Gametocyte-specific and all-blood-stage transmission-blocking chemotypes discovered from high throughput screening on Plasmodium falciparum gametocytes

Abstract

Blocking Plasmodium falciparum human-to-mosquito transmission is essential for malaria elimination, nonetheless drugs killing the pathogenic asexual stages are generally inactive on the parasite transmissible stages, the gametocytes. Due to technical and biological limitations in high throughput screening of non-proliferative stages, the search for gametocyte-killing molecules so far tested one tenth the number of compounds screened on asexual stages. Here we overcome these limitations and rapidly screened around 120,000 compounds, using not purified, bioluminescent mature gametocytes. Orthogonal gametocyte assays, selectivity assays on human cells and asexual parasites, followed by compound clustering, brought to the identification of 84 hits, half of which are gametocyte selective and half with comparable activity against sexual and asexual parasites. We validated seven chemotypes, three of which are, to the best of our knowledge, novel. These molecules are able to inhibit male gametocyte exflagellation and block parasite transmission through the Anopheles mosquito vector in a standard membrane feeding assay. This work shows that interrogating a wide and diverse chemical space, with a streamlined gametocyte HTS and hit validation funnel, holds promise for the identification of dual stage and gametocyte-selective compounds to be developed into new generation of transmission blocking drugs for malaria elimination.

Fig. 1. HTS assay optimization.

a Representative image of the gametocyte stages used in the HTS optimization: Giemsa-stained blood smears of infected red blood cell (RBC) cultivation after three-day treatment with 50 mM NAG followed by additional 5 days of cultivation. Early-stage V gametocytes are visible and a 2% gametocytaemia can be calculated. Black scale bar = 10 µm b Luminescence counts, expressed as Relative Light Units (RLU, averages and standard deviations are depicted, n = 4 independent experiments or n = 2 independent experiments for MB), of haematocrit dilutions of uninfected and of vehicle or 10 µM Methylene Blue (MB) treated infected RBC at 48- and 72 h incubation. Cultures were treated with NAG as above described. c Luminescence reporter assay metrics of MB treated vs. untreated culture at 48 h (averages and standard deviations are depicted, n = 48 technical replicates). Coefficient of variation percentage (CV%) are reported within each bar. Assay with a precision of CV% < 10 are considered optimal. For both a and b due to the high difference in magnitude between high and low conditions, a statistical analysis to assess significance was not needed.

Fig. 2. HTS Z’ value and inhibition frequency.

a Diagram of number of 384 wells plates grouped by the same Z’ value. In total, the Z’ value was calculated on 378 plates. b Inhibition percentage frequency distribution of the 119,059 compounds tested. The dotted line represents the 41% inhibition cut off limit.

Fig. 3. Hit confirmation and selectivity analysis.

The number of replicates for each experimental point was one for all the assays due to the technical challenges related to gametocyte assays. a Distribution chart representing the percentage of inhibition of 960 compounds in HeLa cells proliferation assay at 10 µM and in 3D7elo1-pfs16-CBG99 gametocyte assay at 20, 4, and 0.8 µM. The light blue and violet areas represent the selected and the discharged compounds, respectively. Dot horizontal line inside the areas indicates the median inhibition percentage (thick line) and top and bottom quartiles (thin lines). b Scatter plot of the percentage inhibition of PfLuc assays of 638 compounds on NF54 pfs16-GFP-PyLUC strain vs NF54 hsp86-PpyRE13 strain. Cross red lines, settled at 40% inhibition for each axis, divides the plot in four fields. c Bubble scatter plot (log–log) of the selected 84 compounds potency (nM) on asexual stages (pLDH assay) vs gametocytes (PfLuc assay) in the NF54 hsp86-PpyRE13 strain. The bubble colours correspond to the degree of potency of the HeLa proliferation assay coded by the column on the right.

Fig. 4. Male gamete exflagellation assay.

Open circles represent single experimental points. Averages and standard deviations are depicted for each condition (n = 2 independent experiments). Average DMSO exflagellation events per microscopy field: 17.6. Methylene Blue (MB) and compound DDD01035881 (84) were used as positive controls.

Fig. 5. P. falciparum transmission blocking activity to Anopheles stephensi mosquitoes of selected compounds by Standard Membrane Feeding Assays (SMFA).

a Structure of the 9 compounds selected for SMFA. b The graph shows the individual mosquito luminescence values used to determine oocyst intensity for DMSO (N), uninfected mosquitoes (P), positive control MMV0048 and the tested compounds at both concentrations. Averages and standard deviations are depicted for each conditions (n > 14 biological replicates).

Fig. 6. Screening funnel.

The CNCCS collection of approximately 120,000 was screened with gametocytes of the 3D7elo1-pfs16-CBG99 strain with a PfLuc assay at 10 µM concentration. To select the active compounds, a cut-off limit of 41% inhibition was applied resulting in 960 compounds (hit rate of about 0.8%). Three counter screening assays were performed on these compounds at fixed doses, a HeLa cell proliferation assay and two PfLuc assays on gametocytes of the NF54 hsp86-PpyRE13 and of the NF54 pfs16-GFP-PyLUC strains. An additional quality control assay and a chemical clustering analysis restricted the number of compounds to 84. These were filtered based upon potency as determined by full dose response assays complemented with a pLDH asexual stage assay on the NF54 hsp86-PpyRE13 strain. Results led to select 30 compounds to be tested in a male gamete exflagellation assay, whose results led to the final selection of 9 compounds tested in the standard membrane-feeding assay (SMFA), of which 8 turned out to be able to possess transmission blocking activity. Three of the latter represent, to our knowledge, novel chemotypes of P. falciparum transmission blocking molecules.