Evaluation of Rapid Blood Sample Collection in the Detection of Circulating Filarial Antigens for Epidemiological Survey by rWbSXP-1 Capture Assay

Background Lymphatic filariasis is a neglected tropical disease leading to profound disfiguring causing socio economic burden in the tropics. Current diagnosis strategies available during field surveys and epidemics are based on traditional microscopic detections and a few antigen/antibody assays. We have compared different sampling methodologies and standardized the highly sensitive and reliable rWbSXP-1 antigen detection assay to our new sampling methodology. Methodology Samples collected as serum, whole blood, whole blood on filter paper and whole blood on microscopic slides from patients belonging to various clinical groups of filariasis [endemic normal(EN), chronic pathology(CP), microfilaraemic(MF) and non-endemic normal(NEN)] were collected and standardized the rWbSXP-1 antigen detection assay using monoclonal antibody raised against rWbSXP-1 protein. The whole blood collected on microscopic slide based sampling method was employed in the field and the presence of circulating filarial antigen (CFA) was assessed using the rWbSXP-1 assay. Principal Findings The sampling methods were compared and no significant difference was observed for the detection of CFA (MF, P = 0.304, EN, P = 0.675, CP, P = 0.5698, NEN, P = 0.4494). Further the optimized sampling method was utilized to collect the 1106 samples from Polur, Tiruvannamalai. The rWbSXP-1 assay gave 98 antigen positive results whereas the microscopic method gave only 17. Conclusions Four sampling methodologies were analyzed and the new sampling methodology of whole blood collected on microscopic slide was found to be convenient for the detection of CFA using rWbSXP-1 antigen detection assay. The 1106 samples from Polur were collected using the new method. The rWbSXP-1 antigen assay perceived a 7.32% increased result which was read as false negatives on the conventional microscopic staining method. This new sampling methodology coupled with the rWbSXP-1 antigen assay can be used in epidemiological surveys for lymphatic filariasis and the same sampling methodology can be expanded to other antigen based high affinity assays.


Introduction
Human Lymphatic filariasis (LF) is a neglected tropical infectious parasitic disease caused by lymph dwelling nematodes Wuchereria bancrofti, Brugia malayi, & Brugia timori. Amongst which Wuchereria bancrofti is the most dominant driving force for the infection in the sub continent [1]. The parasite is transmitted by arthropod vectors belonging to the genera Culex, Anopheles, Aedes, and Manosonia. Culex quinquefasciatus is accountable for 50% of transmission throughout the world [2,3]. The infection is acquired at early childhood but realization of the diseased state is often only after a late onset of disfiguring morbidity. The Global Program to Eliminate Lymphatic Filariasis (GPELF) initiated by The World Health Organization targets to eliminate the disease by the year 2020, but to accomplish such a feat, sensitive and reliable diagnostic tests are required for early clinical detections, field evaluations and post therapy monitoring [4]. The diagnosis of the infection during surveys and post infection is done by the thick microscopic smear stained with JSB or giemsa stains [5], which is always subjected to chance when working with infected individuals with low microfilaria (mf) density. Such cases require agitation of the parasite by using anti-filarial drugs, thus leading to underestimation of mf prevalence rates in epidemiological surveys [3,6,7]. Whereas the circulating filarial antigen (CFA) detections in Og4C3, ICT card, and other similar antigen, antibody tests prove to be more sensitive, efficient, quick, easy and cost-effective [8].
The ICT filarial antigen test and the Og4C3 assay are based on detection of adult worm antigens [9,10]. Later, detection assays utilizing the mf stage antigens were developed using various targets such as rWbShp-1, Bm14 and rWbSXP-1 [11,12,13].
In this study we have used a similar CFA detection by using the rWbSXP-1 antigen sandwich ELISA. The Wb-SXP-1 protein is expressed in the mf stage of the parasite. The rWbSXP-1 antigen used in our assay was seen to be highly sensitive and specific as a diagnostic tool and will provide an early detection for the filarial infection [14,15]. The study samples were collected through four different methods and detected by using the rWbSXP-1 antigen detection assay with monoclonal antibodies raised in mouse. The monoclonal antibody 1F6H3 which is highly reactive with brugian and bancroftian parasites were used as capture antibody and polyclonals raised in rabbit for rWbSXP-1 was used for detection as described previously [12]. We have evaluated the samples for the detection of CFA with the assay and derived at an optimized detection parameter for the rWbSXP-1 antigen detection assay to the samples collected as whole blood on microscopic slides, thereafter we have surveyed the filarial endemic village Polur of Tiruvannamalai district in Tamil Nadu, India using this new and easier mode of sampling and the results were compared with the conventional microscopic staining detection. before collecting the samples and conducting the field survey. The procedures followed were in accordance with the guidelines issued by the same department. Written and oral consent was taken from individuals before conducting the sampling. The institutional review board at the Centre for Biotechnology, Anna University, India also approved the protocols.

