Journal of Microbiological Methods 108 (2015) 19–24

Contents lists available at ScienceDirect

Journal of Microbiological Methods journal homepage: www.elsevier.com/locate/jmicmeth

A sensitive enhanced chemiluminescent-ELISA for the detection of Plasmodium falciparum circumsporozoite antigen in midguts of Anopheles stephensi mosquitoes Bryan Grabias a, Hong Zheng a, Godfree Mlambo b,c, Abhai K. Tripathi b,c, Sanjai Kumar a,⁎ a Laboratory of Emerging Pathogens, Division of Emerging and Transfusion Transmitted Diseases, Office of Blood Research and Review, Center for Biologics Evaluation and Research, Food and Drug Administration, Rockville, MD USA b The Johns Hopkins Malaria Research Institute, Bloomberg School of Public Health, The Johns Hopkins University, Baltimore, MD 21205, United States c Department of Molecular Microbiology and Immunology, Bloomberg School of Public Health, The Johns Hopkins University, Baltimore, MD 21205, United States

a r t i c l e

i n f o

Article history: Received 4 August 2014 Received in revised form 10 October 2014 Accepted 10 October 2014 Available online 22 October 2014 Keywords: ELISA Plasmodium Oocyst Malaria Vector screening

a b s t r a c t Efforts to develop a successful malaria vaccine are hampered due to lack of assays that are predictive of protective immunity without conducting large clinical studies. The effect of experimental vaccines and drugs on malaria transmission is yet more difficult to measure. Knowledge on the Plasmodium infection rate in mosquito populations will aid the measurement of effects from intervention measures for malaria control. Here, we report the development of a chemiluminescent sandwich ELISA (ECL-ELISA) that can detect Plasmodium falciparum circumsporozoite protein (Pf CSP) produced in recombinant form at concentrations of 4.4 pg and in P. falciparum sporozoites (Pf SPZ) derived from mosquito salivary glands at levels corresponding to 5 Pf SPZ. Most importantly, we demonstrate reliable Pf CSP-based detection of 0.056 day 8 P. falciparum oocysts developing inside mosquito midguts in whole mosquito lysates. Cumulatively, the ECL-ELISA is 47× more sensitive for the detection of Pf CSP than a colorimetric ELISA while greatly simplifying sample preparation, obviating the need for cumbersome midgut dissections and allowing high throughput screening of Plasmodium infection in mosquito populations. The ECL-ELISA may also have broader application in diagnosis of infectious diseases and the prognostic value in cancer and other diseases such as auto-immunity and genetic disorders based on antigen detection, or quality validation of biological vaccine components. Published by Elsevier B.V.

1. Introduction Malaria is a persistent threat to global public health with an estimated 219 million cases and 627,000 deaths annually (WHO, 2012). Successful propagation of the causative Plasmodium parasites requires a developmental cycle in Anopheles spp. mosquito vectors before being transmitted to the human host through an infectious bite. This critical stage of the parasite life cycle represents a bottleneck or vulnerability in the growth and survival of Plasmodium (Wang and Jacobs-Lorena, 2013) and intervention strategies that specifically target the mosquito host to disrupt the parasite transmission are being designed (Fang et al., 2011; Parvez and Al-Wahaibi, 2003; Rajakumar and Abdul Rahuman, 2011). Such strategies include the novel drugs and vaccines that target the sexual stage of life cycle and application of insecticideAbbreviations: ECL-ELISA, enhanced chemiluminescent ELISA; rPf CSP, recombinant Plasmodium falciparum circumsporozoite protein; SPZ, sporozoites. ⁎ Corresponding author at: 10903 New Hampshire Ave. Bldg 52/72 Rm 5304 Silver Spring, MD 20993, USA. Tel.: +1 301 827 7533. E-mail addresses: [email protected] (B. Grabias), [email protected] (H. Zheng), [email protected] (G. Mlambo), [email protected] (A.K. Tripathi), [email protected] (S. Kumar).

http://dx.doi.org/10.1016/j.mimet.2014.10.006 0167-7012/Published by Elsevier B.V.

treated nets, residual sprays inside residence and disruption of mosquito habitats. However, the success of any of these mosquito-borne transmission blocking strategies would depend on the availability of effective methods to measure the effect of intervention measures on parasite infection rate in mosquito populations in endemic areas. Outside of traditional microscopy, the enzyme-linked immunosorbent assay (ELISA) is a commonly used method to screen biological samples for the presence of infectious agents. Indeed, ELISAs are routinely employed to screen for pathogens by the detection of specific antigens and measure antibody responses in patients (Cao et al., 2014; Li et al., 2014), monitor physiological responses to drug treatments (Pereira et al., 2014) and screen for chemical contaminants in the food supply (Quan et al., 2011). The overall sensitivity of these assays is typically limited by the affinity of the antigen–antibody interaction as well as the nature of the reporter system used to quantify the analyte under investigation. A number of studies have established the effectiveness of diagnostic calorimetric ELISA assays in the detection of specific Plasmodium antigens in blood, such as histidine rich protein 2 and lactate dehydrogenase, at levels corresponding to parasite densities of approximately 1–20 parasites/μL or approximately 0.08-2 ng protein/ μL (Atchade et al., 2013; Bashir et al., 2013; Martin et al., 2009), and

