensitivity of 62.6% and 94.1% for the cients of determination (R2) higher than 0.99. Tese char- detection of D. repens and A. reconditum, respectively. acteristics are directly derived from the design protocol, A substantial (k = 0.75) and an almost perfect (k = 0.97) where the formation of heterodimers and hairpins inside agreement with the gold standard test was observed for and between primers and probes was avoided. Primer the detection of D. repens and A. reconditum, respec- sets share a similar melting temperature which is lower tively (Additional fle 15: Table S11). than that of the probes, ofering a better sensitivity of the qPCR reaction. Discussion Te sensitivity of the pan-flarial 28S qPCR system was qPCR system validation and assay performance much higher than the triplex qPCR for the detection of characteristics D. immitis, D. repens and A. reconditum DNA, where −4 −2 Te newly developed PCR assay systems have shown the detection limit was 1.5 × 10 and 5 × 10 mf/ml, specifc detection of the target DNA for which they respectively. Indeed, the reference baseline for mitochon- were designed. Te pan-flarial 28S qPCR system aims drial DNA retrieved from the EZ1 DNA-tissue kit was at the detection of flarial DNA from biological samples. It 41.4 copies per nuclear genome [45]. Estimated genomic has been adapted for the detection of the flarial para- rRNA copy number of 150 in B. malayi [32] suggests sites known to date, such as members of the subfamilies that the 28S rRNA gene enables a high amplifcation ef- Diroflariinae and Onchocercinae parasitizing mammals, ciency rather than the mitochondrial cox1 gene of flarial reptiles and birds, and those of the subfamily Setariinae, nematodes. However, the sensitivity of the triplex qPCR confned to large mammals, and Oswaldoflariinae para- in detecting single-species DNA of D. immitis was much sites of reptiles, and amphibian parasites of the subfami- higher than that of the duplex ftsZ qPCR in detecting lies Icosiellinae and Waltonellinae [5]. Te LSU rRNA single-species DNA of Wolbachia endosymbiont of D. −3 (28S) gene targeted by this system is known for its con- immitis, where the detection limit was 4.03 × 10 and −1 served regions between the flarial species [41]. Te sec- 4.03 × 10 mf/ml, respectively. Rao et al. [42], reported ond qPCR system was customized for the detection of that flarial DNA is more frequently detected than Wol- Wolbachia DNA irrespective of their lineages. It targets bachia DNA from W. bancrofti microflaremic blood the frst part of the 16S gene which is highly conserved using qPCR assays. Te diference of sensitivity could be between Wolbachia lineages [35]. Another qPCR system explained by the weaker infection density by Wolbachia for Wolbachia targeting the 16S gene has been proposed at this parasite larval stage [42]. as a complementary diagnosis from human blood of the lymphatic flariosis caused by Wuchereria bancrofti [42]. Molecular diagnostic approaches In addition to being specifc, the multiplex qPCRs were Here, we developed and assessed two molecular discriminatory towards targeted DNA without failure approaches in detecting and identifying canine flariosis. (Additional fle 3: Table S3). Tese features are directly Te frst combined the screening and sequence typing of related to the choice of the target genes, which ofer suf- both flarial and Wolbachia DNA. Te genomic DNA was fcient discrimination between species, as is the case identifed with an almost perfect specifcity ranging from with the cox1 gene representing a nematode barcode 99.3 to 100%. However, the sensitivity ranged from mod- [34, 43] used for the development of the triplex qPCR for erate (62.5%) to perfect (94.1%) regarding the presence D. immitis, D. repens and A. reconditum and the duplex or absence of co-infection. Overlapping peaks corre- qPCR targeting D. immitis and A. vasorum agents of sponding to diferent nucleotides on electropherograms heartworm diseases. Wolbachia ftsZ gene, mainly used of the sequenced samples suggest co-infection [46]. Te for the characterization of Wolbachia supergroups [36], second approach combines two multiplex qPCR systems was used for the development of the duplex qPCR for targeting A. reconditum, D. immitis, D. repens and the both Wolbachia genotypes associated with D. immitis Wolbachia genotypes associated with the latter two spe- and D. repens. A real time PCR for Wolbachia endosym- cies. All samples were positive for at least one molecu- biont of Brugia pahangi targeting the same gene has been lar marker which were detected and identifed with described [44]. an almost perfect sensitivity and specifcity using this It is worth noting that molecular diagnosis combin- approach. Tis method is fast, simple to use, sensitive ing the detection of flarial and Wolbachia DNA is an and highly specifc in detecting occult and non-occult Laidoudi et al. Parasites Vectors (2020) 13:319 Page 11 of 15