Human clinical samples for standardization
Samples from asymptomatic microfilaremics (MF) (n = 20) were collected from endemic regions of Tamil Nadu, India. The patients were identified as per the traditional thick smear microscopic data provided by the Department of Public Health and Preventive Medicine and Zonal Entomology Team, Vellore. Endemic normal (EN) (n = 20) samples were collected from individuals of the endemic villages with no mf present in thick smear microscopic test. Chronic pathology (CP) (n = 20) subjects with visible clinical symptoms of lymphedema were collected from Vellore and Tiruvannamalai districts. All the samples were collected by four different sampling methods. Samples were collected as 7 ml venous blood and 5 ml from it was used for separating the serum. The remaining 2 ml blood was stored separately for the whole blood analysis and 30-50 ml blood was used for the assay [16]. A single finger prick was made using a sterile steel lancet (Ghia Surgiblades Pvt. Ltd., Mumbai, India) and 20 ml of blood was absorbed on Whattman filter paper no. 3 in triplicates, air dried and stored. Samples were prepared as described by Hoti et. al. for the filter paper sample [17,18]. A new single 2-3 mm deep finger prick was made using the common diabetic test lancing device and sterile needles (Glucopro, NIPRO Corporation, Japan). A smear (100-150 ml) was made on a glass slide, dried and stored in slide boxes. After 3-4 hours the smear was resuspended with 150 ml of sterile phosphate buffered saline (PBS). The sample was transferred into sterile 1.5 ml eppendorf tubes and stored. This sample was used for the whole blood collected on microscopic slide samples and used for rWbSXP-1 assay experiments, 30 ml of this sample was used for the assay. All the samples were stored in 280uC until further use. The samples were collected in between 21:00 hrs and 23:00 hrs. Non-endemic normal (NEN) (n = 10) samples from healthy volunteers residing in non endemic area was provided by Dr. Murray Selkirk, Professor of Biochemical Parasitology, Division of Cell and Molecular Biology, Imperial College London, London.

Sample Survey
The study was conducted in Polur village of Tiruvannamalai district, Tamil Nadu, India under the supervision of corresponding healthcare officials from the Department of Public Health and Preventive Medicine, Govt. of Tamil Nadu. The survey samples were collected by using the lancing device and sterile needles (Glucopro, NIPRO Corporation, Japan). The samples were collected on glass slides in duplicates. One slide was used for the traditional thick smear microscopic detection by staining with JSB stain [5]. The other slide was used for resuspending the smear with 150 ml PBS and stored as described previously.

Recombinant Protein Expression and Purification
The recombinant clone of WbSXP-1 (Gen Bank Acc. No: AF098861) cloned in pRSETb was used in the study. The construct was transformed into salt-inducible Escherichia coli GJ1158 cells [19] and the recombinant antigen was expressed as a fusion protein with polyhistidine tag. The culture was grown up to ,0.6 optical density (OD) and the expression was induced by NaCl to a final concentration of 250 mM. The culture was allowed to grow for another 4 h at 36uC post induction. The cells were further pelleted by centrifugation and solubilised in binding buffer (100 mM Tris, 100 mM NaH2PO4, 200 mM NaCl, pH8.0). Cells were disrupted by sonication, centrifuged and the supernatant was subjected for purification. HIS-tagged rWbSXP-1 was purified by using Immobilized Metal Affinity Chromatography (IMAC)(Amersham, GE) on chelating sepharose matrix under non-denaturing conditions [20]. The expression and purification was analyzed on 12% sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and confirmed by immunoblotting with anti-HIS antibody [21].

