P. falciparum ALV5, PhIP and GAPM2 proteins co-localise with PhIL1 in the Inner membrane complex
We have previously shown the expression and localization of PhIL1 in the IMC of the P. falciparum parasite and its interaction with some of the components of glideosomal complex (Saini et al., 2017). To investigate the existence of PhIL1-associated novel IMC protein complex in P. falciparum, we selected three proteins: a conserved protein of unknown function (PF3D7_1310700), referred here as PhIL1 Interacting Protein (PhIP); Alveolin 5 (PF3D7_1003600 or IMC1c), an IMC structural protein; and GAPM2 (PF3D7_0423500), and expressed these as GFP-tagged fusion proteins transgenic parasite lines (Supplemental Figure S1A). Expression of fusion protein was confirmed by western blot analysis of lysate from the transgenic parasites using anti-GFP antibodies (Supplemental Figure S1B, C and D). GFP-specific antiserum recognised a band of ~60 kDa for ALV5-GFP, ~43 kDa for PhIP-GFP and ~70 kDa for GAPM2-GFP fusion protein. Transgenic parasites expressing ALV5-GFP showed peripheral localisation in the schizont stage and in merozoites (Figure 1A). Parasites expressing PhIP-GFP or GAPM2-GFP showed similar pattern of peripheral localization in the IMC, as shown for PfPhIL1 (Figure 1B and 1C). These proteins co-localised with PhIL1 in the IMC at schizont stage of the parasite with a Pearson’s colocalization coefficient of more than 0.7 in an indirect immunofluorescence assay (Figure 1D).
Subsequently, to establish the identity of PhIL1 associated novel complex, we performed pull-down assays using GFP-Trap beads with the parasite extracts prepared from these transgenic lines. Immunoprecipitates were digested with trypsin and peptides were analysed by mass spectrometry to identify the interacting partners. The glideosomal proteins GAP50, glideosome associated proteins with multiple membrane spans (GAPMs) 1, -2 and -3, and alveolin/ IMC protein family were identified in each precipitate, together with PhIP (PF3D7_1310700) and ALV5 (Table 1). Overall, these results confirmed the interactions among these proteins as well as with the PhIL1 protein.
PhIL1 associated novel complex is closely associated with the glideosomal complex
To ascertain the association of PhIL1 associated protein complex with the glideosomal complex, we performed sedimentation analysis of P. falciparum 3D7 schizont/ merozoite lysate using glycerol density gradient centrifugation. Western blot analysis of the glycerol gradient fractions using anti-PfALV5, anti-PfPhIP, anti-PGAPM2, anti-PfPhIL1, and anti-PfGAP50 antibodies; revealed that these proteins co-sedimented together in fractions 5 to 11, particularly in fraction 9 corresponding to ∼250 kDa molecular mass (Figure 2A and supplemental figure S2), suggesting that these proteins probably are associated together in the parasite.
To further substantiate these results, schizont stage parasite lysate was subjected to blue native PAGE (BN-PAGE) followed by immunoblot analysis. As shown in figure 2B, we detected a high–molecular mass complex of ∼800 kDa consisting of ALV5, PhIP, GAPM2, PhIL1 and GAP50. In addition, we observed another low molecular weight complex with anti-PhIL1 and anti-GAPM2 antibodies. Together, these results suggested that PhIL1 is associated with two different complexes; a high molecular weight complex of ~800kDa consisting of components of PfPhIL1 associated complex and GAP50 and a low molecular weight complex of ~250kDa consisting of PhIL1 and GAPM2 (Figure 2B), indicating the heterogeneity among these two complexes, which are composed of different but overlapping proteins .
Taken together, results presented here (Figures 1 and 2) validate the association of these proteins with each other, and in particular, their interactions with select, but not all components of the glideosome machinery. The data thus illustrates that these proteins probably form an independent complex in the IMC, which may have a diverse role than the glideosome complex. Based on the above results, we propose that PhIL1 forms a novel complex probably in the outer IMC, having overlapping components with the glideosomal motility complex. The organisation of the proposed novel PhIL1 associated complex is depicted in Figure 2C.
PhIL1 associated complex plays an important role in parasite growth and invasion
To address the role of PfALV5, PfPhIP and PfGAPM2 proteins in the P. falciparum IMC, we generated conditional knock-down parasite lines expressing respective genes in fusion with HA-glmS. The glmS ribozyme is expressed downstream of the target gene, which is efficiently knocked down in response to glucosamine (GlcN). Strategy for generating the knock-down lines is presented in Supplemental Figure S3A. Integrates were enriched with blasticidin selection and cycling, followed by clonal selection of a single transgenic parasite by limiting dilution. Integration into the parasite genome was confirmed by PCR (Supplemental Figure S3B, C and D).
