A miniPCR-duplex Lateral Flow Dipstick Platform in Combination with the Microuidic Device for Rapid and Visual Diagnosis of Lymphatic Filariae Infection

Background: Lymphatic lariasis is a major neglected tropical disease that is a leading cause of permanent and long-term disability worldwide. Signicant progress made by the Global Programme to Eliminate Lymphatic Filariasis (GPELF) has led to a substantial decrease in the levels of infection. Accordingly, access is now needed to a cost-ecient, non-laborious diagnostic assay(s) capable of detecting low levels of microlariae and without the need for infrastructure and expensive equipment. Methods: We developed a diagnostic assay that combines a miniPCR with a duplex lateral ow dipstick (DLFD) assay. The design of the PCR primers was guided by sequence alignments of the HhaI and SspI genes of Brugia malayi and Wuchereria bancrofti, respectively. The limits of detection, accuracy, and cross-reactivity of the assay were evaluated, after which the assay was deployed to assess infection using blood samples obtained from Myanmar study participants living along the Thailand-Myanmar border, a region in Tak province endemic W. bancrofti. In addition, blood samples were provided by from Thais in Narathiwat province residing in an area endemic for brugian lariasis. We also combined a previously reported semi-automated microuidic device with the miniPCR-DLFD to facilitate rapid detection and species identication of microlariae in human blood. Results: The miniPCR-DLFD assay exhibited a detection limit of two microlariae per mililiter blood sample, and cross-amplication was not observed with from other parasites. For eld validation, microlariae and DNA of W. bancrofti were detected from two (0.6%) and ve (1.5%) out of 328 (100%) blood samples, respectively. For the Narathiwat samples, microlariae and DNA of B. malayi were detected from one (0.46%) and two (0.9%) out of 216 (100%) blood samples, respectively. A rapid and visual of species identication was accurately obtained in all microlariae entrapped by our previously developed microuidic device. Conclusion: A miniPCR-DLFD platform alone or coupled with the microuidic device provided rapid and visual diagnosis of lymphatic lariasis. The microuidic device, the miniPCR, and the DLFD are all portable. Coupled with a portable microcentrifuge for DNA extraction, this is a promising platform for the diagnosis of lymphatic lariasis in point-of-collection settings with a modest cost (~USD $ 5) per sample.


Introduction
Lymphatic lariasis (LF), caused by lymphatic parasites, Wuchereria bancrofti, Brugia malayi and B. timori, is one of the most neglected tropical diseases. LF is a mosquito-borne disease and remains a major public health concern, especially in Asia, Africa, the Western Paci c, and parts of the Caribbean and South America [1][2][3].
In addition to Giemsa staining for micro laria, a variety of detection methods such as antigen detection, antibody detection and molecular diagnosis have been developed to improve diagnostic performance and utility [4,5,6]. An immunochromatographic test (ICT) for the detection of W. bancrofti circulating antigen is available commercially [6]. ICTs are easy to perform and their results are easily interpreted.
Nevertheless, larial antigenemia that persists for years, even after treatment, limits the value of ICTs, as evidenced by follow-up studies post-treatment [7][8][9]. A few articles have reported new tests that can detect circulating antigens from Brugia malayi in human blood. However, those tests have not yet been independently veri ed, and none are commercially available [10][11][12][13][14]. Anti larial IgG4 antibody detection has also been developed [5,6,15], and it is a useful addition to the limited array of brugian lariasis diagnostic tools available. However, this antibody test does not discriminate between larial species, and antibody titres fall only slowly following anthelmintic treatment [16].
As the Global Programme to Eliminate Lymphatic Filariasis moves towards the achievement of its goal, the prevalence of LF and the rates of disease transmission have decreased substantially [17,18]. The success of the programme has led to dramatic reductions in both micro laremia and antigenemia levels in those countries that have completed multiple rounds of mass drug administration [19]. Paradoxically, with the steady decline in the disease levels, there is an increasing need for more sensitive and speci c diagnostic assays capable of detecting low levels of parasites [20][21][22]. In addition, proactive monitoring for LF in historically endemic areas is necessary in order to eliminate the disease [1,17,23]. Thus, there is a need to develop a reliable, rapid, user-friendly, less labour-intensive detection method of LF diagnosis that has the capability to screen for, and identify, causative species [4].
As to molecular diagnosis for LF, PCR assays have the advantage of detecting active infections.
Moreover, PCR detects infections with nocturnally periodic strains of lariae in daytime blood collections [20,22,24], and PCR species identi cation is highly reliable [21,24,25]. The disadvantages of standard PCR are that it is time-consuming, requires dedicated laboratory instruments and reagents, and is available at only some facilities.
Several compact PCR-based methods and devices have been validated for use for point-of-collection detections [26,27]. The miniPCR thermocycler represents an attractive and affordable device with the potential for use at collection sites in endemic-country settings. This user-friendly and portable instrument is commercially available at a modest price [26,27]. However, the miniPCR still requires gelanalysis steps involving gel electrophoresis and imaging. In order to obviate these latter steps, several studies have proposed the use of a PCR-nucleic acid lateral ow immunochromatographic assay (PCR-NALFIA) using a lateral ow dipstick (LFD) [27,28]. LFD-based assays can detect speci c DNA products in as little as 10 minutes [27, 28].
The present study proposed for the rst time to developed a miniPCR assay coupled with a duplex lateral ow dipstick (DLFD) to detect both B. malayi and W. bancrofti DNA in samples of human blood. We also applied this miniPCR-DLFD in combination with our recently developed semi-automated micro uidic device [29] to facilitate mass screening and rapid visual species identi cation of micro lariae of lymphatic lariae in human blood.