20

B. Grabias et al. / Journal of Microbiological Methods 108 (2015) 19–24

with diagnostic sensitivities and specificities in the 90%–100% range (Atchade et al., 2013; Noedl et al., 2006). Similarly, a few studies have reported on the Plasmodium falciparum circumsporozoite (Pf CSP)based ELISA for detection of Plasmodium parasites in mosquito vectors (Burkot et al., 1984; Collins et al., 2004). More specifically, standard calorimetric ELISA experiments directed against Pf CSP utilizing whole mosquito lysates or dissected salivary glands have demonstrated detection limits of 3–50 sporozoites/μL (Fontenille et al., 2001). Ongoing optimization of the calorimetric ELISA protocol has further increased the sensitivity of such assays to 0.25 sporozoites/μL, specifically when using purified sporozoites that have been isolated from dissected mosquito salivary glands (De Arruda et al., 2004). In the absence of proper dissection of mosquito glands and midguts before ELISA analysis, a high diagnostic sensitivity of 100% (Fontenille et al., 2001) can be reduced to 71.6% (Beier et al., 1987). In this study, we report the development of an enhanced chemiluminescent ELISA (ECL-ELISA), which is capable of detecting as few as 5 P. falciparum sporozoites (Pf SPZ) and identifying 0.056 Day 8 P. falciparum oocysts (Pf Oocysts) directly from whole mosquito lysates. Briefly, the ECL-ELISA is a sandwich ELISA in which immobilized antiCSP monoclonal antibody (mAb) clone 2A10 captures available Plasmodium CSP antigen. Antibody-bound CSP is subsequently detected and quantified utilizing a biotinylated form of the 2A10 mAb and an avidin-conjugated HRP reporter system. The ECL-ELISA described here obviates the need for dissection of mosquito midguts and salivary glands while simultaneously providing high assay sensitivity and specificity. The performance characteristics and the broad accessibility and ease of use provided by the ECL-ELISA may facilitate the establishment of a comprehensive vector surveillance program and find application as a general diagnostic tool for the detection of other pathologically relevant antigens or host biomarkers of cancer and other disease as well as a critical high throughput validation assay in vaccine development and manufacturing.

2. Materials and methods 2.1. Recombinant Pf CSP Recombinant Pf CSP (rPf CSP) and Py CSP (rPyCSP) antigens were prepared as described previously (Kumar et al., 2013). Briefly, the amino acid sequence 27-123[NANPNVDP] 3[NANP] 21300-411 of P. falciparum (3D7 strain) or Plasmodium yoelii (17XNL strain) CSP was expressed in E. coli and purified on a heparin sepharose affinity column. The concentration of purified rPf CSP was then determined via the Coomassie protein assay reagent (Thomas Scientific, 1856209, Swedesboro, NJ). The stock concentration of the batch of rPf CSP used in this study was 250 ng/μL.

2.2. Anti-P. falciparum CSP mouse monoclonal antibody (mAb) 2A10 The mAb 2A10 was utilized for both capture and detection of target antigen. This antibody specifically recognizes the repeat NANP sequences within Pf CSP. The hybridoma expressing 2A10 mAb was procured from MR4/ATCC, in Manassas, Virginia (Nardin et al., 1982). The antibody was generated as ascitic fluid in 12-week-old male BALB/c mice, and IgG was purified using protein G affinity chromatography (by a commercial source, Harlan Laboratories INC. Madison, WI USA). The protein concentration of the stock mAb 2A10 was 1.55 mg/mL. For the quantification of bound antigen via an avidin-HRP luminescent reporter system, a detection antibody was prepared by biotinylating an aliquot of approximately 150 μg of 2A10 mAb using the Lightning-Link™ Biotin Conjugation Kit (Innova Biosciences, 704–0010; Cambridge, UK) according to the manufacturer’s instructions.

