Novel High-Throughput Fluorescence-Based Assay for the Identification of Nematocidal Compounds That Target the Blood-Feeding Pathway

Abstract

Hookworm infections cause a neglected tropical disease (NTD) affecting ~740 million people worldwide, principally those living in disadvantaged communities. Infections can cause high morbidity due to their impact on nutrient uptake and their need to feed on host blood, resulting in a loss of iron and protein, which can lead to severe anaemia and impaired cognitive development in children. Currently, only one drug, albendazole is efficient to treat hookworm infection and the scientific community fears the rise of resistant strains. As part of on-going efforts to control hookworm infections and its associated morbidities, new drugs are urgently needed. We focused on targeting the blood-feeding pathway, which is essential to the parasite survival and reproduction, using the laboratory hookworm model Nippostrongylus brasiliensis (a nematode of rodents with a similar life cycle to hookworms). We established an in vitro-drug screening assay based on a fluorescent-based measurement of parasite viability during blood-feeding to identify novel therapeutic targets. A first screen of a library of 2654 natural compounds identified four that caused decreased worm viability in a blood-feeding-dependent manner. This new screening assay has significant potential to accelerate the discovery of new drugs against hookworms.

Keywords: blood-feeding; drug-screening; fluorescence; helminth; hookworm; motility; viability.

Figure 1

(a) Live, quinidine-treated (100 µM, 4 days), or boiled (dead) larvae were exposed to Sytox Green for 24 h prior to microscopic analysis using brightfield or fluorescence emissions. The results were derived from three independent experiments using 100 N. brasiliensis iL3s per well. Fluorescence images were analysed using a look-up-table (LUT) “fire” in Fiji (purple low intensity, yellow high intensity). (b) Spectrophotometric measurement of Sytox Green. The fluorescence intensity was log10-transformed. Data were pooled from three independent experiments and analysed using the Kruskal-Wallis test. (c) Linear regression analysis of the number of dead iL3 and Sytox Green fluorescence, measured 24 h after labelling. Data were pooled from three independent experiments and expressed as mean ± standard error of the mean (SEM) ANOVA were performed and post-hoc significance is indicated, ** p < 0.01, *** p < 0.001.

Figure 2

(a) N. brasiliensis iL3s (n = 100) were co-cultured in the presence of RBCs with 100 µM known anthelmintics (pyrantel, piperazine, imidazole, albendazole, chloroquine, benzimidazole, metronidazole and quinidine). After 3 days, Sytox Green was added to the wells and 24 h later, fluorescence (arbitrary units, a.u.) was measured by spectrophotometry. The average of duplicate values for each drug are shown as a heat-map ranging from medium viability (MED) to low viability (DEAD i.e., equivalent to killed by boiling) (arbitrary units a.u.). (b) Drug dilution series for pyrantel, albendazole (Alb.), quinidine (QND), and quinine (QN) using the same assay as described in (a). Data were normalised to those for “dead” controls and pooled from three independent experiments and represented as mean ± standard error of the mean (SEM). The compound concentrations were log10-transformed and fitted using a variable slope four-parameter equation, using ordinary least squares fit model. (c) Using the same conditions as described in (b), motility was assessed by microscopy 4 days after culture. Motility scores were calculated for individual compounds, normalised with reference to the negative control (100% motility) and recorded as percentages. Data points represent one experiment conducted in triplicate; mean ± standard error of the mean (SEM).

Figure 3

Results of the primary screen of a natural compound library (n = 2654) against third-stage larvae of Ni. brasiliensis with reference to quinidine as positive control (QND; 32 replicates per plate) and to DMSO 1% as negative control (DMSO, 32 replicates per plate). All compounds were screened in duplicate using different batches of the parasite (top or bottom panel). All values were normalised for each plate to mean (DMSO) + 3 ∗ SD(DMSO) = 10,000 units (Dashed black line). All compounds and positive controls were tested at 100 µM. Each grey or red dot represents an individual test compound. Mean +/− standard deviation (dashed coloured line +/− coloured zone) for controls were calculated as an average of all replicates (n = 224) and are indicated in green for QND and purple for DMSO. Red dots are hits identified (as described in Figure S1).

Figure 4

Comparative heat-map of hits displaying scaled readouts for viability, motility, and pigmentation is shown (two duplicates for each drug and two different batches of larvae, “1” and “2” were used). Viability: 40 iL3 per well were incubated with compounds of interest at 100 µM for 4 days. After 72 h, Sytox Green was added to the culture and viability was acquired by spectrophotometric measurement 24 h later. Motility: compounds with an effect that was greater than that obtained for the reference compound (QND) were further scored for motility under bright-field microscopy. Pigmentation: Compounds with an effect that was greater than that obtained for the reference compound (QND) were also scored under bright-field microscopy for the presence/degradation of a hemozoin-like compound. All measurements were performed in the same experiment. Values were Z-scored for each readout (centred of the mean of the data), negative Z-scores (lower fitness) are shown in dark blue, and positive Z-scores (higher fitness) are shown in yellow. Clusters are shown on the left and numbered 1–8.

Figure 5

Results of the secondary screen of a natural compound library (n = 66) against third-stage larvae N. brasiliensis with reference to quinidine as positive control (QND, 32 replicates per plate) and to DMSO 1% as negative control (DMSO, 32 replicates per plate). All test and positive control compounds were tested at 100 µM in the presence or absence of blood in duplicates. Each grey or red dot represents an individual test compound. Mean +/− SD for controls were calculated as an average of all replicates (n = 224) and are indicated in purple for DMSO. All values were normalised for each plate to mean (DMSO) + 3 ∗ SD(DMSO) = 10,000 units (Dashed black line). Red dots are compounds identified as hits in presence of blood (39 compounds, screen 2, as described in Figure S1). (a) Normalised fluorescence intensity obtained for each of the 66 hits in absence of blood. (b) For each of the 66 compounds, the fold change of Sytox green intensity between screen 2 (in presence of blood) and screen 3 (in absence of blood) is represented as a function of the Sytox intensity as compared to the reference drug quinidine in presence of blood. The dash grey line has been set manually to separate hits with the higher fold change between absence and presence of blood.