Ethics approval
Before commencement of the study, ethics approval was obtained from the Ethics Committee of the Faculty of Medicine Siriraj Hospital, Mahidol University (approval number, Si129/2016). The research also fully complied with the ethics principles and guidelines for human experimentation issued by the National Research Council of Thailand. The formal consent of the participants was obtained verbally.
Extraction of DNA from blood samples DNA was extracted from 50 µL of ethylenediaminetetraacetic acid blood samples using the High Pure PCR Template Preparation Kit (Roche Diagnostics GmbH, Penzberg, Germany) according to the manufacturer's protocol. Following extraction, the DNA was eluted in 100 µL of elution buffer and stored at -20º C until use. The DNA concentration was determined using a Nanodrop 1000 spectrophotometer (Thermo Fisher Scienti c Inc, Waltham, Mass., USA) before the DNA was subjected to PCR and miniPCR analysis.

Primer design and optimization of standard PCR condition
The primer designs were based on alignments of the HhaI and SspI repetitive non-coding DNA sequences of B. malayi and W. bancrofti, respectively [30,31,32]. These regions contain species-speci c sequences that are useful for identifying B. malayi and W. bancrofti. The HhaI and SspI sequences were obtained from the National Center for Biotechnology Information database (GenBank accession numbers M12691.1 for B. malayi and L20344.1 for W. bancrofti). Fig. 1 presents the sequences of the forward and reverse primers. A BLAST analysis (National Center for Biotechnology Information, Bethesda, Md., USA) was performed to check the speci city of the primers to the target sequences. The presence of the 145-bp and 182-bp ampli cation products of HhaI and SspI, respectively, was veri ed by 2% agarose gel electrophoresis.
The PCR ampli cations were carried out in a volume of 20 µl, consisting of 10 µl of PCR master mix (Quantabio; Qiagen Beverly, LLC., Mass., USA); 0.2 µM of each forward and reverse primer; 7.2 µl of dH 2 O; and 2 µl of DNA template. Nuclease-free water was used as a negative control. DNA from the blood samples containing mf of B. malayi and W. bancrofti were used as positive controls.
To optimize the PCR conditions, a gradient PCR was performed using the Veriti 96-Well Thermal Cycler (Applied Biosystems, Thermo Fisher Scienti c Inc, Waltham, Mass., USA) with the annealing temperature ranging from 55-62º C for both sets of primers. At an annealing temperature of 56° C, both set of primers (HhaI and SspI) ampli ed clearly detectable products. The optimized ampli cation conditions for both primer sets included an activation step at 95° C for 5 min, followed by a 30-step ampli cation of 30 s at 95° C, 30 s at 56° C, and 30 s at 70° C, with a last step at 70° C for 5 min.