2.3. Production and enumeration of Pf sporozoites (SPZ) All P. falciparum transmission in A. stephensi mosquitoes was performed at the Bloomberg School of Public Health at John Hopkins University (Baltimore, MD, USA) by Dr. Abhai Tripathi and Dr. Godfree Mlambo who are coauthors on this manuscript. Three- to five-day-old female A. stephensi mosquitoes were membrane fed with P. falciparum (NF54) gametocytes that were prepared by enrichment of asexual blood stage P. falciparum parasites cultured in human erythrocytes and serum (Miura et al., 2013). Gametocyte fed mosquitoes were maintained in cages kept at 26 °C and 80% relative humidity. On day 8 or 9 post-infected blood meal, eight random mosquitoes from each cage were removed, and their midguts were dissected in order to estimate the population’s average oocyst burden. The remaining mosquitoes in each cage were maintained until day 15 when the salivary glands of infected mosquitoes were dissected and free SPZ were collected as described elsewhere (Kumar et al., 2013; Mlambo et al., 2010; Pacheco et al., 1979). Pf SPZ count was determined in a hemocytometer with light-microscopy by two independent operators to eliminate any sources of reader bias. 2.4. Isolation and enumeration of Pf oocysts in whole mosquito lysates All Pf Oocyst burdens of mosquitoes fed P. falciparum infected blood meal were estimated and extrapolated from counts obtained via microscopic dissection of at least 20 mosquitoes randomly sampled from each independent batch receiving an infected blood meal. Specifically, dissected mosquito midguts were placed into wells filled with 0.05% mercurochrome solution (Sigma Aldrich, M7011, St. Louis, MO) on a cavity microscope slide and stained for 20–45 min. Then the midguts were transferred onto a regular microscope slide in a drop of mercurochrome and covered with a cover slip for counting. Developing Pf oocysts in mosquito midguts were analyzed in lysates obtained from whole mosquito specimens on days 8–10 post-feeding with infected blood meal. Briefly, 50 μL of lysis buffer (1% sarcosyl, 0.05% Tween-20 in 1× PBS) was added to each individual mosquito, and the specimens were homogenized using a piston which was washed with 100 μL of additional lysis buffer and the entire volume was thoroughly vortexed for 20 s. Lysate was boiled for 5 min before centrifugation for 10 min at 16,000 g to remove particulate matter. 2.5. Preparation of rPf CSP and Pf SPZ standard samples rPf CSP stock (250 ng/μL) was diluted in 1× lysis buffer to a final concentration of 1 ng/10 μL. Accurate protein quantitation was confirmed using a Bradford protein assay (Thermo Scientific, PI-23200; Rockford, IL). For the ECL-ELISA, an initial standard containing 50 pg/μL rPf CSP in blocking buffer (1% BSA, 0.5% Tween-20 in 1 × PBS) was prepared and then serially diluted until a final concentration of 0.02 pg/μL rPf CSP was achieved. Stock preparations containing 1250 Pf SPZ/μL were aliquoted into 1% sacrosyl in phosphate-buffered saline (PBS, pH 7.4) lysis buffer to a final concentration of 62.5 sporozoites/μL and boiled for 5 min. Samples were then further diluted in blocking buffer to prepare an initial standard containing 6.25 sporozoites/μL. Additional standards were obtained via serial two-fold dilutions until a final concentration of 0.6 sporozoites/μL was achieved. For analysis by ECL-ELISA, a 5 μL aliquot of the whole mosquito lysates (see Section 2.4) was added to 45 μL of blocking buffer and then the entire volume was added directly to the ELISA well. 2.6. ECL-ELISA The ECL-ELISA employs chemiluminescence to enhance sensitivity. Clear-bottomed black 96-well plates (Fisher Scientific, 07-200-567;

B. Grabias et al. / Journal of Microbiological Methods 108 (2015) 19–24

21

Fig. 1. Limit of Detection of rPf CSP. Serial dilutions containing 4–125 pg of recombinant Pf CSP were subjected to ECL-ELISA and the luminescence intensities plotted and fit to a nonlinear four parameter logistic equation (R2 = 0.97) (A). These same data points were also plotted on log-log axes and fit to a straight line (R2 = 0.99) to clearly demonstrate the effective linear range of the assay (B). Individual data points are plotted as mean ± SEM; n = 3.