flariosis within the infected hosts. Te present results Taylor et al. [52] have indicated that experimental crosses reinforce the utility of multiplex qPCR in detecting co- between B. pahangi and B. malayi have demonstrated infections, confrm the resolution limits of the sequence Wolbachia transmission through female worms only [52]. typing method in the identifcation of co-infections [47], Teoretically, exchange of Wolbachia between D. immi- and avoiding the sequencing procedure needed using tis and D. repens is hardly possible in natural conditions, PCR with flaria generic primers [48]. because these flariae do not share the same site and the adult worms will not have contact inside the host organ- Linkage between Wolbachia strains and flarial species ism [53]. In addition, it has been reported that each gen- within the infected host otype of Wolbachia has a specifc flarial host [36], and As expected, Wolbachia DNA was signifcantly asso- that live worms can release their Wolbachia endosymbi- ciated with Diroflaria species in 80% (20/25) of the onts into host tissues [52]. We believe that the presence samples positive for at least one Diroflaria spp. DNA, of a specifc genotype of Wolbachia is a reliable marker reinforcing the idea that this endosymbiosis relationship for the presence of its flarial host. is present in Diroflaria spp. and not in A. reconditum [27]. Of the samples positive for D. immitis DNA, 75% Performance characteristics of the heartworm antigen (9/12) were also found to be positive for the Wolbachia detection tests genotype known to be associated with this flarioid, In the present study, all diagnostic approaches did not resulting in a signifcant correlation (Table 3). As previ- react with samples from the negative control group. We ously reported, Wolbachia DNA was detected in 64.0% assessed the diagnostic value of LISA (DiroCHEK®) in of the samples positive for D. immitis [27], and in 81.6% detecting heartworms. Te direct exploration of heart- of the samples positive for D. repens [49]. In the present worm antigen from sera without heating the sera showed study, we investigated the link between Wolbachia geno- a moderate performance, with sensitivity and specifcity type and D. repens infection. Te samples positive for a values of 65.52% and 93.3%, respectively. Positive anti- single-species DNA of D. repens had a signifcant corre- gen tests were obtained from 19 out of 29 (65.52%) of the lation with the Wolbachia genotype known to be associ- samples determined as true positives for heartworm and ated with this flarioid. Tis result corroborates the data these often harbored both D. immitis and its Wolbachia by Vytautas et al. [49]. Interestingly, the presence of the sp. DNA. However, 10 samples (34.38%) of which 8 (80%) single-species DNA of D. repens was also strongly corre- harbored only DNA of the Wolbachia endosymbiont of lated with the presence of both Wolbachia strains asso- D. immitis, remained undetected by serology. Te lack ciated with Diroflaria spp. Tis association could be of sensitivity of this assay was unexpected. Tis may be explained, either by the presence of an occult co-infection due in part to the presence of juvenile parasites, which with D. immitis, or by an exchange of Wolbachia strains do not produce detectable antigens [54]. Nevertheless, between Diroflaria spp. Te frst suggestion is supported similar discordances have recently been reported, where by the fact that co-infection of D. repens and D. immitis is 41 (38.7%) positive PCR-confrmed microflaremic sam- often associated with an occult form. Tis phenomenon ples were negative for heartworm antigen [55]. Terefore, results from a competitive suppression between micro- flarial Wolbachia interact with the host by activating the flariae species [19]. On the other hand, Wolbachia sp. T1 type protective-immune response [56], which could of D. immitis was detected in 29.63% (8/27) of the sam- be implemented in the clearance of heartworm antigens. ples in which D. immitis DNA was not detected and, in Finally, nine out of 28 (32.14%) samples were positive the same samples, an antigen was detected after heat for A. reconditum and/or D. repens; these flariae are pre-treatment of sera. Tis result confrms the possibil- known to generate a cross-reactivity in heartworm anti- ity to detect Wolbachia DNA in occult infections. Te gen tests even in the absence of heat treatment of the sera utility of Wolbachia as a diagnosis target for the occult [14]. However, 29 (96.67%) of the samples positive for at heartworm disease has been demonstrated in the dead- least one molecular heartworm marker tested positive end host, such as humans and cats, where the parasite for heartworm antigen after heating the sera; this step cannot achieve its maturation and the infection might has recently been added to improve the sensitivity of this be amicroflariaemic [23]. However, the second sugges- test under certain conditions [15, 41]. In contrast, the