Assay conditions to validate the miniPCR
The PCR assays for B. malayi and W. bancrofti were separately performed using a miniPCR instrument (DBA miniPCR bio; Amplyus LLC., Cambridge, Mass., USA), employing the same reagents and conditions used for the standard PCR, as noted above. For use in a DLFD, the 5' ends of the designed forward primers of HhaI and SspI were labelled with uorescein isothiocyanate (FITC). The 5' ends of reverse primers of HhaI and SspI were labelled with digoxin (DIG) and biotin, respectively. The sequences of the primers and their corresponding amplicons are presented in Fig. 1.

Construction of DLFD for nucleic acid detection
A DLFD is composed of four parts: the sample pad, conjugate pad, nitrocellulose membrane and absorbent pad. The sample pad is pretreated with buffer, and it can offer suitable pH and ion strength for the detection. The conjugate pad is used for the storage of reporter molecules (colloidal gold conjugated mouse anti-FITC). For the development of our DLFD, anti-digoxin (anti-DIG; Test line 1) and streptavidin (Test line 2) were sprayed on the nitrocellulose membrane to create test zones using the IsoFlow Reagent Dispenser (Imagene Technology, Inc., Hanover, N.H., USA), whereas anti-mouse antibody was sprayed on the nitrocellulose membrane to form a control zone (control line) by the AirJet Quanti 3000 Nanoliter aerosol dispenser (BioDot, Inc., Irvine, Calif., USA). The membrane was then dried at 37° C for 12 hours. The nitrocellulose membrane was attached to the central part of an adhesive plate. The lateral ow dipstick was then assembled, as shown in Fig 2; a schematic illustration of the DLFD is also presented in that gure.

Optimization of anti-DIG and streptavidin concentration
To optimize the concentration of anti-DIG, various concentrations of anti-DIG (0.5 µg/strip, 0.75 µg/strip or 1 µg/strip; Test line 1) were sprayed on the nitrocellulose membrane to optimize the concentration of anti-DIG, using the IsoFlow Reagent Dispenser (Imagene Technology, Inc.). The concentrations were at 1 µl per mm. The lateral ow dipstick was then assembled. Various concentrations of streptavidin (i.e., 0.5 µg/strip, 0.75 µg/strip, or 1 µg/strip) (Test line 2) were sprayed on the nitrocellulose membrane to optimize concentration of streptavidin using the AirJet Nanoliter aerosol dispenser (BioDot, Inc.). The assay was performed using positive (DNA of B. malayi) and negative (dH 2 O) controls.

DLFD for detection of ampli cation products
To detect the ampli cation products of each sample, 1 µl of ampli cation product from each set of primers (speci c to HhaI and SspI) was added into a well of the 96-well plate containing 100 µl of sample buffer. The dipstick was placed into the well vertically, and the reaction was read within 5-10 min. The appearance of positive, pink-coloured lines was observed using the naked eye on both the test and control lines. With respect to negative results, the pink-coloured line was apparent solely on the control line.

Detection limit of miniPCR-DLFD
To study the detection limit of the miniPCR-DLFD, mf of B. malayi was spiked into ethylenediaminetetraacetic acid blood specimens obtained from healthy subject, as listed in Table 1. Four sets of the samples were prepared (sample IDs 1-24). The DNA of B. malayi as well as negative blood samples were used as positive and negative controls. DNA that had been extracted from the blood samples underwent ampli cation using the standard PCR as well as the miniPCR, followed by DLFD assay.
Comparison of the Giemsa-stained thick blood smears, standard PCR based ampli cation and the miniPCR-DLFD assay The study population consisted of 10 subjects positive for B. malayi mf, 14 subjects positive for W. bancrofti mf, and 50 healthy controls. Blood was also collected from the 14 W. bancrofti mf positive subjects on day 7 following treatment with diethylcarbamazine and albendazole treatment ( Table 2). Giemsa staining of thick blood smears, standard PCR ampli cation, and miniPCR-DLFD assays were performed.

