Development and validation of a quantitative PCR for the detection of Guinea worm (Dracunculus medinensis)

Dracunculus medinensis (Guinea worm) is a parasitic nematode that can cause the debilitating disease dracunculiasis (Guinea worm disease) in humans. The global Guinea Worm Eradication Program has led intervention and eradication efforts since the 1980s, and Guinea worm infections in people have decreased >99.99%. With the final goal of eradication drawing nearer, reports of animal infections from some remaining endemic countries pose unique challenges. Currently, confirmation of suspected Guinea worm infection relies on conventional molecular techniques such as polymerase chain reaction (PCR), which is not specific to Guinea worm and, therefore, requires sequencing of the PCR products to confirm the identity of suspect samples, a process that often takes a few weeks. To decrease the time required for species confirmation, we developed a quantitative PCR assay targeting the mitochondrial cytochrome b (cytb) gene of Guinea worm. Our assay has a limit of detection of 10 copies per reaction. The mean analytical parameters (± SE) were as follows: efficiency = 93.4 ± 7.7%, y-intercept = 40.93 ± 1.11, slope = -3.4896 ± 0.12, and the R2 = 0.999 ± 0.004. The assay did not amplify other nematodes found in Guinea worm-endemic regions and demonstrated 100% diagnostic sensitivity and specificity. Implementation of this quantitative PCR assay for Guinea worm identification could eliminate the need for DNA sequencing to confirm species. Thus, this approach can be implemented to provide more rapid confirmation of Guinea worm infections, leading to faster execution of Guinea worm interventions while increasing our understanding of infection patterns.


Introduction
Dracunculus medinensis (Guinea worm; Family Dracunculidae, Order Spirurida) is a parasitic nematode that causes the debilitating disease dracunculiasis (Guinea worm disease) in humans. Dracunculiasis is caused by the development and emergence of a large, gravid female worm measuring up to a meter long [1,2]. In particular, the emergence of the gravid female through the skin of a distal extremity can cause considerable pain for several weeks and can leave the affected individual with an open wound that is susceptible to secondary bacterial infections [1,2]. The Guinea Worm Eradication Program (GWEP) has worked to eradicate this parasite since the 1980s, during which time human infections have decreased dramatically from an estimated 3.5 million cases in 1986 to just 15 in 2021 [1,3,4].
Dracunculus transmission occurs when an obligate intermediate host, a cyclopoid copepod, infected with third-stage larvae [L3s] is ingested by a susceptible definitive host [2,5]. As measures were implemented to minimize the transmission of Guinea worm (primarily focusing on ensuring access to clean drinking water free of infected copepods), dracunculiasis cases decreased dramatically [1,3,6,7]. However, in recent years, infections in animals have complicated eradication efforts; most animal infections have occurred in domestic dogs (Canis lupus familiaris) in Chad, Africa, with fewer numbers of dogs in other countries, domestic cats (Felis catus) in Chad, and baboons (Papio spp.) in Ethiopia [8]. Accordingly, while traditional Guinea worm control measures remain in place, new measures focusing on preventing transmission to animals have been enacted to address this final barrier to eradication. One vital aspect of these control measures is rapid containment of infected animal hosts and confirmation of the identity of any emerging worms suspected to be Guinea worm.
Accurate surveillance and effective implementation of some intervention measures rely on confirmation that suspected infections are actually D. medinensis. However, expeditious molecular confirmation of suspected nematodes as Guinea worm can be challenging. Currently, samples are processed via a conventional PCR that is not specific to Guinea worm, amplifying other nematode DNA; therefore, this PCR must be followed by Sanger sequencing in order to confirm the identity. While this method reliably differentiates Guinea worm from other nematodes, the turnaround time from sample receipt at appropriate diagnostic laboratories to species confirmation can take up to two weeks. Therefore, the purpose of this Guinea worm quantitative polymerase chain reaction assay (GW qPCR) is to provide a highly sensitive and specific diagnostic tool for rapid identification of suspect infections with adult, female Guinea worms in humans and domestic and wild animals. This GW qPCR platform avoids the need for sequencing and can produce results (including the time needed for DNA extraction) within 24 hours.

Methods
The development and validation of the GW qPCR protocol were carried out at the Southeastern Cooperative Wildlife Disease Study (SCWDS) at the University of Georgia in Athens, Georgia, USA. First, primer and probe sets were designed to target Guinea worm based on closely related nematode genomic data in the National Center for Biotechnology Information (NCBI) database. Second, the conditions for the GW qPCR were optimized using a synthetic Guinea worm amplicon serially diluted as a standard curve along with extracted DNA from Guinea worms as positive samples and DNA extracts from the related Dracunculus insignis as negative control samples. Next, the analytical sensitivity was estimated using standard curve C q values generated in-house (n = 9) and at other labs (n = 3). Last, the diagnostic sensitivity (D Sn ) and specificity (D Sp ) were estimated with non-target samples (n = 180) and Guinea worm samples (n = 200) run in triplicate.

