Fig 1.
Apicoplast-targeting drugs are lethal to Plasmodium parasites in the asexual blood stage due to direct or indirect inhibition of IPP production.
A) Delayed death is a consequence of inhibition of plastid housekeeping functions in P. falciparum. Daughter parasites inherit a dysfunctional apicoplast and so are unable to synthesise the essential metabolite isopentenyl pyrophosphate (IPP), resulting in parasite death only in the cycle following treatment. Supplementation with the prenyl precursor GGOH temporarily rescues delayed death, allowing parasites to progress to the end of the second cycle before succumbing to IPP starvation. B) IPP is synthesised in the apicoplast and exported into the cytosol. Multiple IPP units are condensed together to form polyisoprenoids chains which are used for dolichol synthesis in the ER. Dolichol-phosphate-mannose (Dolichol-P-Man) is synthesised at the cytosolic face of the ER, then flipped into the ER lumen, where it donates essential mannose residues to build GPI anchors which are used to tether proteins with glycophosphatidylinositol (GPI) attachments motifs to membranes. Parts of this figure were created using Biorender.com.
Fig 2.
Metabolic labelling of P. falciparum glycolipids with GDP-[3H]-mannose.
A) Fluorograph of HPTLC separating P. falciparum 3H-Man-labelled GPIs putatively identified according to Gerold et al. 1999 [34]. Chemical and enzyme treatments of labelled glycolipids fractions. B) Fluorograph of HPTLC separating [3H]-Man-labelled P. falciparum glycolipids from untreated and indolmycin [50 μM] treated parasites (1x109 cell equivalents/mL) collected 72–78 h post drug administration (equivalent to 28–32 hpi in the second IDC after treatment). The glycolipids were divided into equal fractions and chemically treated as indicated. Dol-P-Man was detected in untreated and was resistant to mild base treatment (+ Base). GPI intermediates were detected in untreated and were resistant to mild acid treatment (+ Acid). In indolmycin treated parasites Dol-P-Man was below the limit of detection and there was less [3H]-man incorporation into GPI. C) Parasite lysate (1x109 cell equivalents/mL) were analysed by SDS-PAGE and Coomassie blue as a loading control.
Fig 3.
Surface GPI-anchored proteins lose their membrane association in second cycle partially rescued delayed death (PRDD) P. falciparum.
A) The localisation of surface GPI-anchored proteins was examined along with proteins associated with membranes via dolichol independent mechanisms. The GPI-anchored protein merozoite surface protein 2 (MSP2) (green) was co-localised (red) with B) GPI-anchored MSP1 C) transmembrane protein apical membrane antigen (AMA1), D) membrane spanning pore exported protein 2 (EXP2) and E) glideosome associated protein 45 (GAP45)–membrane associated by attachment of palmityl and myristoyl groups. Apicoplast inhibition with translation inhibitors results in surface GPI-anchored proteins lose their membrane association. Nuclei stained with 4,6-diamidino-2-phenylindole (DAPI; blue). Scale bar = 5 μm. All experiments were performed at least in triplicate and additional images are shown in S2–S5 Figs). Parts of this figure were created using Biorender.com.
Fig 4.
Detergent solubility assays further confirm the PV localisation of GPI-anchored proteins following drug treatment.
A) Following permeabilisation by saponin and centrifugation, the resulting supernatant (SN) is collected to assay for proteins soluble in the red blood cell (RBC) or parasitophorous vacuole (PV). The parasite pellet is then treated with Triton X114 (TX-114) and phase separation is used to partition proteins into an aqueous and detergent phase to enrich for soluble and membrane associating proteins respectively. B) Representative western blots showing that GPI-anchored proteins MSP1 and MSP2 partition in the detergent phase, with AMA1 and BIP as the as the membrane bound and soluble control respectively. After apicoplast inhibition MSP1 and MSP2 are found in the saponin supernatant, consistent with a portion of these proteins having lost their GPI and releasing into the PV. C) Treatment of parasite extracts with a GPI-specific phospholipase D removes the membrane anchor of MSP1, causing it to become enriched in the aqueous phase. Parts of this figure were created using Biorender.com.
Fig 5.
Second cycle PRDD schizonts cannot efficiently segment or egress from their host RBC.
A) Giemsa blood smears show that PRDD stall at the schizont stage and are unable to establish a third cycle of infection. B) To test parasites egress, an assay was performed with parasites transfected with an exported Nanoluciferase (Nluc). Upon RBC lysis, Nluc is released into the supernatant, which is collected. The NanoGlo substrate is added, with the relative light released used as a proxy for egress. Parasite egress can be reversibly blocked with the inhibitor compound 1 (C1). C) Purified schizonts were plated out in the presence of either DMSO vehicle or C1 at 4 μM and the supernatant collected after 4 or 8 hours. PRDD parasites showed egress comparable to incubations with C1. Each dot represents a technical replicate normalised to the DMSO 100% control, with 3 technical replicates in each of the 3 biological experiments. Error bars = SD. Unpaired t-test, p value < 0.05. D) IFAs showing persistence of the parasitophorous vacuole membrane (PVM) structure as seen by maintained EXP2 labelling. The egress defect of PRDD schizonts involves an inability to completely rupture their PVM long after typical egress occurs. BF, bright field E) These schizonts are also unable to segment appropriately, with new inner membrane complex (IMC) forming around only some nuclei as seen by labelling of the IMC protein GAP45. Nuclei stained with DAPI (blue). Images represent single Z stacks. Scale bar = 5 μm. F) Electron microscopy shows that while some nuclei (Nu) are not successfully segregated and enveloped by a new parasite plasma membrane (PPM), apparently normal rhoptry formation still occurs in PRDD parasites. Hpi, hours post invasion; M, merozoite; R, rhoptry; S PPM, schizont parasite plasma membrane; M PPM, merozoite parasite plasma membrane. Scale bar = 1 μm. Parts of this figure were created using Biorender.com.
Fig 6.
Merozoites derived from PRDD second cycle schizonts are incapable of invasion.
A) Giemsa smears show parasite development. Magnet enriched schizonts matured in the presence of E64 before their merozoites were mechanically extracted from the host RBC. Merozoites were incubated with fresh RBCs and parasite growth was monitored over 20 and 44 hours. Scale bar = 7 μm. B) Live cell imaging of purified merozoites with MitoTracker staining shows parasites are viable following their purification. To determine whether merozoite have invaded, flow cytometry was performed 20 hours after parasites were added to fresh RBCs. Scale bar = 1 μm. C) Gate setting for flow cytometry analysis performed using FLowJo software. RBCs were selected using SSC-H and FSC-A gating to exclude uninvaded parasites. D) PRDD parasites have negligible invasion, comparable to that of parasites treated with the invasion inhibitor heparin. Each dot represents a technical replicate normalised to the DMSO 100% control, with 3 technical replicates in each of the 3 biological experiments. Error bars = SD. Unpaired t-test, p value < 0.05.
Fig 7.
Dolichols are essential for survival of asexual stage P. falciparum parasites.
Parasites treated with apicoplast inhibitors cannot synthesise IPP in their subsequent cycle. Without this essential metabolite, these parasites become depleted in dolichols which are required for the synthesis of GPI anchors and the normal membrane association of their anchored proteins. Upon mislocalisation of GPI-anchored proteins normal schizont development is halted with incomplete segmentation. These schizonts proved incapable of egress, and the merozoites that do form are unable to invade new RBCs even when mechanically purified from mature segmented schizonts. Parts of this figure were created using Biorender.com.