Rockville, MD) were incubated with 0.2 μg of 2A10 antibody overnight at 4 °C. The following day, wells were washed 5× for 3 min each with washing buffer (0.1% Tween-20 in 1× PBS) at 37 °C for 1 h. Fifty microliters of standard samples (rPf CSP or Pf SPZ), test mosquito samples or no antigen controls (blocking buffer) were added to the wells immediately after the blocking step and allowed to incubate at 37 °C for 1 h and then subjected to 5 × washes with washing buffer. Biotinylated 2A10 mAb was subsequently diluted in blocking buffer at a 1:2500 dilution (~ 0.5 ng/μL) and added to the wells for 1 h at 37 °C. Excess antibody was removed via another set of 5 washes and a 1:4000 dilution of Streptavidin-HRP conjugate (Thermo Scientific, N100; Rockford, IL) was added to wells and incubated at room temperature for 1 h. The plate was then again washed 5× before the addition of 100 μL/well of SuperSignal ELISA Pico Chemiluminescent Substrate (Thermo Scientific, N100; Rockford, IL). After developing the plate for 5 min, luminescence was measured by using a Viktor III Luminometer (Perkin Elmer, 1420–012; Waltham, MA). We also directly compared the performance of the ECL-ELISA assay to the standard calorimetric ELISA format. To this end, side-by-side experiments were carried out for two ELISA protocols; for the calorimetric method, plates were developed utilizing 100 μL/well of ABTS Peroxidase Substrate solution (KPL, 50-64-02; Gaithersburg, MD) for 30 min. Absorbance readings were taken at 410 nm via Spectramax 340PC384 microplate reader (Molecular Devices; Sunnyvale, CA).

Fig. 2. Performance comparison of the ECL-ELISA and a standard calorimetric ELISA. In order to compare the sensitivity and dynamic range of the ECL-ELISA assay to the more traditional calorimetric format, recombinant CSP standards were simultaneously assessed via ECL-ELISA (left axis, 4–125 pg) and calorimetric ELISA (right axis, 62.5–2000 pg, linear range: 250–1000 pg). Individual data points are plotted as mean ± SEM; n = 3.

2.7. Statistical analysis 2.7.1. The four parameter logistic model Data acquired from the ECL-ELISA assay were fit to a nonlinear four parameter logistic (4PL) model of the form: Luminesence ¼

1þ



A−D

pg Antigen C

B þ D

where A is the lower asymptote of the curve, D is the upper asymptote, C represents the inflection point, and B is the Hill slope, a constant that governs the steepness of the curve. To account for unequal variance across multiple sample concentrations, the data were subjected to nonlinear regression weighted by a factor of 1/(y2). A best fit curve generated from this model was employed to determine the concentration of antigen in unknown test samples and to simultaneously assess the quality of the data points via assessment of the coefficient of determination (R2). The limit of detection for each experiment was determined by defining the cutoff luminescence threshold as the mean of luminescence intensities for negative control “buffer only” samples plus three times the standard deviation (SD) of those samples. This value was subsequently substituted into the 4PL best fit model equation to back calculate a specific concentration or target density. Linear or dynamic range was determined by optimizing the coefficient of determination (R2) for the sample data points above the limit of detection when plotted on log-log axes and fit to a log-log line.

Fig. 3. Detection of native CSP in Pf SPZ. Sporozoites extracted from Plasmodium-infected mosquitoes were serially diluted to contain from 1.25 to 40 sporozoites per well and subjected to the ECL-ELISA. The luminescence intensities were then plotted and fit to both a nonlinear four parameter logistic equation (R2 = 0.96) (A) and a straight line on log-log axes (R2 = 0.93) to clearly demonstrate the effective linear range of the assay, which encompassed most of the samples examined, from 5 to 40 sporozoites (B). Individual data points are plotted as mean ± SEM; n = 3.

22

B. Grabias et al. / Journal of Microbiological Methods 108 (2015) 19–24

Fig. 4. Sensitivity of ECL-ELISA against oocysts from infected mosquito lysates. Lysates from 6 infected mosquitoes with an average oocyst burden of 150 oocysts/mosquito were pooled before being serially diluted and subjected to the ECL-ELISA assay. Luminescence intensities were then plotted and fit to a 4 PL (A) or linear model (B). The dynamic linear range encompassed all of the samples examined including 0.25–16 oocysts/μL and the LOD is calculated to be 0.056 oocysts. Individual data points are plotted as mean ± SEM; n = 3.

2.7.2. Assay precision In order to assess reproducibility of the assay, the intra-assay variability and the inter-assay variability were calculated. These are two distinct measures that validate the ability of the assay to produce consistent test results. For the computation of the intra-assay variability, rPf CSP standard samples comprising multiple concentrations were each run four times in a single experiment. The mean and the SD of counts obtained from those four replicates were then employed to calculate the intra-assay CV, calculated as follows:

Intra‐assay CVð%Þ ¼

SD  100 Mean

The inter-assay variability was calculated by using the means of these four replicates from experiments conducted on three different days. Specifically, the mean and the SD of the luminescence counts obtained from the three assay means were used as defined below:

Inter‐assay CVð%Þ ¼

SD of Means  100 Grand Mean

2.7.3. Specificity of ECL-ELISA for Pf CSP The specificity of the ECL-ELISA was determined based on its ability to distinguish between the detection of rPf CSP and recombinant CSP based on the sequence from P. yoelii, a mouse malaria parasite. The results were evaluated by employing a Student’s t-test for each series compared to no antigen control wells.