In silico qPCR primer and probe design
Available mitochondrial genome sequences for Guinea worm were downloaded from Gen-Bank. Sequences for close phylogenetic relatives (D. insignis and D. lutrae) were generated by assembling the Illumina sequencing reads for these species from [9] using v1.9 of the mitoBIM pipeline [10] with mira v4.0.2 [11], run for 30 iterations and a baiting kmer length of 41, in 'quick' mode using the reference mitochondrial genome for Guinea worm [9] as bait sequence. Assembled mitochondrial sequences were then annotated using Prokka v1.14.5 [12], and the ORF with the highest scoring blastp hit with e-value < = 0.01 when compared against each O. volvulus mitochondrial protein set [13] was annotated as the likely homolog for that sample. For Guinea worm, four mitochondrial genes (cytochrome c oxidase subunit 1 [COI], cytochrome c oxidase subunit 3 [cox3], cytochrome b [cytb], and NADH dehydrogenase 3 [nd3]) had significant coverage in GenBank and were identified as candidate genes for primer and probe design [14]. For each gene, sequences were aligned using the MUSCLE alignment feature [15] in Geneious Prime (Ver 2019.0.4). The consensus sequences for each gene alignment were generated at 100% identity threshold, revealing homologous regions within the cytb gene that were suitable candidates for primer and probe design. A homologous region in the cox3 gene was also identified (see S1 Methods).
The consensus sequence generated from all available Guinea worm cytb gene sequences was imported into PrimerQuest software (Integrated DNA Technologies, Coralville, IA; Table 1) for primer and hydrolysis probe design. Primer and probe sequences were then aligned with cytb consensus sequences from the other nematode taxa to confirm mismatches. Thus, the in silico specificity of the primer and probe combination was limited to Guinea worm. Primers and probe were synthesized by MilliporeSigma (Burlington, MA, USA).

Standard curve production and estimation of qPCR assay parameters
To develop a standard curve, a gBlock Gene Fragment (Integrated DNA Technologies, Iowa, USA) was generated using a known Guinea worm cytb gene target sequence plus a 50-bp overhang on both ends of the target sequence, resulting in a 224-bp sequence for positions 156-379. Due to the low GC content, adaptor sequences were added to both ends of the cytb gene fragment to allow for successful synthesis, resulting in a final gene fragment length of 283-bp. Standard curves were generated using 10-fold serial dilutions and tested in triplicate at SCWDS to estimate the efficiency [10-1/slope-1; 16], repeatability, and limit of detection (LOD). The 10-fold serial dilutions of the gBlock (ranging from 1 x 10 8 copies to 1 copy per reaction) were run in triplicate to optimize the qPCR protocol (e.g., determining the most efficient annealing temperature at 55.1˚C) and estimate the analytical sensitivity and specificity of the assay. A total of 12 standard curve experiments were performed to estimate the analytical efficiency, slope, y-intercept, correlation coefficient (R 2 ), LOD/analytical sensitivity, repeatability, and reproducibility. Three of these 12 experiments were performed by labs at the University of Georgia that were not associated with this project. The LOD was defined as the dilution with the lowest copy number detected 50% of the time [17]. Repeatability and reproducibility were calculated as the percent coefficient of variation using the mean and standard deviation (SD) of C q values within or among, respectively, the experiments.