tion related to the exchange (horizontal transfer) of Wol- heat pre-treatment of sera strongly altered the specifc- bachia strains between Diroflaria species is in contrast ity of the heartworm antigen test. Cross-reactivity was with the published data. Wolbachia transmission princi- observed overall in the samples positive for any flarial pally occurs via eggs of female worms (vertical transfer) parasite, as well as in samples positive for unknown anti- [36, 50]. Te vertical transfer of Wolbachia leads to the gens. Filarial (D. repens and A. reconditum) and non- specialization of the host-symbiotic relationship [51]. flarial nematodes (A. vasorum and S. lupi), are known to Laidoudi et al. Parasites Vectors (2020) 13:319 Page 12 of 15

Fig. 4 Diagram showing the diagnostic procedure proposed in this study

cross-react with heartworm antigen tests [15, 57]. Tere- Conclusions fore, molecular and serological diagnosis of A. vasorum Te molecular approach developed herein represents an from canine blood showed a close efciency [58]. Regard- improvement in the diagnosis of canine flariosis. It relies ing the life-cycle of A. vasorum [59], the absence of its principally on TaqMan multiplex qPCR technologies. We DNA from all blood samples tested in the present study encourage researchers to follow the molecular procedure is not sufcient enough to rule out infection. However, summarized in Fig. 4. Te approach allows the detection the qPCR targeting A. vasorum developed in this study of flarial parasites as well as their Wolbachia endosymbi- can be used as a simplex for detecting parasite larvae onts at the family level from canine blood. Furthermore, from other biological samples, such as faeces, pharyn- we have implemented a highly sensitive and specifc tri- geal swabs as well as in the intermediate hosts. In the plex qPCR assay for the simultaneous detection of D. absence of circulating microflaria, the antigen detection immitis, D. repens and A. reconditum, the most frequent can neither confrm nor exclude the occult heartworm agents of canine flariosis. A duplex qPCR is presented in area endemic for the other species that cross-react for the simultaneous identifcation of Wolbachia geno- with the test, as is the case of D. repens (Fig. 2). Dirof- types from D. immitis and D. repens as a complementary laria repens microflariae induce a suppressive efect on diagnostic of canine diroflariosis and their occult forms. those of D. immitis [18] which induces the occult form Two primer sets are proposed for PCR/sequencing of of the latter. Te European Society of Diroflariosis and flariae and Wolbachia DNAs. Finally, the approach is Angiostrongylosis (ESDA) does not recommend routine complemented by a duplex qPCR for D. immitis and A. heat pre-treatment of sera in an area endemic for these vasorum, agents of canine heartworm disease. Moreo- parasites; this is recommended for use only to resolve the ver, this approach is useful in epidemiological surveil- discrepancy between other tests, especially when a dog is lance, in diagnosis and therapeutic follow-up of both positive for the microflaria test and negative for serology, flarial parasites and their Wolbachia endosymbionts. or to confrm a suspicion of clinical disease suggestive of Te specifc detection of Wolbachia genotypes could be microflaremia [54]. used for the diagnosis of flariosis and the assessment of related pathogeny within dead-end hosts, as is the case of D. immitis infection in humans and cats. Heartworm Laidoudi et al. Parasites Vectors (2020) 13:319 Page 13 of 15

antigen testing without heat treatment of the sera is not Authors’ contributions YL, BD, MV and OM designed the study. YL, BD, OM designed, and YL carried reliable in an endemic area with other flarial species, out the data analysis. YL, MV, EAN and OM drafted the manuscript. All authors including A. reconditum and D. repens. We discourage read and approved the fnal manuscript. the use of heat pre-treatment of sera, which signifcantly Funding alters the specifcity of the assay due to the cross-reactiv- This study was supported by the Institut-Hospitalo-Universitaire (IHU) Méditer- ity between many flarial and non-flarial nematodes and ranée Infection, the National Research Agency under the program “Investisse- the possibility of false-positive results that may induce ments d’avenir”, reference ANR-10-IAHU-03, the Région Provence-Alpes-Côte d’Azur and European funding FEDER PRIMI. Funding sources had no role in the unnecessary heavy treatment of heartworm disease. design or conduct of the study; collection, management, analysis or interpre- tation of the data; or preparation, review or approval of the manuscript. Supplementary information Availability of data and materials Supplementary information accompanies this paper at https​://doi. The data supporting the conclusions of this article are included within the org/10.1186/s1307​1-020-04185​-0. article and its additional fles.