Expression and efficient knockdown of the fusion protein was analysed by western blot analysis of lysate from transgenic parasites using anti-HA antibody under the effect of GlcN inducer (Figure 3A, B and C inset). Ring stage parasites at 16-20 hpi were treated with GlcN (2.5 mM) and parasites were harvested at 42-44 hpi. The saponin lysed treated parasites were subjected to lysis in RIPA buffer followed by freeze-thaw cycles and the lysate was subjected to SDS-PAGE. PfBiP, a constitutively expressed endoplasmic reticulum chaperon protein was used as a loading control. Apparent knockdown of up to ~80-85% was achieved for the expression of ALV5-HA, PhIP-HA and GAPM2-HA in the respective parasite lysates.
To study the growth of knockdown parasites, GlcN was added at the ring stage parasites 16-20 hpi at varying concentrations (0.6 mM, 1.25 mM, 2.5 mM, and 5 mM) and the growth was monitored till the formation of new rings i.e. up till one invasion cycle. Loss of ALV5 had little effect in the invasion of human RBCs by merozoites in comparison to the wild type parasites (Figure 3A), however PhIP depleted parasites showed ~80% invasion inhibition at 5 mM glucosamine (Figure 3C) while reduction in GAPM2 levels exhibited an invasion inhibitory potential of ~80% at 1.25 mM concentration of GlcN (Figure 3E).
Representative Giemsa stained smears of the ALV5 knockdown parasites showed no significant difference apart from a slightly delayed parasite growth cycle in the GlcN treated (5 mM) and untreated parasites, suggesting that ALV5 is not essential for parasite growth (Figure 3B). In comparison, the PhIP-HA-glmS parasites showed arrested development of schizonts in the knockdown line (1.25 mM GlcN treatment) (Figure 3D). Following PhIP knockdown, a proportion of schizonts phenotypically displayed incomplete segmentation showing agglomerates of unsegmented merozoites in close proximity to the food vacuole, which might be due to failure of IMC formation. Some of the parasites, which were fully segmented and merozoites egressed normally, however, these merozoites were unable to invade and newly released merozoites were seen arrested on the surface of RBC suggesting that, despite the initial attachment, parasite was unable to penetrate into the host RBC . In PhIP knockdown parasites two distinct parasites were observed, ~43% showed unsegmented merozoites, while 57% of the PhIP depleted parasites were found to be arrested at erythrocyte surface after attachment (Figure 3D-zoom and supplemental figure S4). In the absence of GlcN, distinct merozoites were observed enclosed in schizonts and these merozoites invaded normally as seen with their ability to progress to ring stage (Figure 3D). By contrast, treatment of GAPM2-HA-glmS with GlcN resulted in normal merozoite egress, however released merozoites were found to be stuck at erythrocyte surface, indicative of their inability to invade the RBC. However, we did not observe any defect in the attachment of these merozoites to RBC surface (Figure 3F).
PfPhIP knock-down causes underdeveloped IMC/plasma membrane during schizogony
We further studied the defects in merozoite segmentation in PfPhIP knock-down parasites for the formation of the parasite plasma membrane (PPM), formation of and secretion by apical secretory organelles and IMC formation by immunostaining these parasites with either with anti-GAP50, or anti-MSP1, or anti-AMA1 antibodies. Briefly, schizonts maintained with and without GlcN from the early ring-stage were treated with 10 mM E64 at 42 hpi, to prevent release of daughter merozoites In PhIP knockdown parasites, multiple daughter cells remained partially attached to each other. In these parasites, we were able to observe residual signal for PhIP as the efficiency of knockdown was ~80%, not the complete knockout. In PfPhIP-knockdown parasites, segmented merozoites showed detectable PfPhIP staining by IFA, whereas in the residual unsegmented agglomerate PfPhIP staining was not detected (Figure 4A). PhIP depleted parasites showed apparent loss of signal for GAP50 in the multi-nucleated agglomerates suggesting defect in IMC formation in agglomerates while [-] GlcN parasites, showed well-formed IMC around each nucleus of the segmented schizont (Figure 4B). Parasite plasma membrane which coats the individual newly formed daughter cells was examined by Merozoite Surface Protein 1 (MSP1). PhIP depleted schizonts showed MSP1 staining enclosing multiple nuclei of the agglomerate inside one contiguous membrane in contrast to untreated parasites that showed MSP1 surrounding each segmented daughter merozoite nucleus discretely (Figure 4C and Supplemental Figure S5A). Thus, in the agglomerates, PhIP knock-down parasites failed to direct the PPM around single daughter nuclei. Simultaneously, microneme formation and secretion was assessed using anti-PfAMA1 (Apical Membrane Antigen 1) antibody in [-] GlcN as well as [+] GlcN parasites. [-] GlcN PhIP-HAglmS parasites showed AMA1 staining around each merozoite, whereas PhIP depleted parasites demonstrated surface AMA1 staining in fully segmented merozoites, whereas loss of AMA1 signal on merozoite surface was observed for the multi-nucleated agglomerates (Figure 4D and Supplemental Figure S5B). 3D reconstruction of schizont stage transgenic parasites with and without GlcN with respect to different marker antibodies is illustrated using Imaris, version 7.6.1 (Bitplane) (Figure 4E).