Sample acquisition
Gravid female Guinea worms preserved in 70% ethanol that were morphologically and genetically confirmed as Guinea worm were obtained from Guinea worm endemic countries in Africa from 2006-2020 [ Table 2; 9,14]. Sources of these samples represented the full extent of the currently known definitive non-human hosts (dog, cat, baboon, and leopard) and extant geographic range (Angola, Cameroon, Chad, Ethiopia, Mali, and South Sudan) of GW and thus sufficiently represent the samples on which this assay will be used [18]. DNA extracts from sections of the worms containing larvae, or aliquots of larvae expressed from adult females after collection, were used to optimize and validate this protocol. Host tissue and additional parasite samples were obtained through multiple sources to test assay specificity. Intermediate [i.e., cyclopoid copepods; 1,5,19], potential paratenic and transport [20][21][22][23], and definitive host species [i.e., domestic cat, domestic dog, and baboon; 5,14,24] were included (Table 3). Morphologically-identified and molecularly-confirmed nematodes (n = 91) and one trematode species were obtained from research samples or wildlife submitted for diagnostic evaluation by the SCWDS diagnostic service (Table 4). These non-target samples included other dracunculids (n = 62) that are most likely to cross-react in the assay [14,25,26]. Dracunculus insignis samples were primarily obtained from raccoons, a common host of this species in North America, with additional samples from mustelids and skunks [27][28][29][30][31]. Two adult, female Dracunculus ophidensis [32] from Lake Erie watersnakes (Nerodia sipedon insularum) in the USA were available; these nematodes were presumed to be D. ophidensis, as no males, which are necessary to morphologically identify species in this genus [32][33][34][35], were available for morphological confirmation, and minimal genetic information was available on Genbank for comparison at the time of this validation. Other non-target parasites were obtained from potential Guinea worm hosts.

Ethical clearance
All samples included in this research were derived from materials collected either under procedures sanctioned by the World Health Organization and national governments for containment and treatment of GW, or were collected under authorizations for research or diagnostic services. These authorizations are granted by Institutional Animal Care and Use Committees, which provide oversight of each Animal Use Protocol (AUP). Some of the samples collected for the purpose of research and diagnostics have not been included in published reports prior to this manuscript.

Extraction of DNA from tissue samples
Host and parasite tissues were preserved and stored either frozen at -20˚C or preserved in 70-100% ethanol at ambient temperature. Sections of worms, portions of tissues, or whole larvae in ethanol were incubated at room temperature in 1.5 mL microcentrifuge tubes for at least 12 hours to allow residual ethanol to evaporate. Frozen worm or tissue samples were thawed and

Detection of Guinea worm DNA using qPCR
The total reaction volume of qPCR reactions was 15 μl, with 7.5 μl Taq

Estimation of diagnostic sensitivity and specificity
Guinea worm (n = 200), host (n = 88), and non-target parasite (n = 92) samples (Tables 2-4) were tested in triplicate to assess the diagnostic sensitivity and specificity of the GW qPCR assay. A sample was considered "positive" when the C q values for all three replicates surpassed the detection threshold at � 40 cycles, "suspect" when only two samples were above the detection threshold, and "negative" when one or none of the replicates was above the threshold. "Suspect" samples were retested in triplicate to confirm status. To ensure the protocol was sufficiently optimized for various laboratory settings, each of the three external labs that participated in the analytical sensitivity/specificity experiments also tested 17 blind DNA samples in triplicate (source groups indicated with^in Tables 2-4).  For the designed cytb primer/probe set, three of the available D. insignis sequences showed complete homology to the cytb forward primer but not the reverse primer or probe (Fig 1). "ND" denotes where no data was available.

In silico qPCR primer and probe design
Indicates groups/samples used for validation at other laboratories at UGA.

▲
Indicates potential paratenic/transport hosts of Guinea worm. ¶ Indicates potential definitive hosts of Guinea worm. ⬢ Submitted to the SCWDS diagnostic service, Animal Use Protocol: A2020 11-010-Y2-A3. Initial optimization and specificity testing of the cytb gene resulted in nonspecific amplification of D. insignis samples, which was resolved by increasing the annealing temperature and adding detection enhancer to the qPCR reaction mix.

Standard curve production and estimation of qPCR assay analytical parameters
The LOD was determined to be 10 copies per reaction, which was detected in 86% (31/36 replicates) of the reactions throughout the experiments (n = 12; Table 5); detection of 1 copy per reaction was <50%. Average (± SE) parameters over 12 experiments were as follows: efficiency = 93.4 ± 7.7%, y-intercept = 40.93 ± 1.11, slope = -3.4896 ± 0.12, and the R 2 = 0.999 ± 0.004 (Fig 2). The intra-and inter-assay coefficients of variability ranged from < 0.02-12.0% and 7.1-9.9% (Table 4), respectively. In all 12 experiments with a LOD of 10 copies, the analytical specificity and sensitivity were both 100%.  Table 5. Guinea worm qPCR inter-assay variability of standard curve dilution series (10 8 to 10 1 ) across 12 experiments (plates) at the University of Georgia. Nine experiments were conducted at the Southeastern Cooperative Wildlife Disease Study, and three were performed at other UGA laboratories.