Page 7/20
For eld validation of the miniPCR-DLFD, a total of 216 blood samples were collected from Thai residents in Narathiwat Province, an area endemic for brugian lariasis. A further 328 blood samples were obtained from Burmese immigrants who resided along the Thailand-Myanmar border in Tak Province, which is a region endemic for W. bancrofti ( Table 2). All of the blood samples from these study participants were collected during the daytime.
Micro uidic device coupled with miniPCR-DLFD Detection of micro lariae was undertaken in all study samples using the micro uidic device, as described [29]. Brie y, 50 μl of blood was suspended in 150 μl of lysis buffer A, mixed, and then incubated at room temperature for 10 min. The sample was uploaded for analysis by drawing buffer B, 15 μl, into a one ml syringe that connects to the adapter and to the inlet port. The pump was started and the sample solution introduced into the micro uidic device via the inlet port. The micro lariae were trapped in the micro uidic chip while the remaining solution exited the micro uidic chip via the outlet port into the waste tube. The trapped micro lariae were inspected by light microscopy at 10x magni cation. To identify species of the trapped mf in the micro uidic chips, 100 μl Tris-EDTA buffer was dispensed into the chip via the inlet port. The chip was placed on a hot plate at 56°C for 15 min, after which the solution containing the trapped mf was withdrawn through the outlet port and transferred to a 1.5 ml Eppendorf tube. The tube was subjected to centrifugation at 15,520× g for 10 min. The supernatant was discarded, and DNA was extracted from the pelleted material using the Roche high pure PCR template preparation kit, according to the manufacturer's instructions (Roche Diagnostics GmbH, Penzberg, Germany). Species identi cation was performed by HhaI/SspI gene ampli cation by miniPCR (Amplyus) followed by use of the DLFD testing using the above-mentioned miniPCR assay primers and employing the same reaction recipes and conditions as described above.

Optimization of anti-DIG and streptavidin concentration
The optimal concentrations of anti-DIG and streptavidin which showed clearly visible results for test line 1 (B. malayi) and test line 2 (W. bancrofti) of the DLFD were 0.75 µg anti-DIG/strip and 1 µg streptavidin/strip, respectively. The optimal amount of PCR products from the miniPCR that yielded a clear band on the DLFD strip was 1 µl of ampli cation product from each set of primers (HhaI and SspI) in 100 µl of sample buffer. Clearly detectable, positive, pink-coloured lines were observed within 10 minutes.
The miniPCR-DLFD speci city As illustrated in Fig. 3A, the DNA from 11 other parasites were all negative by the miniPCR-DLFD assay.
Page 8/20 The detection limit of the miniPCR-DLFD The miniPCR-DLFD still detected a positive band in the samples numbered 17-20, each of which contained 2 mf per millilitre of blood ( Fig. 3A and Table 1). Table 1 The sample ID, number of spiked micro laria (mf) mf of Brugia malayi in a one millilitre blood sample, concentration of extracted DNA (ng/µL), and outcome of the miniPCR-DLFD assay and standard PCR with gel electrophoresis as positive (+) or negative (-). Comparison of the Giemsa-stained thick blood smears, standard PCR based ampli cation and the miniPCR-DLFD assay For all 24 mf positive blood samples as well as the 50 controls tested, the standard PCR results agreed with the miniPCR results using both test strip-based detection and electrophoresis-based detection ( Table  2). As to the 14 W. bancrofti mf positive subjects who received diethylcarbamazine and albendazole treatment, mf and DNA of W. bancrofti were detected in four of the 14 blood samples. Only DNA was detected in two of the post-anthelmintic treatment blood samples. No mf nor DNA of W. bancrofti were detected from the remainder of the samples (Table 2). Figure 4 presents the ndings obtained with the tested samples.