Additional fle 1: Table S1. DNA used as a positive control in this study. Ethics approval and consent to participate Not applicable. Additional fle 2: Table S2. DNA used as a negative control in this study. Additional fle 3: Table S3. In vitro validation of the triplex cox1-based Consent for publication qPCR. Not applicable.

Additional fle 4: Figure S1. Assessment of the specifcity of the triplex Competing interests cox1-based qPCR in detecting the target DNA. The authors declare that they have no competing interests. Additional fle 5: Table S4. Analytical sensitivity of the pan-flarial 28S-based qPCR in detecting single-species DNA. Author details 1 Microbes, Evolution, Phylogeny and Infection (MEPHI), UMR Aix-Marseille Additional fle 6: Figure S2. a, b, c-1. Efciency of pan-flarial 28S-based University, IRD, APHM, IHU Méditerranée Infection, 19–21, Bd Jean Moulin, qPCR using single-species DNA of D. immitis, D. repens and A. reconditum. 13005 Marseille, France. 2 IHU Méditerranée Infection, 19–21, Bd Jean Moulin, a, b, c-2. Standard curves generated from serial 10-fold dilution of each 13005 Marseille, France. 3 Ceva Santé Animale, 10 Avenue de la Ballastière, DNA. 33500 Libourne, France. 4 VITROME, UMR Aix-Marseille University, IRD, SSA, Additional fle 7: Table S5. Analytical sensitivity of the triplex cox1-based APHM, IHU Méditerranée Infection, 19–21, Bd Jean Moulin, 13005 Marseille, qPCR using pooled DNA. France. Additional fle 8: Figure S3. a Efciency of the triplex cox1-based qPCR using pooled DNA. b Standard curves generated from serial 10-fold dilu- Received: 15 October 2019 Accepted: 13 June 2020 tion of DNA. Additional fle 9: Table S6. Analytical sensitivity of the triplex cox1 and the duplex ftsZ-based qPCRs in detecting single-species DNA of D. immitis and its Wolbachia. References Additional fle 10: Figure S4. a, b Efciency of the triplex cox1 and the 1. Ravindran R, Varghese S, Nair SN, Balan VM, Lakshmanan B, Ashruf RM, duplex ftsZ-based qPCRs using single-species DNA of D. immitis and its et al. Canine flarial infections in a human Brugia malayi endemic area of Wolbachia. c, d Standard curves generated from serial 10-fold dilution of India. Biomed Res Int. 2014;2014:630160. DNA. 2. Latrofa MS, Annoscia G, Colella V, Cavalera MA, Maia C, Martin C, et al. A real-time PCR tool for the surveillance of zoonotic Onchocerca lupi in Additional fle 11: Table S7. Samples distribution according to flariae dogs, cats and potential vectors. PLoS Negl Trop Dis. 2018;12:e0006402. and Wolbachia DNA detected by the sequence typing approach. 3. Seixas F, Travassos P, Coutinho T, Lopes AP, Latrofa MS, dos Pires MA, et al. Additional fle 12: Table S8. Samples distribution according to flariae The eyeworm Thelazia callipaeda in Portugal: current status of infection in and Wolbachia DNA detected by the multiplex approach. pets and wild mammals and case report in a beech marten (Martes foina). Vet Parasitol. 2018;252:163–6. Additional fle 13: Table S9. Samples tested positive by at least by one 4. Otranto D, Dantas-Torres F, Brianti E, Traversa D, Petrić D, Genchi C, et al. diagnostic approach. Vector-borne helminths of dogs and humans in Europe. Parasit Vectors. Additional fle 14: Table S10. Performance of molecular and serological 2013;6:16. assays in detecting D. immitis infection. 5. Adamson RC. Nematode parasites of vertebrates. Their development and Additional fle 15: Table S11. Performance of molecular approaches in transmission. Wallingford: CAB International; 2000. detecting D. repens and A. reconditum. 