Fig 1.
The procedure for automating the analysis of mosquito larval swimming.
In each trial, 30 images are acquired at 10 ms intervals and stored for later offline analysis. An index of the amount of movement is obtained by measuring the variance for each pixel over time. Pixels for which the variance more than 3 standard deviations from the mean variance are scored as 1, the remainder as 0. The movement index for each well is taken as the sum of these scores for that well. The output of the algorithm for quantifying movement plotted against the concentration of temephos (bottom right), a larvicide commonly used in the control of mosquitoes, is shown at the end of the pipeline. Alongside this similar data for larvae of An. gambiae are shown. A concentration-dependent inhibition of movement is seen in studies on larvae of An. gambiae and Ae. aegypti.
Fig 2.
Screening a chemical library on An. gambiae larvae using the INVAPP / Paragon system.
Concentration response curves for the five compounds in the Medicines for Malaria Venture (MMV) Pathogen Box library identified as active in the primary screen. The movement index for each well was measured before adding the compounds, and then again 240 min later. The second reading is divided by the first to normalize for background variations between the wells caused, for example, by differences in the number of larvae dispensed in each well. The compounds were applied over the range 3.0 × 10−9 M—1.0 × 10−4 M, although only concentrations above 7.8 × 10−7 M are shown for clarity. Larval progeny from both resistant An. coluzzii (Tiassale) and sensitive An. gambiae (G3) strains were studied and labelled “G” and “T” in the figure respectively. Of the compounds tested, only MMV688934 (tolfenpyrad) and MMV637229 (Clemastine) have chemical names.
Fig 3.
Motility measurement following deltamethrin exposure in concentration- and time-dependent manner for An. Gambiae susceptible (G3) and resistant (Tiassale) strains as well as for Ae. Aegypti susceptible (New Orleans) and resistant (Cayman) strains.
The normalized movement index, which is the movement index (see Methods) divided by the index recorded before the application of insecticide, is plotted against the concentration after 60 min applications (A) and the time following exposure (B) to 10−6 M deltamethrin. Increased rate of movement inhibition is apparent in the larvae from adult susceptible strains (blue: G3 (G), New Orleans (N)) as compared to the adult resistant strains (red: Tiassale (T), Cayman (C)). The dotted line indicates the value before insecticide application. Error bars indicate ± 1 s.e.m.
Fig 4.
The application of the standard WHO larval assay to detect deltamethrin resistance in Ae. aegypti.
Deltamethrin was tested over a range between 5.0 x 10−10 M to 1.0 x 10−6 M. (A) The concentration-response curves are given for the wild-type New Orleans strain (blue) and the Deltamethrin-resistant Cayman strain (red). (B) The fitted PIC50s estimated from fitting the curves in A to a Sigmoid curve. The WHO assay yielded LC50s of 8.60 ± 0.04 (2.5 nM) and 7.29 ± 0.14 (51.3 nM) for New Orleans and Cayman respectively (2-tailed, independent t-test, T(4) = -3.9, P = 0.0). Each point in (B) represents the mean of 3 (WHO) separate experiments and error bars indicate the standard error of the mean.
Fig 5.
The output of a smartphone application that measures mosquito larvae motility following deltamethrin treatment, quantified by the average pixel variance over 2 s.
(A) and (B) show the pixel variance over time (motility) of Aedes aegypti New Orleans and Cayman larvae with 10−5 M deltamethrin in water and DMSO/water only controls. The spike at t 200s is the addition of the deltamethrin or control (water). The solid lines show the mean and the shaded area, the standard deviation (n = 3). The susceptible strain (A) motility decreases over time with DM added, whilst the resistant strain (B) does not. (C) and (D) show the pixel variance over time of Anopheles gambiae G3 and An. coluzzii Tiassale larvae with 10−5 M deltamethrin in water and DMSO/water controls. In this case, the resistant strain variance also decreases, but at a slower rate. (E) Fitting the variance over time of all the DM samples to exponential decay functions allows the extraction of the decay rate in each case (see methods). These are plotted on the bar chart with error bars showing the standard deviation of the fitted decay rate (n = 3). A higher decay rate is observed for the larvae from susceptible strains compared to their respective resistant strains. Aedes aegypti p = 1.9 × 10−5 (one-tailed t-test, T(2) = 19.8); Anopheles p = 1.1 × 10−2 (one-tailed t-test, T(2) = 6.6).