Field validation
Regarding the eld validation in Narathiwat Province, mf and DNA of B. malayi were detected in one of the 216 blood samples, whereas only DNA of B. malayi was detected in one of those samples. As to the 328 blood samples collected from Tak Province, mf and DNA of W. bancrofti were detected in two of the samples, while only DNA of W. bancrofti was detected in three of the samples (Table 2). Table 2 Comparison of the standard PCR-based ampli cation, miniPCR-DLFD based ampli cation and detection of micro lariae (mf) detection by stained thick blood smear for diagnosis of infection with Wuchereria bancrofti and Brugia malayi in study blood samples Samples from eld study in Tak Province  328  2  326  5  323  5  323 Micro uidic device coupled with the miniPCR-DLFD After processing of the micro laria detection using the micro uidic device, mf of the 24 mf-positive human blood samples were trapped in the detection zone within the micro uidic chips. Species of the entrapped mf were further identi ed by the miniPCR-DLFD. The 10 samples were identi ed as B. malayi mf, and 14 were con rmed to be W. bancrofti (Fig 5).

Discussion
In the present study, we developed a miniPCR-DLFD for rapid and visual diagnosis of lymphatic lariasis. The duplex LF detection is another format of lateral ow dipstick that can be used for detection of more than one target larial species, and the assay is performed on a strip containing test lines equal to the number of target species to be analysed. In addition, DLFD exhibits increased sensitivity for PCR product detection compared with gel electrophoresis, and it takes less time [28,33,34].
The detection limit of the thick blood smear staining technique for mf detection is at least 16-20 mf in one ml of blood [24,25]. The result obtained from the present study, as the detection limit of the miniPCR-DLFD was two mf of B. malayi per ml of blood sample. This is 10-fold more sensitive than thick blood smear staining. Moreover, almost all species and strains of lymphatic lariae worldwide display nocturnal periodicity/sub-periodicity, given that samples collected during the daytime show a reduced sensitivity for the detection of mf [8]. Notably, of all 14 cases from whom blood samples were collected post-treatment, W. bancrofti DNA was still detected in six infected study participants whereas mf was detected in only four cases. These data encourage the usage of the miniPCR-DLFD as a sensitive tool for follow-up after treatment.
A small, portable, miniPCR instrument was used in this study to replace the need for a bulkier, standardsized thermocycler. The portable miniPCR from Amplyus used by the current investigation is diminutive in size (5.1 x 12.7 x 10.2 cm) and not heavy (450 g), and it only requires 100-240 V (AC) and 50-60 Hz, 90 W to perform a PCR run of 16 ampli cations/round. Furthermore, its cost (~ US $800) is not prohibitive. Taken together, these features make it an attractive and reliable alternative tool for point-of-care. With respect to cost, the reagents for the DNA extraction and miniPCR steps cost about $3/sample, and one DLFD strip costs $2. Thus, the total cost of the miniPCR-DLFD assay is of the order of $5 per sample. One limitation of using the miniPCR-DLFD for point-of-collection settings, however, is that the DNA extraction steps require somewhat cumbersome instruments including a high-speed microcentrifuge. We suggest that use of a portable microcentrifuge, e.g., WiseSpin CF-10 Diahan-Scienti c, which weights ~ 315 5 kg, We have previously developed a semi-automated micro uidic device for rapid, higher-throughput detection of micro lariae in animal blood. We next ]performed a real time PCR with high resolution melting analysis (HRM-real time PCR) for species identi cation of the trapped mf [29]. The e ciency of the device was evaluated in laboratory and it was applied to study the prevalence of canine and feline micro laremia [29,37]. In this study, we applied the micro uidic device to detect micro lariae of B. malayi and W. bancrofti in human blood. We also replaced the HRM-real time PCR with the miniPCR-DLFD for the species identi cation step in order to reduce the need for infrastructure and expensive equipment. We posit that use of the micro uidic device, which is portable and weights only 3 kg, coupled with the miniPCR-DLFD can enable population-scale, mass screening for LF in endemic areas.

Conclusions
To overcome the impediments to facilitate rapid diagnosis for LF at the collection site, associated with

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.