6. Ilie MS, Imre K, Hotea I, Dărăbuș G. Survey of canine diroflariosis from south-western Romania—preliminary results. In: 3rd European Diroflaria Days, 21–22 June 2012, Parma, Italy; 2012. p. 68. Abbreviations 7. Simón F, González-Miguel J, Diosdado A, Gómez PJ, Morchón R, Kartashev FAM: 6-FAM (6-carboxyfuorescein); VIC: VIC (2′-chloro-phenyl-1,4- dichloro- V. The complexity of zoonotic flariasis episystem and its consequences: a 6-carboxyfuorescein); Cy5: cyanine 5; TAMRA: tetra-methyl-rhodamine; BHQ: multidisciplinary view. Biomed Res Int. 2017;2017:6436130. black hole quencher. 8. Genchi C, Kramer L. Subcutaneous diroflariosis (Diroflaria repens): an infection spreading throughout the old world. Parasit Vectors. Acknowledgements 2017;10:517. We gratefully thank Sebastien Cortaredona for assistance with statistical analy- 9. Magnis J, Lorentz S, Guardone L, Grimm F, Magi M, Naucke TJ, et al. sis. We also thank Professor Domenico Otranto who provided us with positive Morphometric analyses of canine blood microflariae isolated by the control DNA. We express our gratitude to Zoetis who provided us with the Knott’s test enables Diroflaria immitis and D. repens species-specifc and DiroCHEK® tests. Acanthocheilonema (syn. Dipetalonema) genus-specifc diagnosis. Parasit Vectors. 2013;6:48. Laidoudi et al. Parasites Vectors (2020) 13:319 Page 14 of 15

33. Ferri E, Barbuto M, Bain O, Galimberti A, Uni S, Guerrero R, et al. Integrated 10. Nutman BT. Filarial infections. In: Jong EC, Sanford CA, editors. The travel taxonomy: traditional approach and DNA barcoding for the identifcation and tropical medicine manual. 4th ed. Philadelphia: Saunders/Elsevier; of flarioid worms and related parasites (Nematoda). Front Zool. 2009;6:1. 2008. p. 611–25. 34. Yilmaz E, Fritzenwanker M, Pantchev N, Lendner M, Wongkamchai S, 11. Luca F, Luigi V. Filarial infections. In: Day MJ, editor. Arthropod-borne Otranto D, et al. The mitochondrial genomes of the zoonotic canine infectious diseases of the dog and cat. 2nd ed. Boca Raton: CRC Press; flarial parasites Diroflaria (Nochtiella) repens and Candidatus Diroflaria 2016. (Nochtiella) honkongensis provide evidence for presence of cryptic spe- 12. Simsek S, Ciftci AT. Serological and molecular detection of Diroflaria spe- cies. PLoS Negl Trop Dis. 2016;10:e0005028. cies in stray dogs and investigation of Wolbachia DNA by PCR in Turkey. J 35. Willems A, Watteyne S, Mertens J, Gillis M, Vandekerckhove TTM, Swing Arthropod Borne Dis. 2016;10:445–53. JG. Phylogenetic analysis of the 16S rDNA of the cytoplasmic bacterium 13. Bowman DD, Atkins CE. Heartworm biology, treatment, and control. Vet Wolbachia from the novel host Folsomia candida (Hexapoda, Collem- Clin North Am Small Anim Pract. 2009;39:1127–58. bola) and its implications for wolbachial taxonomy. Fems Microbiol Lett. 14. Little S, Saleh M, Wohltjen M, Nagamori Y. Prime detection of Diroflaria 1999;180:279–86. immitis: understanding the infuence of blocked antigen on heartworm 36. Bandi C, Anderson TJC, Genchi C, Blaxter ML. Phylogeny of Wolbachia in test performance. Parasit Vectors. 2018;11:186. flarial nematodes. Proc R Soc B Biol Sci. 1998;265:2407–13. 15. Aroch I, Rojas A, Slon P, Lavy E, Segev G, Baneth G. Serological cross- 37. Bio‐Rad, L. Real-time PCR applications guide. Hercules: Bio-Rad Laborato- reactivity of three commercial in-house immunoassays for detection ries Inc; 2006. p. 41. of Diroflaria immitis antigens with Spirocerca lupi in dogs with benign 38. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment esophageal spirocercosis. Vet Parasitol. 2015;211:303–5. search tool. J Mol Biol. 1990;215:403–10. 