3. Results and discussion Highly sensitive immunoassays to assess the Plasmodium infection rate in mosquitoes are urgently needed for epidemiological studies and to measure the efficacy of drugs, vaccines and vector control programs on parasite transmission in endemic areas. Recently, we have reported on an ECL-Western blot assay that had the linear detection range of 3–12 pg of rPf CSP and the range for Pf SPZ detection was between 0.0625 and 1 parasite (Kumar et al., 2013). To our knowledge, this is the most sensitive assay reported for the detection of any malarial antigen. However, Western blot is not amenable for adaptation to a high throughput format, and thus not applicable for use in large field studies. In this paper, we report on the development and analytical characterization of an ECL-ELISA for Pf CSP detection based on a 96-well plate format and results are obtained as luminescence counts.

3.1. Analytical sensitivity 3.1.1. rPf CSP To determine the analytical sensitivity of our devised ECL-ELISA, twofold serial dilutions of rPf CSP constituting 4 pg–125 pg (or concentrations of 0.6–75.8 pM) per well were analyzed in three independent experiments. Luminescence counts of each test sample were normalized by the subtraction of readings obtained from “no antigen” controls that contained only blocking buffer. The resulting standard curves derived from average intensity readings obtained from three independent experiments are shown in Fig. 1A and B. Overall, the data adhered very closely to the 4PL model (R2 = 0.97) (Fig. 1A). In particular, a clearly defined linear region was observed for all the tested samples examined, from 4 to 125 pg of CSP (R2 = 0.99) (Fig. 1B).

Fig. 5. Specificity of Pf ECL-ELISA. In order to verify the stringency of the assay for the detection of CSP epitopes specific to Plasmodium falciparum (Pf.), serially diluted standards (4–125 pg) prepared from both Pf. and Plasmodium yoelii (Py.) CSP protein were evaluated simultaneously. No statistically significant difference between the Py. samples and the no antigen controls were observed. Individual data points are plotted as mean ± SEM; n = 3.

B. Grabias et al. / Journal of Microbiological Methods 108 (2015) 19–24 Table 1 Intra-assay variability.

23

Table 3 Intra-assay variability.

Day 1

Luminescence counts

rPf CSP (pg)

Sample 1 Sample 2 Sample 3 Sample 4 Mean

125 62.50 31.25 15.63 7.81 3.91

166394.4 170510.4 166335.4 181003.4 171060.9 6910.4 64135.4 66129.4 63897.4 67644.4 65451.7 1771.6 29500.4 27191.4 29433.4 31712.4 29459.4 1845.9 14125.4 14554.4 13675.4 14809.4 14291.2 498.2 5821.4 4642.4 4190.4 4446.4 4775.2 721.6 1921.4 1907.4 1209.4 1698.4 1684.2 332.5

Standard deviation

Day 3

Luminescence counts

CV (%)

rPf CSP (pg)

Sample 1 Sample 2 Sample 3 Sample 4 Mean

4.0 2.7 6.3 3.5 15.1 19.7

125 62.50 31.25 15.63 7.81 3.91

134085.8 124747.8 132621.8 130322.8 130444.6 4101.5 63185.8 60044.8 56636.8 61627.8 60373.8 2802.0 26862.8 28572.8 26600.8 27331.8 27342.0 874.4 11897.8 13225.8 12840.8 12828.8 12698.3 564.6 5157.8 5788.8 4591.8 5121.8 5165.0 489.8 1663.8 1834.8 1825.8 1599.8 1731.0 117.6