Detection of Guinea worm DNA using qPCR and estimation of qPCR diagnostic parameters
Including the panels sent to external labs, all Guinea worm samples were detected as "positive" with no instances of "suspect" amplification; thus, the D Sn of this assay is 100%. Non-target samples, including host DNA, non-Dracunculus sp. nematodes, D. insignis, and D. ophidensislike DNA, did not amplify; thus, the D Sp of this assay was calculated at 100%.

Discussion
The primary objective of this work was to develop an assay that can provide rapid, genetic confirmation of suspect Guinea worm samples. The validation of a rapid molecular-based test that genetically confirms suspect Guinea worm samples allows for faster responses to reports of dracunculiasis and subsequently aids in implementing quick and effective control measures. Accordingly, the development and implementation of such a test are vital for advancing eradication and facilitating certification of dracunculiasis eradication. During the initial stages of assay development, we faced challenges due to the lack of available sequence data for our chosen gene targets, particularly from non-target Dracunculus spp. and other relevant nematodes, which made it challenging to determine in-situ assay specificity. While the cytb assay was designed to exclude D. lutrae based on available mitochondrial sequences (Fig 1), there were no reliable D. lutrae samples available for testing. Dracunculus lutrae has been shown to exhibit high mitochondrial (COI) genetic diversity [40]; however, considering the lack of D. lutrae cytb sequences available on Genbank at the time of assay design, it is unknown if this gene is equally diverse. Obtaining and testing confirmed D. lutrae samples in the future would help to clarify this point. Regardless, D. lutrae is considered a host-specific parasite of North American river otters (Lontra canadensis) that has not been found outside the United States and Canada [40]. Accordingly, the likelihood of D. lutrae being present in Africa is extremely low. However, detection of Guinea worm from new or unusual hosts or in new locations should be supported by morphological and sequencing confirmation methods.
Thus far, we have determined that this assay has not amplified any known parasitic nematode species of animals found in Guinea worm-endemic areas. Other Dracunculus spp. currently documented in Africa besides Guinea worm include snake-infecting species such as D. doi on Madagascar and D. dahomensis in Benin [35]. Unfortunately, sequences for these snake-infecting Dracunculus spp. are lacking, but we obtained D. ophidensis samples (presumptive identification based on adult female morphology, host, location, and 18S rDNA sequencing) from North American snakes that were included in our assay validation [32,35]. These D. ophidensis samples did not amplify with this assay; thus, it is likely that other snakeinfecting species would also be divergent enough from Guinea worm to not amplify; this has been seen with a previous 18S rRNA gene phylogenetic analysis, which included D. oesophageus from a European snake [41]. The Dracunculus sp. nematode (OPO28; GenBank accession no. MK085893.1) that most closely groups with the Dracunculus nematode recovered from a dog in Spain [26] also did not amplify. There are likely additional Dracunculus species not yet formally described [for example, the Vietnamese Dracunculus specimen recovered from a human; 42] for which there are no cytb sequences; therefore, continued validation of this assay with new Dracunculus spp. as they are discovered and verified would ensure the specificity of this assay.
The shift from infections only in humans to an increasing number of reported infections in domestic animals and wildlife complicates the eradication of Guinea worm [8,9,43]. The current system of confirming morphologically identified/suspect nematodes using Sanger sequencing is time-consuming and presents no opportunity to be adapted for use in the field. In contrast, the extremely high diagnostic sensitivity demonstrated by the GW qPCR assay described here lends strong support for its use in diagnostic laboratories, and it could be adapted for patient-side use through field-deployable qPCR technologies. Importantly, this assay was developed and validated using sections of worms or individual larvae. Additional validation would be needed should this assay be used to detect larvae within water samples/ copepods or fish or frog tissues, as we did not evaluate the impacts of sampling protocols or how the number of copepods within a sample or the amount of tissue present would impact the sensitivity of the assay.
This validated qPCR assay for the species confirmation of Guinea worm-suspect samples will aid the Guinea worm eradication efforts by increasing the rapidity of diagnosis and, therefore, implementation of control measures. In the future, this assay could be adapted for use with field-deployable real-time PCR technology to further advance rapid Guinea worm species confirmation in Guinea worm endemic areas. For example, this could help obtain a preliminary diagnosis of suspect worms of new regions or hosts, or to investigate the role of transport and paratenic hosts [20][21][22]36,44]. A field-deployable assay would further decrease the time from sample collection to species confirmation. Implementation of this highly sensitive and specific assay by Guinea worm diagnosticians will be extremely beneficial in the fight against Guinea worm disease.
Supporting information S1 Methods. The mitochondrial cytochrome c oxidase III (cox3) gene was tested as a possible target site for designing a qPCR protocol specific to Dracunculus medinensis. (DOCX) S1