PfCERLI1 is a conserved rhoptry associated protein essential for Plasmodium falciparum merozoite invasion of erythrocytes

Abstract

The disease-causing blood-stage of the Plasmodium falciparum lifecycle begins with invasion of human erythrocytes by merozoites. Many vaccine candidates with key roles in binding to the erythrocyte surface and entry are secreted from the large bulb-like rhoptry organelles at the apical tip of the merozoite. Here we identify an essential role for the conserved protein P. falciparum Cytosolically Exposed Rhoptry Leaflet Interacting protein 1 (PfCERLI1) in rhoptry function. We show that PfCERLI1 localises to the cytosolic face of the rhoptry bulb membrane and knockdown of PfCERLI1 inhibits merozoite invasion. While schizogony and merozoite organelle biogenesis appear normal, biochemical techniques and semi-quantitative super-resolution microscopy show that PfCERLI1 knockdown prevents secretion of key rhoptry antigens that coordinate merozoite invasion. PfCERLI1 is a rhoptry associated protein identified to have a direct role in function of this essential merozoite invasion organelle, which has broader implications for understanding apicomplexan invasion biology.

Fig. 1. Phylogeny of PfCERLI1 (Pf3D7_0210600) and development of genetic tools to investigate function.

a PfCERLI1 is 446 amino acids in length and predicted to contain a signal peptide. b The amino acid sequence of PfCERLI1 was compared against Plasmodium spp. orthologues by multiple pairwise alignments. Pr = P. reichenowi, Pg = P. gaboni, Pm = P. malariae, Po = P. ovale, Pk = P. knowlesi, Pv = P. vivax, Pb = P. berghei, Py = P. yoelii, Pc = P. chabaudi. c Schematic representation of the HA-GlmS riboswitch system used to study PfCERLI1. A plasmid vector, containing a 3’ flank of the Pfcerli1 sequence (1277bp-2046bp) with a haemagglutinin (HA) tag and under the control of a GlmS ribozyme was transfected into wildtype parasites by 3’ single crossover recombination. Glucosamine (GLCN) binds to the GlmS ribozyme, promoting Pfcerli1 mRNA degradation. d Plasmid integration was confirmed by PCR using primers that amplify only WT Pfcerli1 locus (primer A and B) or primers that amplify only integrated Pfcerli1^HAGlmS locus (primer A and C). These PCR reactions showed that the majority of parasites had integrated Pfcerli1^HAGlmS into the correct flanking region (M = size ladder). e Western blot of ring (R), early trophozoite (ET), late trophozoite (LT) or schizont (S) stage PfCERLI1^HAGlmS lysates probed with anti-HA (PfCERLI1) or anti-EXP2 (loading control) antibodies. f Western blot of schizont stage PfCERLI1^HAGlmS lysates either in the presence (+) or absence (−) of 2.5 mM GLCN, which was then probed with anti-HA (PfCERLI1) and anti-EXP2 (loading control) antibodies, showing effective knockdown of PfCERLI1 in the presence of GLCN. g Synchronous PfCERLI1^HAGlmS or 3D7 trophozoite-stage parasites were treated with increasing concentrations of GLCN for 48 h, with the number of trophozoites the following cycle measured to determine knockdown-mediated growth inhibition (n = 3 biological replicates. Parasite growth expressed as a % of non-inhibitory media controls, error bars = standard error of the mean (SEM). Source data are provided in source data file). X-axis presented as a log 2 scale for viewing purposes.

Fig. 2. PfCERLI1 knockdown does not inhibit merozoite development but does prevent merozoite invasion.

a Flow cytometric detection of GFP-expressing PfCERLI1^HAGlmS/GFP ring stage parasites after merozoite invasion indicated a direct inhibition of merozoite invasion with protein knockdown (results presented as a % of media control, n = 4 biological replicates). PfCERLI1^HAGlmS Growth is replicated from Fig. 1g for direct comparison between growth and invasion inhibition. X-axis presented as log 2 scale for viewing purposes (Source data are provided as a source data file). b Mean number of fully segmented merozoites per schizont was determined by microscopy analysis of Giemsa-stained thin blood smears. Smears were blinded and counted with each data point representing the mean number of merozoites per schizont pooled from 20 schizonts, n = 3 biological replicates. c The percentage of schizont rupture that occurred over a 6-h window either with, or without, GLCN treatment (n = 3 biological replicates), d The number of free merozoites after GLCN treatment was assessed using flow cytometry, with results presented as % of total erythrocytes (n = 4 biological replicates). e Giemsa-stained PfCERLI1^HAGlmS schizonts, either GLCN treated or untreated, matured normally in the presence of E64. Scale bar = 2 µm. All error bars = SEM. *p < 0.05, ns = p > 0.05 by unpaired t-test.

Fig. 3. PfCERLI1 localises to the rhoptry bulb of merozoites.

a Immunofluorescence microscopy showing 3D reconstructed projections of confocal images with anti-HA (PfCERLI1) in green, colocalising more strongly with the rhoptry bulb marker RAP1 than with the micronemal marker CyRPA, rhoptry neck marker RON4 or the inner-membrane complex marker GAP45. b Quantification of the colocalisation, by Pearson’s correlation coefficient when PfCERLI1 (anti-HA) staining is defined as the region of interest, with the following merozoite organelle markers: GAP45, CyRPA, RON4, and RAP1. (****p < 0.0001 by Analysis of variance (ANOVA), n = 3 biological replicates, with 6 merozoite containing schizont images taken from each, error bars = SEM). c Maximum intensity projections of super-resolution immunofluorescence microscopy (Airyscan) of PfCERLI1^HAGlmS schizonts stained with anti-HA and anti-RAP1 antibodies. Yellow box (top left panel) indicates the zoom area for the free merozoite depicted in the other three panels.