16. Venco L, Manzocchi S, Genchi M, Kramer LH. Heat treatment and false- 39. Beall MJ, Arguello-Marin A, Drexel J, Liu J, Chandrashekar R, Alleman AR. positive heartworm antigen testing in ex vivo parasites and dogs naturally Validation of immune complex dissociation methods for use with heart- infected by Diroflaria repens and Angiostrongylus vasorum. Parasit Vectors. worm antigen tests. Parasit Vectors. 2017;10:481. 2017;10:476. 40. Landis JR, Koch GG. An application of hierarchical kappa-type statistics 17. Rinaldi L, Cortese L, Meomartino L, Pagano TB, Pepe P, Cringoli G, et al. in the assessment of majority agreement among multiple observers. Angiostrongylus vasorum: epidemiological, clinical and histopathological Biometrics. 1977;33:363–74. insights. BMC Vet Res. 2014;10:236. 41. Chilton NB, Huby-Chilton F, Gasser RB. First complete large subunit ribo- 18. Monica A, Matei IA, Amico GD, Bel LV, Dumitrache MO, Modrý D, et al. somal RNA sequence and secondary structure for a parasitic nematode: Diroflaria immitis and D. repens show circadian co-periodicity in naturally phylogenetic and diagnostic implications. Mol Cell Probes. 2003;17:33–9. co-infected dogs. Parasit Vectors. 2017;10:116. 42. Rao RU, Atkinson LJ, Ramzy RM, Helmy H, Farid HA, Bockarie MJ, et al. A 19. Mircean M, Ionică AM, Mircean V, Györke A, Codea AR, Tăbăran FA, et al. real-time PCR-based assay for detection of Wuchereria bancrofti DNA in Clinical and pathological efects of Diroflaria repens and Diroflaria immitis blood and mosquitoes. Am J Trop Med Hyg. 2006;74:826–32. in a dog with a natural co-infection. Parasitol Int. 2017;66:331–4. 43. Prosser SWJ, Velarde-Aguilar MG, León-Règagnon V, Hebert PDN. Advanc- 20. Sassi AJ, Geary JF, Leroux LP, Moorhead AR, Satti M, Mackenzie CD, et al. ing nematode barcoding: a primer cocktail for the cytochrome c oxidase Identifcation of Diroflaria immitis proteins recognized by antibodies subunit I gene from vertebrate parasitic nematodes. Mol Ecol Resour. from infected dogs. J Parasitol. 2014;100:364–7. 2013;13:1108–15. 21. Solgi R, Sadjjadi SM, Mohebali M, Zarei Z, Golkar M, Raz A. Development 44. Simoncini L, Casiraghi M, Bazzocchi C, Sacchi L, Bandi C, Genchi C. Real- of new recombinant DgK antigen for diagnosis of Diroflaria immitis infec- time PCR for quantifcation of the bacterial endosymbionts (Wolbachia) tions in dogs using ELISA technique and its comparison to molecular of flarial nematodes. Parassitologia. 2001;43:173–8. methods. Iran Biomed J. 2018;22:283–9. 45. Fazzini F, Schöpf B, Blatzer M, Coassin S, Hicks AA, Kronenberg F, et al. 22. Martin C, Gavotte L. The bacteria Wolbachia in flariae, a biological Russian Plasmid-normalized quantifcation of relative mitochondrial DNA copy dolls’system: new trends in antiflarial treatments. Parasite. 2010;17:79–89. number. Sci Rep. 2018;8:15347. 23. Turb ME, Zambon E, Zannoni A, Russo S, Gentilini F. Detection of Wol- 46. Kommedal Ø, Karlsen B, Sæbø Ø. Analysis of mixed sequencing chro- bachia DNA in blood for diagnosing flaria-associated syndromes in cats. J matograms and its application in direct 16S rRNA gene sequencing of Clin Microbiol. 2012;50:2624–30. polymicrobial samples. J Clin Microbiol. 2008;46:3766–71. 24. Dingman P, Levy JK, Kramer LH, Johnson CM, Lappin MR, Greiner EC, et al. 47. Irshad M, Gupta P, Mankotia DS, Ansari MA. Multiplex qPCR for serodetec- Association of Wolbachia with heartworm disease in cats and dogs. Vet tion and serotyping of hepatitis viruses: a brief review. World J Gastroen- Parasitol. 2010;170:50–60. terol. 2016;22:4824–34. 25. Landum M, Ferreira CC, Calado M, Alho AM, Maurício IL, Meireles JS, 48. Rishniw M, Barr SC, Simpson KW, Frongillo MF, Franz M, Dominguez et al. Detection of Wolbachia in Diroflaria infected dogs in Portugal. Vet Alpizar JL. Discrimination between six species of canine microflariae by a Parasitol. 