These results established that PfPhIP-deficient parasites show developmental defect during the final stages of schizont segmentation that fail to reinstate the asexual blood cycle due to structural defects.
PfPhIP and PfGAPM2 knock down results in generation of non-invasive merozoites defective in apical organelle secretion
Since we observed that merozoites in PfPhIP and GAPM2 knock-down parasites were able to attach to the RBC surface, but could not invade, we next performed immunostaining of these RBC attached but non-invasive merozoites, to determine whether the inability of merozoites to invade the RBCs is due to defects in apical organelles biogenesis or their secretion, which is crucial for formation of invasion complex or due to the inability of motility complex which fails to propel the invading merozoite into the host RBC. To understand whether the IMC formation is affected in these knock-down parasites, we stained these attached merozoites with anti-PfGAP50 antibody. Importantly, PhIP deficient merozoites showed loss of signal for GAP50 suggesting defects in IMC formation (Figure 5A; panel 1) while GAPM2 depleted merozoites displayed intact IMC encircling the nascent attached cell (Figure 5B; panel 1). To assesses the merozoites released from the PhIP or GAPM2 deficient schizonts for the formation and secretion of apical organelles, we used α-EBA175, and αAMA-1 antibodies as markers for micronemes, and α-PfRON2 antibody as a rhoptry marker. We observed typical micronememal staining as visualized by the presence discrete dots with anti-EBA175 and anti-AMA1 antibodies. However, we could not locate these proteins on merozoite surface in the merozoites released from the PhIP and GAPM2 knockdown schizonts, therefore indicating the failure to discharge the micronemal contents post egress (Figure 5A and B; panel 2 and 3). We additionally evaluated the presence of PfRON2, a rhoptry marker in merozoites in PfPhIP- and PfGAPM2-knockdown parasites. These knock-down parasites displayed characteristic rhoptry localisation (Figure 5A and B; panel 3). Together, these results indicated that merozoites formed and released in PfPhIP- and PfGAPM2-knockdown parasites looked morphologically normal as indicated by their staining with the panel of cell markers. However, apical organelle secretion of the invasion ligands seems to be affected in the merozoites released from schizonts with PfPhIP and PfGAPM2 deficiency (Supplemental Figure S6A and B). 3D reconstruction of stuck merozoites from transgenic parasites with GlcN with respect to different marker antibodies are illustrated using Imaris, version 7.6.1 (Bitplane) (Figure 5C).
Since, the secretion of EBA175, AMA1, and RON2, failed to initiate following merozoite attachment in PfPhIP- and PfGAPM2-deficient merozoites, it clearly suggested that the signal for invasion in these merozoites post attachment to the erythrocytes was not triggered in PfPhIP- and PfGAPM2-knockdown parasites. On detailed analysis, we observed ~87% of the attached merozoites failed to align their apical end towards the erythrocyte surface i.e. the apex of the merozoite is not in direct proximity of the erythrocyte surface as indicated by staining with apical marker proteins. Apical reorientation is imperative for triggering commitment to invasion. Taken together, the analysis of attached PfPhIP- and PfGAPM2-deficient merozoites suggested roles of PfPhIP and GAPM2 in the reorientation of merozoites so that the apical organelles are aligned to the erythrocyte membrane (Figure 6). This study also highlights that merozoite reorientation might directly or indirectly mediated by the motor complex due to existence of the overlapping components among these two complexes.