Inhibitors of malaria parasite cyclic nucleotide phosphodiesterases block asexual blood-stage development and mosquito transmission

Abstract

Cyclic nucleotide-dependent phosphodiesterases (PDEs) play essential roles in regulating the malaria parasite life cycle, suggesting that they may be promising antimalarial drug targets. PDE inhibitors are used safely to treat a range of noninfectious human disorders. Here, we report three subseries of fast-acting and potent Plasmodium falciparum PDEβ inhibitors that block asexual blood-stage parasite development and that are also active against human clinical isolates. Two of the inhibitor subseries also have potent transmission-blocking activity by targeting PDEs expressed during sexual parasite development. In vitro drug selection experiments generated parasites with moderately reduced susceptibility to the inhibitors. Whole-genome sequencing of these parasites detected no mutations in PDEβ but rather mutations in downstream effectors: either the catalytic or regulatory subunits of cyclic adenosine monophosphate-dependent protein kinase (PKA) or in the 3-phosphoinositide-dependent protein kinase that is required for PKA activation. Several properties of these P. falciparum PDE inhibitor series make them attractive for further progression through the antimalarial drug discovery pipeline.

Fig. 1.. Structures and activity of the PDE inhibitor series against ABS parasites in biochemical and cell-based assays.

(A) Structure of an example of the 2-alkyl, 5-aryl, and 5-benzyl PDE inhibitor subseries. (B) Plot showing the correlation between IC₅₀ values for inhibition of cAMP hydrolysis by PDEβ in P. falciparum schizont lysates and EC₅₀ values for inhibition of P. falciparum ABS growth for all three subseries. Green, 5-benzyl; red, 5-aryl; blue, 2-alkyl. Circles correspond to the main compounds in this study, and squares indicate compounds tested but not further described herein. Pearson’s correlation coefficient, r = 0.841. (C) Bar charts showing the fold change in elevated cAMP levels in rings (n = 9) and elevated cGMP levels in schizonts (n = 6) [measured by enzyme-linked immunosorbent assay (ELISA)] following treatment with a 5-benzyl compound (SAL-0010042) compared to dimethyl sulfoxide (DMSO) treatment. Error bars show the means ± SD with individual replicate measurements shown as black dots. (D) Bar chart showing the effects of a 5-benzyl compound (SAL-0010042) on schizont formation (assessed by measuring DNA content by flow cytometry) when added to ring-stage parasites at four different time windows after erythrocyte invasion. (E) Plot showing the P. falciparum ABS killing profiles for representatives of the three subseries generated using parasite reduction ratio (PRR) assays. The reduction in the numbers of viable parasites measured is shown for each test compound (colored lines), over time compared to control antimalarials (colored hatched lines). All compounds were used at a concentration of 10× EC₅₀.

Fig. 2.. Activity of the inhibitors against sexual-stage parasites.

(A) Bar chart showing the fold change in elevated cGMP levels (measured by ELISA) in stage IV to V gametocytes following treatment with representatives of the three subseries or with XA compared to DMSO treatment. Experiments were carried out three times in duplicate. Error bars show the means ± SD. Compounds subseries are indicated by the following colors: green, 5-benzyl; red, 5-aryl; blue, 2-alkyl. (B) Bar charts showing the ability of representatives of each subseries to stimulate gametogenesis following three different treatment durations. The proportions of the treated cultures that were stage IV gametocytes, stage V gametocytes, and male and female gametes, indicated by the color of the bars, were identified by examination of parasite morphology in Giemsa-stained blood films. *P < 0.05 and **P < 0.01 using an unpaired Student’s t test. (C) Images of Giemsa-stained blood films containing stage V gametocytes or “rounded-up” gametocytes/gametes following treatment with DMSO, the 5-benzyl compound SAL-0010042, or XA. Scale bar, 7 μm. (D) Images of Giemsa-stained blood films containing exflagellating male gametocytes/gametes following treatment with the 5-benzyl compound SAL-0010042. Scale bar, 7 μm. (E and F) Plots showing infection prevalence and intensity, respectively, for test and control compounds. Experiments were carried out with stage V gametocytes of the P. falciparum NF54 HGL reporter strain following incubation with the 5-benzyl compound SAL-0010003 for 48 hours before feeding to A. stephensi mosquitoes. Eight days after feeding, infection status was assessed by luminescence analysis of individual mosquitoes. All conditions were tested in duplicate feeders (respective intensities distinguished by blue or black dots) with 24 mosquitoes analyzed per feeder. Plots showing the results of SMFA data for SAL-0010003 compared to dihydroartemisinin (DHA) and DMSO controls. IC₅₀ values for both the prevalence and intensity of infection are indicated. RLU, relative light units.

Fig. 3.. Activity of the inhibitor series against clinical isolates and selectivity against human PDE isoforms.

(A) SAL-0010255 (2-alkyl) was tested for ex vivo activity against eight P. falciparum clinical isolates from Porto Velho, Rondônia in the Brazilian Amazon (black dots) and against the P. falciparum 3D7 control line (green dots). Chloroquine and artesunate were also tested in parallel. Red bars indicate the geometric mean EC₅₀ values, and the error bars depict the 95% confidence interval for the clinical isolates. The values of the geometric means are shown for each compound. (B) SAL-0010042 (5-benzyl) and SAL-0010333 (5-aryl) and two control antimalarials were tested for activity against 47 clinical isolates from the Tororo and Busia Districts of Uganda. Black dots represent individual isolate EC₅₀ values of the clinical isolates; enlarged blue dots indicate mean Dd2 control EC₅₀ values; enlarged green dots indicate mean 3D7 control EC₅₀ values. The red bars indicate the geometric mean, and the error bars depict the 95% confidence interval. The values of the geometric means are shown for each compound. (C) In light blue (top), a high-resolution crystal structure of the catalytic site of human PDE5 (66) (code 1TBF) is shown bound to sildenafil. des-Me sildenafil is visualized assuming the same coordinates as for sildenafil in 1TBF. The atoms of sildenafil and the side chain of Ala⁷⁶⁷ are shown with ball and stick. Other atoms are depicted as stick or cartoon. In beige (middle), the sequence of PDEβ has been modeled onto 1TBF using Phyre2 (67). The structures of sildenafil and des-Me sildenafil are shown in the bottom panels.

Fig. 4.. In vitro selection of parasites with mutations in three effectors of cAMP signaling that operate downstream of PDEβ.

Schematic showing the key players in the P. falciparum cAMP signaling pathway with red indicating pathway components that had mutations selected under drug pressure.