Utilizing the 4PL model and cutoff values calculated from 3 standard deviations plus the mean of the negative control samples, the analytical sensitivity of the ECL-ELISA for rPf CSP was calculated to be 4.4 pg. We next compared the analytical sensitivity of the ECL-ELISA format to the traditional calorimetric ELISA platform that is typically employed to screen for Plasmodium-infected mosquito vectors. The experiment employing serially diluted standards of rPf CSP was repeated for two independent ELISA plates processed simultaneously, one with standards containing 4–125 pg of CSP and another with standards containing 62.5–2000 pg of CSP. The plate with the lower range of standards (4–125 pg of CSP) was analyzed using the ECL-ELISA, while the higher range of standards (62.5–2000 pg of CSP) was analyzed and quantified by the traditional ELISA using the oxidation of chromogenic 2,2′azino-di-(3-ethylbenzthiazoline-6-sulfonate) substrate solution (Fig. 2). The limit of detection of the calorimetric format ELISA was 166 pg (100.6 pM), while the sensitivity of the ECL-ELISA was observed to be approximately 47 × higher at 3.5 pg (2.1 pM). The increased detection efficiency of the ECL-ELISA was also complemented by a much broader dynamic range than the standard calorimetric format. 3.1.2. Pf SPZ CSP As previously discussed, many published reports have utilized ELISA as screening test for Plasmodium-infected mosquitoes by the detection of Pf CSP in sporozoites extracted from dissected mosquito salivary glands. In order to provide adequate comparison of our ECL-ELISA to these existing protocols and determine the detection limits of the ECLELISA for Pf SPZ, we tested the assay against a broad concentration range of native Pf CSP derived from the known counts of Pf SPZ harvested from the day 14 salivary glands of infected Anopheles mosquitoes. Stock preparations of Pf SPZ were lysed and then serially diluted to generate standard samples that ranged from 1.25 to 40 SPZ/well. The standard curve generated from three independent experiments is shown in Fig. 3A and B. The limit of detection of our assay for purified Pf SPZ was 5 sporozoites. Studies enumerating the sporozoites found in mosquitoes from endemic areas have identified median counts of 4,000–6,000 sporozoites in the salivary glands (Pringle, 1966) or typically 1,300 to 4,600 sporozoites per developing oocyst in the midgut of

Table 2 Intra-assay variability. Day 2

Luminescence counts

rPf CSP (pg)

Sample 1 Sample 2 Sample 3 Sample 4 Mean

125 62.50 31.25 15.63 7.81 3.91

132937.6 135516.6 137130.6 155302.6 140221.9 10201.1 58995.6 62055.6 51555.6 62720.6 58831.9 5114.9 32218.6 30011.6 27751.6 30437.6 30104.9 1837.1 18247.6 10881.6 11589.6 15714.6 14108.4 3486.6 7190.6 5441.6 6049.6 4835.6 5879.4 1004.9 1922.6 1710.6 2009.6 2314.6 1989.4 250.6

Standard deviation

CV (%) 7.3 8.7 6.1 24.7 17.1 12.6

Standard deviation

CV (%) 3.1 4.6 3.2 4.4 9.5 6.8

infected mosquitoes (Harris et al., 2012; Rosenberg and Rungsiwongse, 1991). Thus, the detection limits of our assay are well below the typical sporozoite burden found in wild caught mosquito specimens, and we would expect near 100% sensitivity when screening vector populations. 3.1.3. Pf CSP on day 8 oocysts in whole mosquito lysates We next wanted to evaluate the ability of the ECL-ELISA to detect oocysts derived from whole mosquito lysates. Six mosquito specimens from a population that had received a P. falciparum gametocyte infected blood-meal and possessing an estimated average Pf Oocyst burden of 150 per mosquito were individually homogenized and then pooled before being serially diluted and subjected to ECL-ELISA analysis. Pf Oocyst counts in the standard samples ranged from 0.25 to 16 oocysts per well. The luminescence intensities were then used to generate a standard curve (Fig. 4A, B) and an LOD of 0.056 oocysts per well was obtained by substituting a threshold value into the best fit 4PL equation. The cutoff intensity was derived from three standard deviations plus the mean luminescence of the no antigen control wells. Importantly, the difference in sensitivity of the assay against sporozoites (LOD: 5 SPZ) and oocysts (LOD: 0.056 oocysts) can be explained by the fact that each developing oocyst within the mosquito midgut contains several thousand sporozoites (Rosenberg and Rungsiwongse, 1991). Thus, even the most dilute oocyst sample applied to the ECL-ELISA still contains several hundred sporozoites. In endemic areas, oocyst burden in mosquitoes is known to be dependent on the intensity of transmission. For example, the oocyst burden of infected mosquitoes from highly endemic regions has been observed to be at least twice that of mosquitoes from low transmission areas (Boudin et al., 2004). Similarly, reports on oocyst intensity in vectors from endemic regions have identified median infections of 2–3 oocysts per mosquito (Rosenberg, 2008). Following our sample preparation protocol, our assay can readily detect approximately 1.7 oocysts per mosquito; thus, ECL-ELISA should be able to detect even the lowest oocyst burden that is known to occur during malaria transmission. 3.2. Analytical specificity We next assessed the specificity of ECL_ELISA by determining the nonspecific binding of CSP from other species of Plasmodium. To this end, multiple samples containing different concentrations of Plasmodium falciparum CSP were examined with a concentration range of recombinant CSP from P. yoelii (rPyCSP). Overall, there was no statistically

Table 4 Inter-assay variability. Grand mean

Standard deviation

CV (%)