Fig. 4. PfCERLI1 is peripherally-associated with the cytosolic face of the rhoptry membrane.

a PfCERLI1^HA schizont cultures were used for a proteinase K protection assay. Parasites were treated with either saponin alone, saponin and digitonin or saponin, digitonin and proteinase K and probed with anti-HA antibodies (PfCERLI1). GAP45 (inner-membrane complex, exposed to the cytosol) and RAP1 (rhoptry lumen) serving as positive and negative controls, respectively, for proteinase K digestion. Selected images representative of three independent experiments. b To determine membrane association, PfCERLI1^HA schizont cultures were subjected to a solubility assay. Saponin lysed parasite cultures were hypotonically lysed before being treated with sodium carbonate and then Triton-X-100 (Tx100), with supernatants and the Tx100 insoluble fraction being reserved after each treatment. Resulting samples were probed with anti-HA antibodies (PfCERLI1). The presence of a strong band in the carbonate treatment indicates release of the majority of PfCERLI1 protein into the supernatant with this treatment. Membranes were also stained with anti-EXP2 (transmembrane domain containing), anti-GAPDH (cytosolic) and anti-RH5 (peripheral) solubility controls. Selected images representative of three independent experiments. c A single z-slice of the double-bulbous rhoptries displaying the scheme that was used to measure the fluorescence intensity peaks (d) and diameter (e) of anti-HA and anti-RAP1 staining across the rhoptry (****p < 0.0001 by unpaired two-tailed t-test, n = 5 biological replicates, 1139 rhoptries measured for PfCERLI1^HA and 1040 for RAP1, Error bars = SEM). f 3D rendered image of a free merozoite showing the nucleus (DAPI) localising at the basal surface, PfCERLI1 (HA) wrapping around RAP1 at the rhoptry bulb and RON4 localising in the rhoptry neck at the far apical tip. g Representative image of Compound 1 treated PfCERLI1^HAGlmS schizonts that were fixed, labelled with anti-HA antibodies and probed with 18 nm colloidal gold secondary antibodies, before being imaged using transmission electron microscopy. White arrows mark rhoptry bulb (RB) and neck (RN), while black arrows mark PfCERLI1 foci. Scale bar = 500 nm.

Fig. 5. PfCERLI1^HAGlmS knockdown alters rhoptry antigen distribution.

a 3D reconstructions of super-resolution (Airyscan) micrographs for PfCERLI1^HAGlmS schizont cultures that were either left untreated or treated with glucosamine (+2.5 mM GLCN) before being stained with DAPI (nucleus), anti-HA (PfCERLI1), anti-RAP1 (rhoptry bulb) and anti-RON4 (rhoptry neck) antibodies. According to the image analysis pipeline detailed in Supplementary Fig. 7, the (b) % of RON4 that colocalised with RAP1, (c) Pearson’s correlation coefficient (PCC) of the amount of RON4 colocalising with the RAP1 signal, and (d) thresholded Mander’s correlation coefficient (MCC) of RON4 overlap with RAP1 signal was compared between PfCERLI1^HAGlmS knockdown and control parasites. Each data point represents a single schizont image (n = 5 biological replicates, 38 schizonts imaged for untreated and 36 for +2.5 mM GLCN schizonts). e Rhoptry bulb (RAP1) diameter was measured between untreated and PfCERLI1^HAGlmS knockdown (+2.5 mM GLCN) parasites. Each data point represents a single rhoptry (n = 5 biological replicates, 1220 rhoptries counted for untreated and 1217 for +2.5 mM GLCN). (ns = p > 0.05, **p < 0.01, ***p < 0.001 by one-way unpaired t-test). All error bars = SEM.

Fig. 6. PfCERLI1^HAGlmS knockdown alters rhoptry antigen processing and secretion.

a PfCERLI1^HAGlmS parasites were either treated with 2.5 mM GLCN (+) or left untreated (−), with the erythrocyte surface receptors of the culture cleaved by enzyme treatment to prevent invasion but allow release of merozoite antigens. The parasite lysate (P) and culture supernatant (SN) were stained with anti-RON4 (rhoptry neck), anti-EBA175 (micronemes), anti-RH4 (rhoptry neck), anti-RAP1 (rhoptry bulb), anti-HA (PfCERLI1) and anti-ERC (loading control) antibodies and analysed by western blot. Selected blots representative of five independent experiments (M = size markers). b Western blots of rhoptry antigens, and secreted EBA175, were normalised to the loading control (ERC) and quantified, with data represented as +2.5 mM GLCN signal expressed as a % of the signal for the untreated control (n = 5 biological replicates, source data are provided as a source data file). c Using the parasite lysates from the rhoptry secretion experiment, each of the three individual RAP1 bands present on the western blots were quantified and presented as a percentage of the total RAP1 signal (n = 5 biological replicates, ns = p > 0.05, *p < 0.05, **p < 0.01 by two-way ANOVA). All error bars = SEM.