Immunoreactivity of rWbSXP-1 protein
Immunoreactivity of the rWbSXP-1 was checked with clinical sera of MF, EN, CP and NEN using pooled sera from 10 human serum samples. The purified rWbSXP-1 antigen was separated on 12%SDS-PAGE, electrotransferred onto nitrocellulose membrane. After blocking overnight with 5% skimmed milk the membrane strips were incubated individually with (1:50) pooled sera from the different clinical groups. The bound antibodies were probed with alkaline phosphatase labelled goat antihuman IgG, and colour was developed with NBT/BCIP chromogenic substrate complex (ThermoFisher Scientific, Rockford, IL) [22].

Generation of antibodies for ELISA assay
Monoclonal antibody 1F6H3 synthesized earlier in our laboratory were used for the assay and 1F6H3 was identified to have more reactivity to native and recombinant mf antigen. It also had higher affinity and avidity to recombinant antigen [12]. The polyclonals were prepared by immunizing laboratory-bred rabbits with purified rWbSXP-1 protein [23]. Purified recombinant protein (250 mg) was emulsified in Freund's complete adjuvant and administered subcutaneously; consequently 125 mg of antigen was given in Freund's incomplete adjuvant. 10-13 days post final immunization the animals were bled and tested for immunoreactivity against the rWbSXP-1 by ELISA. The reactivity of the purified polyclonal was confirmed by western blots [12].
Detection of circulating filarial antigen by rWbSXP-1 sandwich ELISA Purified 1F6H3 monoclonal antibodies (1 mg/well in PBS) were coated separately on to each well of the high-binding 96 well microtiter ELISA plates (Thermo Scientific, Waltham, MA) and incubated at 4uC overnight as capture antibody. Blocking of the wells is done by 200 ml of 5% skimmed milk and incubating the plates at 37uc for 1 hr. Simultaneous PBS and PBST (0.05% Tween20 dissolved in PBS) washes were done followed by addition of corresponding patient sample (sera, whole blood, filter paper and slide method) and incubation for 2 hours at room temperature. Detection of CFA is attained by adding rabbit anti-rWbSXP-1 polyclonal antibody followed by goat anti-rabbit HRP conjugated secondary antibodies (Santacruz biotechnology, USA) with 1 hour incubation after adding each antibody. The plates were washed extensively and the bound peroxidase activity was detected using tetramethyl benzidine substrate. The reaction was stopped by using 1 M H 2 SO 4 solution. The results were articulated as mean absorbance at 450 nm for each group 6 standard deviation (OD6SD). Samples were identified as positive when the optical density was higher than the mean OD+3SD value of 10 NEN control sera [11].

Statistical Analysis
Statistical analysis for all the experiments was done using the GrapPad Prism software version 5.0 (GraphPad Software, San Diego, CA). Comparison of mean values was done using nonparametric Mann-Whitney analysis for two mean values, and one way analysis of variance (ANOVA) was used for comparison of more than two values. P#0.05 was considered to be statistically significant.

rWbSXP-1 Protein expression and purification
The recombinant Wb-SXP-1 protein was over expressed in salt inducible E.coli GJ1158, using NaCl at a final concentration of 250 mM. The protein was obtained at a molecular weight of 26 KDa ( Figure 1A) and confirmed by western blotting with anti histidine antibody after purification by IMAC ( Figure 1B).

Immunoreactivity of rWb-SXP-1 antigen
The immunoreactivity of rWb-SXP-1 was analysed by checking the reactivity of the purified antigen against pooled sera samples of MF, EN, CP and NEN (samples n = 10). The antigen showed

Sampling method Optimization
Four types of samples were collected from each individual and the samples were used for the standardization of the assay. The MF, EN, CP and NEN samples were identified by the traditional microscopic method and visible clinical symptoms. Samples such as serum separated from the venous blood, whole blood samples, samples collected on filter paper and samples from the slide method were collected as described earlier. Separate assays were performed for the EN, MF, CP and NEN groups with the above samples and the results were taken for optimising the rWbSXP-1 antigen detection assay based on the slide method.
The MF ( Figure 3A), CP ( Figure 3B), EN ( Figure 3C) and NEN ( Figure 3D) samples when analyzed with the rWbSXP-1 antigen detection assay showed no significant difference in their OD values, when comparing the slide method samples with the sera and whole blood samples (Table 1). Four EN samples were detected as positive by our assay. These four samples were positive for all the four types of sampling methods. The positivity of these four samples was confirmed by Signal-MF filarial detection kit (Span Diagnostics, Gujarat, India) [24].
Using this data a standard plot for rWbSXP-1 detection assay for the slide method samples was developed (Figure 4). The antigen levels in MF group samples were significantly higher when compared with EN, CP and NEN groups with a p value of , 0.0001( Table 2). The cut off value of 0.3477 for the antigen positive samples were calculated and assigned using the formula NEN+3SD. Since the samples collected on microscopic slide were comparable with the samples collected by other methods, the same was used for field evaluation of antigen detection assay.

