Plasmodium falciparum sexual parasites regulate infected erythrocyte permeability

Abstract

To ensure the transport of nutrients necessary for their survival, Plasmodium falciparum parasites increase erythrocyte permeability to diverse solutes. These new permeation pathways (NPPs) have been extensively characterized in the pathogenic asexual parasite stages, however the existence of NPPs has never been investigated in gametocytes, the sexual stages responsible for transmission to mosquitoes. Here, we show that NPPs are still active in erythrocytes infected with immature gametocytes and that this activity declines along gametocyte maturation. Our results indicate that NPPs are regulated by cyclic AMP (cAMP) signaling cascade, and that the decrease in cAMP levels in mature stages results in a slowdown of NPP activity. We also show that NPPs facilitate the uptake of artemisinin derivatives and that phosphodiesterase (PDE) inhibitors can reactivate NPPs and increase drug uptake in mature gametocytes. These processes are predicted to play a key role in P. falciparum gametocyte biology and susceptibility to antimalarials.

Fig. 1. NPPs are still active in immature gametocytes.

a Kinetics of sorbitol-induced isosmotic lysis of erythrocytes infected with early rings (4 h-post-invasion (hpi)), late rings (16 hpi), trophozoites (28 hpi), schizonts (40 hpi), stage II gametocytes and uninfected erythrocytes during 60 min. n = 3 independent experiments. b % lysis of stage II GIE in sorbitol at 60 min with or without 100 µM NPPB, Furosemide, Ro5-4864, or PK11195. c Lysis kinetics (left) and % lysis at 60 min (right) of stage II GIE in sorbitol, alanine or PhTMA⁺. d Patch-clamp experiments on erythrocytes infected with trophozoites, (green bar, number of cells = 14), stage II gametocytes (blue bar, number of cells = 14) or uninfected erythrocytes (red bar, number of cells = 11). Left: whole-cell conductance calculated at −100 mV. Right: I−V plot from patch-clamp experiments. e Patch-clamp experiments on stage II GIE in presence or absence of NPPB (number of cells = 14 and 5, respectively). Left: whole-cell conductance calculated at −100 mV. Right: I−V plot from patch-clamp experiments. f Viability (luciferase activity) of early gametocytes of the NF54-cg6-pfs16-CBG99 line 48 h after a 3-h incubation with or without 100 µM NPPB. The graph shows relative viability normalized by the average luciferase activity of control (without NPPB). a.u. arbitrary units. In (b, c, f), circles indicate the number of independent experiments. In (d, e), the data distribution is shown in Supplementary Fig. S1. Error bars show the standard error of the mean (SEM). Statistical significance is determined by a Mann−Whitney test (d−f) or by one-way ANOVA with Dunnet correction (b) or Sidak correction (c) for multiple comparisons. ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05, ns: non-significant difference.

Fig. 2. NPPs decline along gametocytogenesis.

a–c Sorbitol-induced (a), alanine-induced (b) and PhTMA⁺-induced (c) isosmotic lysis of GIE from stage I to stage V. Left: Kinetics of isosmotic lysis during 60 min. Right: % lysis at 60 min with (orange) or without (blue) 100 µM NPPB. Circles indicate the number of independent experiments and error bars show the SEM. Statistical significance is determined by one-way ANOVA with Sidak correction for multiple comparisons. d Left: I−V plot from patch experiments on GIE from stage I to stage V. Right: Whole-cell conductance calculated at −100 mV on GIE from stage I to stage V with (orange) or without (blue) 100 µM NPPB. Circles indicate the number of independent experiments and error bars show the SEM. Statistical significance is determined by a Mann−Whitney test at each gametocyte stage and by ANOVA test for trend between stage 1 and stage 5, p = 0.0013, slope −134.7 ± 40.03 pS/stage transition. ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05, ns: non-significant difference.

Fig. 3. NPP activity is regulated by cAMP-signaling.

a, b Sorbitol-induced isosmotic lysis of stage II GIE with 100 µM H89 or 10 µM KT5720 (a), or with 100 µM 8Br-cAMP (b). c Sorbitol-induced isosmotic lysis of stage II GIE from the NF54 isolate (Control) and the transgenic pHLpfpkar line, cultivated with or without pyrimethamine (Pyri), or preincubated with 100 µM 8Br-cAMP. d, e Sorbitol-induced isosmotic lysis of stage V GIE with 100 µM 8Br-cAMP (d), sildenafil (e) or tadalafil (e). f Sorbitol-induced isosmotic lysis of stage V gametocytes from the NF54 isolate and the transgenic line PDEδ⁻. All experiments were performed in the presence (orange) or absence (blue) of 100 µM NPPB. Circles indicate the number of independent experiments and error bars show the SEM. Statistical significance is determined by one-way ANOVA with Sidak correction for multiple comparisons. ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05, ns: non-significant difference.

Fig. 4. NPPs contribute to the uptake of artemisinin derivatives.

a Diagram illustrating the uptake assay. b Left: Quantification of Fluo-DHA uptake in uninfected erythrocytes and in early GIE by flow cytometry. Right: scatter plots showing the gating strategy for Fluo-DHA uptake. c Competition assay between Fluo-DHA and DHA, and control of HPA-NBD and DHA-induced fluorescence levels. d Inhibition of Fluo-DHA uptake in early GIE upon 100 µM NPPB or Furosemide incubation. e Viability (luciferase activity) of early gametocytes of the NF54-cg6-pfs16-CBG99 line 48 h after a 3-h incubation with 150 nM Fluo-DHA or 5 µM artemisinin, with or without 100 µM NPPB. The graph shows the ratio of luciferase activity (drug-treated/control) and is normalized to the condition without NPPB. a.u. arbitrary units. f Quantification of Fluo-DHA uptake during gametocytogenesis with (orange) or without (blue) 100 µM NPPB. g Fluo-DHA uptake in early GIE upon 100 µM H89 or 10 µM KT5720 incubation. h Fluo-DHA uptake in early GIE from NF54 (Control) and the transgenic pHLpfpkar line, cultivated with or without pyrimethamine (Pyri), or preincubated with 100 µM 8Br-cAMP. i Fluo-DHA uptake in stage V GIE upon 0, 10, 30 and  100 µM tadalafil incubation. j Left: % inhibition of gamete egress after a 24-h incubation with 5 µM artemisinin, with or without 30 µM tadalafil. Right: gamete egress observed by IFAs. Samples were co-stained with mouse anti-glycophorin A (GPA, red) and rabbit anti-Pfg27 (green) IgG. DIC differential interference contrast. Arrow: mature GIE with an intact erythrocyte membrane, arrowhead: egressed gamete. Scale bars: 5 μm. Circles indicate the number of independent experiments and error bars show the SEM. Statistical significance is determined by one-way ANOVA with Dunnet correction (b−d, g, i) or with Sidak correction (f, h, j) for multiple comparisons or by a Mann−Whitney test (e). ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05, ns: non-significant difference.

Fig. 5. Model for cAMP-mediated regulation of NPP activity.

Left: In early GIE, PfPDEδ expression is low, resulting in high cAMP levels in the parasitophorous vacuole and in the host cell, thereby activating the human PKA. PKA phosphorylates one or several proteins directly or indirectly involved in NPP activity that contributes to the uptake of drugs. Middle: In mature GIE, PfPDEδ is highly expressed and degrades cAMP, leading to a decrease in PKA phosphorylation and NPP activity. Right: In mature GIE, inhibition of PfPDEδ by tadalafil results in increased levels of cAMP that reactivate PKA and NPPs, thereby restoring uptake of drugs.