147242.5 61552.4 28968.8 13699.3 5273.2 1801.5

21198.8 3463.7 1445.3 871.7 560.0 164.4

14.4 5.6 5.0 6.4 10.6 9.1

24

B. Grabias et al. / Journal of Microbiological Methods 108 (2015) 19–24

significant difference in the luminescence counts between the rPyCSP and the negative (no antigen) controls (Fig. 5). 3.3. Precision studies For any screening assay to be a valuable diagnostic tool, the reported results must be consistent and reproducible. In order to investigate the precision of the ECL-ELISA and determine the inherent inter- and intravariability, four replicates of rPf CSP samples for each concentration in the linear range of the assay were measured on three consecutive days (Tables 1–3). The means and standard deviations of each concentration on each individual day are reflective of the intra-assay variability. Across multiple sample concentrations, intra-assay CVs ranged from a minimum of 2.7% (Table 1) to a maximum of 24.7% (Table 2). These data were subsequently pooled into grand means, standard deviations, and CVs to calculate inter-assay variations which ranged from 5.0% to 14.4% (Table 4). Importantly, for both data analyses the CVs are within acceptable range for such immunoassays (below ~20%). 4. Conclusions In summary, we have developed a highly sensitive and reproducible ECL-ELISA from whole mosquito that can be used to detect P. falciparum infection in endemic areas; the assay should be effective, even in areas where the oocyst burden is known to be very low. We find that the ECL-ELISA is approximately 47 fold more sensitive than the standard calorimetric ELISA format and can be easily adapted for high throughput analyses. To our knowledge, this is the first report that directly assessed the sensitivity of plate ELISA in detecting Pf oocysts developing in the mosquito midgut. Because of these performance characteristics, the ECL-ELISA should find broad applicability in any potential vector control or prevalence study. Moreover, this assay should be applicable to the other infectious diseases and possibly as a point-of-care test for the diseases such as cancer and auto-immune disorders. Acknowledgments This research was funded through Intramural Research Grants provided by the FDA. References Atchade, P.S., Doderer-Lang, C., Chabi, N., 2013. Is a plasmodium lactate dehydrogenase (pLDH) enzyme-linked immunosorbent (ELISA)-based assay a valid tool for detecting risky malaria blood donations in Africa? Malar. J. 12, 279. Bashir, I.M., Otsyula, N., Awinda, G., Spring, M., Schneider, P., Waitumbi, J.N., 2013. Comparison of PfHRP-2/pLDH ELISA, qPCR and microscopy for the detection of plasmodium events and prediction of sick visits during a malaria vaccine study. PLoS One 8, e56828. Beier, J.C., Perkins, P.V., Wirtz, R.A., Whitmire, R.E., Mugambi, M., Hockmeyer, W.T., 1987. Field evaluation of an enzyme-linked immunosorbent assay (ELISA) for Plasmodium falciparum sporozoite detection in anopheline mosquitoes from Kenya. Am. J. Trop. Med. Hyg. 36, 459–468. Boudin, C., Van Der Kolk, M., Tchuinkam, T., Gouagna, C., Bonnet, S., Safeukui, I., Mulder, B., Meunier, J.Y., Verhave, J.P., 2004. Plasmodium falciparum transmission blocking immunity under conditions of low and high endemicity in Cameroon. Parasite Immunol. 26, 105–110.