Field Survey
The endemic village Polur from Tiruvannamalai district of Tamil Nadu was taken for the field survey. Samples were collected in duplicates by using the optimized slide method of sample collection. A total of 1106 samples were collected of which 525 (47.5%) were females. The age of the individuals selected for sample collection was between 3-85 years. Further detection by traditional microscopic staining method and rWbSXP-1 antigen detection from samples collected by slide method was performed (Table 3).

rWbSXP-1 Detection Assay for evaluation of field samples
rWbSXP-1 detection assay for the samples collected by slide method was carried out and the samples with a OD value higher than the cut off (0.3477) was considered as positive. A total of 17/ 1106 (1.54%) samples were identified as mf positive individuals by the conventional microscopic staining method, the mf density was within the range of 40-380 mf/ml with a geometric mean of 5.73 in the infected blood. Whereas 98/1106 (8.86%) was identified as positive by the rWbSXP-1 assay. The 17 positive samples identified by traditional method was also CFA positive for the rWbSXP-1 assay amongst the additional 81 (7.32%) samples ( Figure 5, Table 3). No mf was detected in any of the night blood smears examined from subjects who were negative in the rWbSXP-1 assay.

Discussion
The effective detection and control of filariasis is anchored on the identification of infection and assessment of the disease. This was thought to be achieved completely by the traditional microscopic staining method, although light was shed into the fact that microscopic detection is not completely reliable and feasible as there were difficulties in sampling, in terms of, time of collection, segregating the large volume of samples, and most importantly the finding of mf in blood smear was based on chance. It depends on factors such as the stage of infection, collection time  and density of mf in the patient. Later detections based on CFA were introduced and the Immunochromatographic card test (ICT) and Og4C3 ELISA became the limelight for filarial detection [9,25], while this CFA detection had their advantages, sensitivity and storage issues were noted. Other detection assays based on filarial antibody detections (WbSXP, Bm14, BmR1) were also developed [15,24,26]. Although antibody detections were developed for the detection of nematode infections, the antigen based detections proved to be more informative than antibody methods [27], due to the fact that endemic controls often harbour high titres of antibody than the infected people [28]. Amongst these CFA detection assays a very high affinity WbSXP-1 assay was introduced by us using monoclonal and polyclonal antibodies specific to the parasite [12]. In all the above techniques the prime issue faced is the inability to collect and process large volumes of samples from the field during surveys or an epidemic. Most sampling is done by means of finger pricking for the microscopic staining and from collecting venous blood to separate the serum for detection of CFA. Commercially available detection methods use field samples collected in the form of filter paper or the whole blood for detection of CFA [16]. The collection of samples for detection of CFA from the serum has its restriction as the collection, segregation and storage of such samples are very hard. A recently developed rapid test (Alere Filariasis Test Strip) for detecting CFA in human blood detects the infection within 10 minutes, whereas gives late positives after 24 hours [16]. Primarily during large scale field surveys and endemic outbreaks when the sample volume will be very high and in thousands, it is time consuming to wait for 10 minutes at each home and conduct tests as per the manufacturer's instruction. Thus when working with large volumes of samples, collection of samples from the field and evaluation at a later time is required.
Therefore in this study we have made an attempt to develop an easy, quick, reliable, cost effective and sensitive sampling and detection technique. Using this method we can conduct large scale surveys and detection of the infection in quick notice, and with minimal hardships. The slide method described in our study is a simple technique that requires not more than 3 people for conducting the sampling and detection. It is a less invasive, time saving, inexpensive technique that does not require a trained specialist for the collection and storing of the samples [4,6]. Also since the needle and the lancing devices are put into use, it is less likely that the pricking process will leave painful injuries on the fingers of the individuals tested. Unlike night blood surveys, the samples can be collected and stored with ease throughout the day, since it is CFA detection.