2014;204:407–10. single polymerase chain reaction. Vet Parasitol. 2006;135:303–14. 26. Rossi MID, Aguiar-Alves F, Santos S, Paiva J, Bendas A, Fernandes O, et al. 49. Sabūnas V, Radzijevskaja J, Sakalauskas P, Petkevičius S, Karvelienė B, Detection of Wolbachia DNA in blood from dogs infected with Diroflaria Žiliukienė J, et al. Diroflaria repens in dogs and humans in Lithuania. immitis. Exp Parasitol. 2010;126:270–2. Parasit Vectors. 2019;12:177. 27. Maia C, Altet L, Serrano L, Cristóvão JM, Tabar MD, Francino O, et al. 50. Rao RU. Endosymbiotic Wolbachia of parasitic flarial nematodes as drug Molecular detection of Leishmania infantum, flariae and Wolbachia spp. targets. Indian J Med Res. 2005;122:199–204. in dogs from southern Portugal. Parasit Vectors. 2016;9:170. 51. Taylor MJ, Bandi C, Hoerauf A. Wolbachia bacterial endosymbionts of 28. Hall T, Biosciences I, Carlsbad C. BioEdit: an important software for flarial nematodes. Adv Parasitol. 2005;60:245–84. molecular biology. GERF Bull Biosci. 2011;2:60–1. 52. Taylor MJ, Hoerauf A. Wolbachia bacteria of flarial nematodes. Parasitol 29. Owczarzy R, Tataurov AV, Wu Y, Manthey JA, McQuisten KA, Almabrazi HG, Today. 1999;15:437–42. et al. IDT SciTools: a suite for analysis and design of nucleic acid oligom- 53. McCall JW, Genchi C, Kramer LH, Guerrero J, Venco L. Heartworm disease ers. Nucleic Acids Res. 2008;36:163–9. in animals and humans. Adv Parasitol. 2008;66:193–285. 30. Ye J, Coulouris G, Zaretskaya I, Cutcutache I, Rozen S, Madden TL. Primer- 54. Nelson CT, McCall JW, Jones S, Moorhead A. Current canine guidelines BLAST: a tool to design target-specifc primers for polymerase chain for the prevention, diagnosis, and management of heartworm (Diroflaria reaction. BMC Bioinform. 2012;13:134. immitis) infection in dogs. American Heartworm Society; 2018. https​:// 31. Bik HM, Fournier D, Sung W, Bergeron RD, Thomas WK. Intra-genomic vari- www.heart​worms​ociet​y.org/veter​inary​-resou​rces/ameri​can-heart​worm- ation in the ribosomal repeats of nematodes. PLoS ONE. 2013;8:e78230. socie​ty-guide​lines​. Accesed 25 Dec 2019. 32. Lagatie O, Merino M, Debrah LB, Debrah AY, Stuyver LJ. An isothermal 55. Alho AM, Landum M, Ferreira C, Meireles J, Gonçalves L, de Carvalho LM, DNA amplifcation method for detection of Onchocerca volvulus infection et al. Prevalence and seasonal variations of canine diroflariosis in Portu- in skin biopsies. Parasit Vectors. 2016;9:624. gal. Vet Parasitol. 2014;206:99–105. Laidoudi et al. Parasites Vectors (2020) 13:319 Page 15 of 15

56. Tabar MD, Altet L, Martínez V, Roura X, Veterinario H, Vicente S, et al. Wol- 59. Ferdushy T, Hasan MT. Angiostrongylus vasorum: the “French heartworm”. bachia, flariae and Leishmania coinfection in dogs from a Mediterranean Parasitol Res. 2010;107:765–71. area. J Small Anim Pract. 2013;54:174–8. 57. Schnyder M, Deplazes P. Cross-reactions of sera from dogs infected with Angiostrongylus vasorum in commercially available Diroflaria immitis test Publisher’s Note kits. Parasit Vectors. 2012;5:258. Springer Nature remains neutral with regard to jurisdictional claims in pub- 58. Jeferies R, Morgan ER, Helm J, Robinson M, Shaw SE. Improved detection lished maps and institutional afliations. of canine Angiostrongylus vasorum infection using real-time PCR and indirect ELISA. Parasitol Res. 2011;109:1577–83.

Ready to submit your research ? Choose BMC and benefit from:

• fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year

At BMC, research is always in progress.

Learn more biomedcentral.com/submissions