Burkot, T.R., Zavala, F., Gwadz, R.W., Collins, F.H., Nussenzweig, R.S., Roberts, D.R., 1984. Identification of malaria-infected mosquitoes by a two-site enzyme-linked immunosorbent assay. Am. J. Trop. Med. Hyg. 33, 227–231. Cao, Y.T., Li, J.H., Wang, Y.T., Fu, Y.W., Xu, J., 2014. Serum ALDH1A1 is a tumor marker for the diagnosis of non-small cell lung cancer. Tumori 100, 214–218. Collins, K.M., Hochberg, L.P., Ryan, P.R., Collins, W.E., Wirtz, R.A., Ryan, J.R., 2004. Quantification of Plasmodium malariae infection in mosquito vectors. Ann. Trop. Med. Parasitol. 98, 469–472. De Arruda, M.E., Collins, K.M., Hochberg, L.P., Ryan, P.R., Wirtz, R.A., Ryan, J.R., 2004. Quantitative determination of sporozoites and circumsporozoite antigen in mosquitoes infected with Plasmodium falciparum or P. vivax. Ann. Trop. Med. Parasitol. 98, 121–127. Fang, W., Vega-Rodriguez, J., Ghosh, A.K., Jacobs-Lorena, M., Kang, A., St Leger, R.J., 2011. Development of transgenic fungi that kill human malaria parasites in mosquitoes. Science 331, 1074–1077. Fontenille, D., Meunier, J.Y., Nkondjio, C.A., Tchuinkam, T., 2001. Use of circumsporozoite protein enzyme-linked immunosorbent assay compared with microscopic examination of salivary glands for calculation of malaria infectivity rates in mosquitoes (Diptera: Culicidae) from Cameroon. J. Med. Entomol. 38, 451–454. Harris, C., Morlais, I., Churcher, T.S., Awono-Ambene, P., Gouagna, L.C., Dabire, R.K., Fontenille, D., Cohuet, A., 2012. Plasmodium falciparum produce lower infection intensities in local versus foreign Anopheles gambiae populations. PLoS One 7, e30849. Kumar, et al., 2013. A chemiluminescent-Western blot assay for quantitative detection of Plasmodium falciparum circumsporozoite protein. J. Immunol. Methods 390, 99–105. Li, Z.N., Lin, S.C., Carney, P.J., Li, J., Liu, F., Lu, X., Liu, M., Stevens, J., Levine, M., Katz, J.M., Hancock, K., 2014. IgM, IgG, and IgA antibody responses to influenza A(H1N1) pdm09 hemagglutinin in infected persons during the first wave of the 2009 pandemic in the United States. Clin. Vaccine Immunol. 21, 1054–1060. Martin, S.K., Rajasekariah, G.H., Awinda, G., Waitumbi, J., Kifude, C., 2009. Unified parasite lactate dehydrogenase and histidine-rich protein ELISA for quantification of Plasmodium falciparum. Am. J. Trop. Med. Hyg. 80, 516–522. Miura, K., Deng, B., Tullo, G., Diouf, A., Moretz, S.E., Locke, E., Morin, M., Fay, M.P., Long, C.A., 2013. Qualification of standard membrane-feeding assay with Plasmodium falciparum malaria and potential improvements for future assays. PLoS One 8, e57909. Mlambo, G., Kumar, N., Yoshida, S., 2010. Functional immunogenicity of baculovirus expressing Pfs25, a human malaria transmission-blocking vaccine candidate antigen. Vaccine 28, 7025–7029. Nardin, E.H., Nussenzweig, V., Nussenzweig, R.S., Collins, W.E., Harinasuta, K.T., Tapchaisri, P., Chomcharn, Y., 1982. Circumsporozoite proteins of human malaria parasites Plasmodium falciparum and Plasmodium vivax. J. Exp. Med. 156, 20–30. Noedl, H., Yingyuen, K., Laoboonchai, A., Fukuda, M., Sirichaisinthop, J., Miller, R.S., 2006. Sensitivity and specificity of an antigen detection ELISA for malaria diagnosis. Am. J. Trop. Med. Hyg. 75, 1205–1208. Pacheco, N.D., Strome, C.P., Mitchell, F., Bawden, M.P., Beaudoin, R.L., 1979. Rapid, largescale isolation of Plasmodium berghei sporozoites from infected mosquitoes. J. Parasitol. 65, 414–417. Parvez, S.D., Al-Wahaibi, S.S., 2003. Comparison of three larviciding options for malaria vector control. East. Mediterr. Health J. 9, 627–636. Pereira, M.C., Cardoso, P.R., Da Rocha Jr., L.F., Rego, M.J., Goncalves, S.M., Santos, F.A., Galdino-Pitta, M.R., Dantas, A.T., Duarte, A.L., Pitta, M.G., 2014. Simvastatin inhibits cytokines in a dose response in patients with rheumatoid arthritis. Inflamm. Res. 63, 309–315. Pringle, G., 1966. A quantitative study of naturally-acquired malaria infections in Anopheles gambiae and Anopheles funestus in a highly malarious area of East Africa. Trans. R. Soc. Trop. Med. Hyg. 60, 626–632. Quan, Y., Chen, M., Zhan, Y., Zhang, G., 2011. Development of an enhanced chemiluminescence ELISA for the rapid detection of acrylamide in food products. J. Agric. Food Chem. 59, 6895–6899. Rajakumar, G., Abdul Rahuman, A., 2011. Larvicidal activity of synthesized silver nanoparticles using Eclipta prostrata leaf extract against filariasis and malaria vectors. Acta Trop. 118, 196–203. Rosenberg, R., 2008. Malaria: some considerations regarding parasite productivity. Trends Parasitol. 24, 487–491. Rosenberg, R., Rungsiwongse, J., 1991. The number of sporozoites produced by individual malaria oocysts. Am. J. Trop. Med. Hyg. 45, 574–577. Wang, S., Jacobs-Lorena, M., 2013. Genetic approaches to interfere with malaria transmission by vector mosquitoes. Trends Biotechnol. 31, 185–193. WHO, 2012. World Malaria Report. World Health Organization.