For the current study we have used the rWbSXP-1 assay , previously described by us which has high sensitivity with clinical sera of Wb (90.8%) and Bm (91.4%), and has no cross reactivity with Oncocerca volvulus infected sera [12,24]. The purified rWbSXP-1 antigen showed reactivity with MF pooled sera and no reactivity with CP, EN and NEN sera while checked on immunoblots. The four different sample collection modes when analyzed with the rWbSXP-1 assay have shown similar or mild variations in their optical density, which makes it clear that the samples collected by  the slide method has sensitivity on a par with sera samples. While comparing the P value of sera based rWbSXP-1 detection versus samples collected by the slide method it was seen that for the MF (P = 0.3040), EN (P = 0.6750), CP (P = 0.5698), NEN (P = 0.4494) samples the values were not significant. It was also noted that the OD values of the samples analyzed are not significant with the OD values of whole blood detections ( Table 1). The filter paper methodology of sampling has shown less sensitivity for the rWbSXP-1 assay, similar incompetency of the filter paper samples with respect to sensitivity and specificity were previously reported [18,29]. During our optimization we have found that four individuals who have been characterised as endemic normal by traditional microscopic method were having the rWbSXP-1 antigen in their system. This clearly entails that the traditional microscopic method failed to prove its reliability by providing with false negative results. A low density of mf in the system or prepatent infection can lead to CFA in the system but lack mf in the blood smear or any clinical symptoms [24], similar findings were reported with microscopic detections earlier [18,30,31]. This cryptic infection which is overlooked during field surveys [32,33] and control of infection leads to the failure of MDA and eradication of filariasis. Such cases that have escaped the detection will further proliferate the infection by substituting as fresh reservoirs for transmission. Thus a conclusion can be drafted that the slide method can be used as an alternative for the surveying and infection analysis using rWbSXP-1 assay.
In the second part of our study we have conducted a field survey of Polur village, from Tiruvannamalai, Tamil Nadu with the new slide based sampling method and analyzed the same with rWbSXP-1 assay. A total of 1106 samples were collected among which 581 males and 525 females were analyzed by both, traditional microscopic staining method and the rWbSXP-1 assay. In our results we have found that 1.54% samples were only identified as mf positive in the conventional thick smear microscopic staining method whereas 8.86% was identified as positive by the rWbSXP-1 assay. The 17(1.54%) samples identified as mf positive were also CFA positive in the rWbSXP-1 assay. The 7.32% increased positivity in our assay could be due to cryptic or occult infection in the patients. Since CP patients can eliminate the SXP antigen from the system owing to the Th1 response they possess, they were detected as negatives in our assay [12,34,35].

Limitations
To confirm the statement of cryptic situation, qPCR measures could give an insight, whereas to do the same genomic dna  extracted from mf is required, which was not present in the slides or the blood obtained from antigen positive individuals. Similarly in regions where co-infection of loasis is seen, this method might not be suggested as cross reactivity with Loa loa serum of SXP antibody detection was noted earlier. However since loasis is not co-endemic with brugian and bancroftian infections in most endemic countries, this could not cause any problem [24].

Conclusions
To summarize four sampling methods were analyzed and an easy, cost effective method which is less invasive, mass survey friendly and reliable was identified and optimised for the high affinity rWbSXP-1 antigen detection assay. Further the endemic village of Polur, Tiruvannamalai, Tamil Nadu was surveyed with the new sample collection method and assayed using the optimized rWbSXP-1 assay. This methodology of sample collection and assay can be further employed in large scale surveys and detections owing to the merits it poses towards sampling and storage, without compromising the sensitivity and reliability of the detection. The survey can be further performed in other endemic areas for filarial infections and also can be used to identify the cryptic infections amongst the population. It can also be used as a yardstick to assess the state and volume of infection in EN populations where the infection is claimed to be absent. Thereafter MDA programmes can be administered long before clinical manifestations or a widespread endemic occurs. The sampling methodology can be further diversified into other parasitic infections but this cannot be discussed in detail until further research is done on this subject.

Supporting Information
Dataset S1 ELISA standardization data. The ELISA data used for comparing different methods of sampling and standardizing the rWbSXP-1 assay to the slide based sample collection method.