Fig 1.
Summary of the modifications to the slot blot assay protocol.
A number of changes were implemented to improve overall assay performance, reduce assay time, and minimize reagent and film requirements to better allow broad usage of the ECL-SB in resource poor field sites. The time required to perform the slot blot has been reduced by more than half of the original protocol length and the sensitivity for Pfoocysts has been improved from to 0.25 oocyst to 0.02 oocyst.
Fig 2.
Optimization of sample processing and antibody incubations.
Before being applied to the ECL-SB membrane, protein samples were typically boiled in lysis buffer for ten minutes. In order to potentially shorten assay time, we compared band intensities when samples containing 2 oocysts per well were subjected to either the standard boiling protocol (left), homogenized directly in hot (100°C) lysis buffer (middle), or lysed and applied to the slot blot all at room temperature (right) (A). Our data indicate that the RT lysis method is at least as good and statistically not significantly different from boiling samples for ten minutes thus boiling was removed from the protocol. We also evaluated different secondary antibody and detergent concentrations during the secondary antibody incubation step and the overall effects on slot blot band intensities and estimated detection limits of recombinant PfCSP are shown (B). Integrated optical densities were obtained for each specimen tested under each antibody and detergent concentration and fit to a Michaelis-Menten regression curve. Utilizing lysates of uninfected mosquito specimens as negative controls, we established a cutoff threshold of mean IOD + 2 standard deviations to estimate assay sensitivity under each condition being evaluated. We observed that highest sensitivity and signal-to-noise ratio was obtained when a secondary antibody dilution of 1:20,000 was employed with 0.1% Tween-20.
Fig 3.
Labeling of primary antibody enhances detection of PfOocysts in the slot blot assay.
We evaluated whether conjugation of primary mAb 2A10 with alkaline phosphatase (1°-AP) enhanced sensitivity for the detection of Pfoocyst prepared from mosquito lysates when compared to the use of AP-conjugated secondary antibody (2°-AP). Detection limits were compared for each protocol by fitting the band intensities of serially diluted oocysts to a Michaelis-Menten regression curve and establishing a cutoff intensity threshold of mean + 2 SD from unfed mosquito specimens run on the same blot. Labeled primary antibody displayed overall higher band intensities across the range of oocyst dilutions examined and achieved lower limits of detection than the typical sandwich antibody format (0.009 oocyst versus 0.02 oocyst, respectively). The removal of an additional antibody incubation step also contributed to an overall shorter assay time in the newly developed slot blot protocol.
Fig 4.
Sensitivity comparison between 2A10 and C3103 mAbs.
Twofold serial dilutions of rPfCSP protein were transferred to a nitrocellulose membrane using the slot blot apparatus and incubated with AP-conjugated 2A10 or C3103 mAb. Chemiluminescent signal was captured using autoradiography film and band intensities were quantitated using ImageJ software. While both antibodies adhere very well to the Michaelis-Menten curve fit, the C3103 antibody developed in our laboratory exhibited approximately 2-fold enhanced detection of CSP antigen.
Fig 5.
Comparison of film and digital scan data capture methods in the detection of rPfCSP.
Twofold serial dilutions of rPfCSP protein were transferred to a nitrocellulose membrane using the slot blot apparatus and incubated with AP-conjugated C3103 mAb. Chemiluminescent signal from the membrane was captured using both autoradiography film and the C-Digit scanner. The resulting band intensities were fit to Michaelis-Menten regression curves to identify changes in the detection limit of the slot blot assay. Cutoff threshold intensities were determined from the mean intensity of negative control samples plus two standard deviations Film and the digital scan displayed similar sensitivities of approximately 0.01 pg of CSP. Intensities obtained on the C-digit scanner were generally bolder at lower intensities but declined more dramatically, with a steeper slope near the detection limit.
Fig 6.
Comparison of film and digital scan data capture methods in the detection of Pf oocyst.
Stock oocyst lysate generated from pooled infected mosquitoes was serially diluted and processed according to the ECL-SB assay protocol. Chemiluminescent signal from the membrane was obtained from both exposure to film and a digital scan. The resulting band intensities were plotted and detection limits were determined. Data were plotted on logarithmic scales to better visualize overall trends as linear scale axes results in four of the lowest concentration data points acquired on the LICOR residing closer to the horizontal Y = 0 axis than the higher concentration samples. Film displayed slightly better sensitivity for Pfoocysts, detecting as few as 0.01 oocyst whereas the C-Digit scanner could only detect as little as 0.02 oocyst. Both methods could demonstrate reactivity for infected mosquitoes with a single developing oocyst.
Fig 7.
Inter-/Intra-assay varibility for the C-Digit blot scanner.
In three separate experiments performed on three different days, 0.25 oocyst was loaded into each slot blot well and data was captured with the C-digit blot scanner. Variability in band intensities from each day never exceeded 18% and overall inter-assay variability was estimated to be 9.85%.
Table 1.
rPfCSP detection—Inter-/Intra-assay variability of 10 samples (S1-S10) over three days.
Fig 8.
Prevalence estimation in high infectivity mosquito population using the C-Digit blot scanner.
A batch of mosquitoes were fed an infectious blood meal with high percent Pf gametocytemia. On day 8 post feeding, a random sampling of 15 mosquitoes was dissected, stained, and analyzed using microscopy. Oocyst prevalence and intensity were estimated to be 86.7% and 20.5 oocysts per mosquito, respectively. This estimate was compared to the prevalence calculated from 20 different mosquitoes of the same batch that were homogenized and analyzed via slot blot data acquired using both film and the C-Digit blot scanner for data capture. Data obtained from both the film and blot scanner agreed very well, with 18/20 mosquitoes testing positive (90% prevalence) in both methods. The cutoff threshold for each analysis was determined by calculating the mean + 2SD of the band intensities of the negative or unfed mosquitoes.
Fig 9.
Prevalence estimation in low infectivity mosquito population using the C-Digit blot scanner.
A batch of mosquitoes were fed an infectious blood meal with high percent Pf gametocytemia. On day 8 post feeding, a random sampling of 15 mosquitoes were dissected, stained, and analyzed using microscopy. Oocyst prevalence and intensity were estimated to be 20% and 1.4 oocyst per mosquito, respectively. This estimate was compared to the prevalence calculated from mosquitoes of the same batch that were homogenized and analyzed via slot blot data acquired using both film and the C-Digit blot scanner for data capture. Data obtained from both the film and blot scanner agreed very well, with 3/20 mosquitoes testing positive (15% prevalence) in both methods.