A revised mechanism for how Plasmodium falciparum recruits and exports proteins into its erythrocytic host cell

Abstract

Plasmodium falciparum exports ~10% of its proteome into its host erythrocyte to modify the host cell's physiology. The Plasmodium export element (PEXEL) motif contained within the N-terminus of most exported proteins directs the trafficking of those proteins into the erythrocyte. To reach the host cell, the PEXEL motif of exported proteins is processed by the endoplasmic reticulum (ER) resident aspartyl protease plasmepsin V. Then, following secretion into the parasite-encasing parasitophorous vacuole, the mature exported protein must be unfolded and translocated across the parasitophorous vacuole membrane by the Plasmodium translocon of exported proteins (PTEX). PTEX is a protein-conducting channel consisting of the pore-forming protein EXP2, the protein unfoldase HSP101, and structural component PTEX150. The mechanism of how exported proteins are specifically trafficked from the parasite's ER following PEXEL cleavage to PTEX complexes on the parasitophorous vacuole membrane is currently not understood. Here, we present evidence that EXP2 and PTEX150 form a stable subcomplex that facilitates HSP101 docking. We also demonstrate that HSP101 localises both within the parasitophorous vacuole and within the parasite's ER throughout the ring and trophozoite stage of the parasite, coinciding with the timeframe of protein export. Interestingly, we found that HSP101 can form specific interactions with model PEXEL proteins in the parasite's ER, irrespective of their PEXEL processing status. Collectively, our data suggest that HSP101 recognises and chaperones PEXEL proteins from the ER to the parasitophorous vacuole and given HSP101's specificity for the EXP2-PTEX150 subcomplex, this provides a mechanism for how exported proteins are specifically targeted to PTEX for translocation into the erythrocyte.

Fig 1. Generation and validation of HSP101-HAglmS parasite line and knockdown analysis.

(A) The P. falciparum HSP101 targeting construct integrated into the endogenous locus by a single crossover recombination event as indicated. Hemagglutinin (HA) and strep II (Str) epitope tags, human DHFR selectable marker (hDHFR), glmS ribozyme and untranslated regions (UTR) are shown. Arrows indicate oligonucleotides used in diagnostic PCR analysis and their product sizes. (B) Diagnostic PCR showing the hsp101 gene contains the integrated sequence. Oligonucleotide pairs shown in (A) were used on genomic DNA prepared from 3D7 wildtype parental parasites (WT) or from two independent clones of drug-resistant parasites obtained after transfection with the targeting construct (C1, C2). DO561 and DO562 recognize the hsp101 locus and serve as a positive control for the PCR. (C) Western blot analysis showing that the HSP101-HAglmS C1 and C2 express the HA epitope tags. Rabbit HSP101 and HSP70-1 antibodies serve as loading controls. (D) Representative western blot demonstrating HSP101 expression can be knocked down using glucosamine (GlcN). Left panel: Synchronised cultures of ring-stage PfHSP101-HAglmS parasites grown in the presence or absence of GlcN for 48 hours were harvested and parasite lysates were probed with anti-HA antibodies to detect HSP101-HA expression. HSP70-1 served as a loading control. Right panel: Densitometry performed on bands observed by western blot using ImageJ to quantitate the level of HSP101 expression relative to HSP70-1 in parasite lines grown in the presence or absence of GlcN. Shown is the mean ± SEM (n = 4 biological repeats). (E) Parasitemia of cultured PfHSP101-HAglmS parasites grown in 0 mM or 2.5 mM GlcN (left panel) and representative Giemsa-stained smears (right panel) shows knockdown of HSP101 leads to the arrest of parasite growth in ring stages, with parasites unable to transition into trophozoite stage. Shown is the mean ± SEM (n = 3 biological repeats). (F) Immunofluorescence analysis of ring-stage parasites treated with 2.5 mM GlcN for 48 hours and fixed with acetone/methanol and labelled with anti-HA antibodies to detect HSP101-HA, or antibodies against the exported proteins SBP1 and GEXP07. This demonstrates that knocking down expression of HSP101-HA blocks the export of SBP1 and GEXP07 into the host erythrocyte. Parasite nuclei were stained with DAPI (4’,6-diamidino-2-phenylindole). Scale bar, 5 μm, DIC, differential phase contrast.

Fig 2. Knockdown of HSP101 does not disrupt the stability of the EXP2/PTEX150 subcomplex.

(A), (B), and (C) BN-PAGE analysis of PTEX-HAglmS parasite lines following protein knockdown enables the effects on PTEX complex formation to be examined. Saponin-lysed pellets were lysed with 1% digitonin, separated on 4–12% NativePAGE gel, and analysed via western blotting using PTEX-specific antibodies. Knockdown of EXP2 and PTEX150 prevented the formation of a >1236 kDa PTEX subcomplex. Knockdown of HSP101 in contrast, did not disrupt the ability of PTEX150 and EXP2 to form a >1236kDa complex free of HSP101. * and ** are two non-specific bands observed with rabbit anti-HSP101 antibody. (D) Schematic summarising the impact of knocking down individual PTEX components with regards to complex formation.

Fig 3. HSP101 exhibits distinct dual localisation inside the parasite and in the parasitophorous vacuole.

(A) Representative IFA images of parasites expressing HSP101-HA, EXP2-HA, or PTEX150-HA. Parasites were labelled with anti-HA IgG and nuclei were stained with DAPI. LR, Late ring stage. MT, Mid trophozoite stage. LT, Late trophozoite stage. Scale bar, 5 μm. (B) Graph of the ratios of internal parasite signal over PV signal for each cell line indicates HSP101-HA exhibits a comparatively greater internal signal than the other PTEX proteins. Data were pooled from at least two biological replicates, with >20 mid trophozoites (MT) parasites measured for each line. Statistical significances were determined using an ordinary one-way ANOVA. ****, p-value<0.0001). (C) Schematic of the proteinase K (Prot. K) protection assay to validate cellular localisation of HSP101. EqtII, equinatoxin II; TX-100, Triton X-100. (D) Western blots of permeabilised mid trophozoite stage parasites digested with increasing concentrations (0–20 μg/mL) of proteinase K. Sn, Supernatant. P, Pellet. (E) Densitometry analysis of the western blot in (D) showing the total protein digested with Prot. K in saponin (Sap) or TX-100 -treated parasites from the combined signals of the Sn and P fractions. Percentage normalised to the total band intensity (Sn and P fractions) of samples treated with 0 μg/mL Prot. K. The data indicate that compared to EXP2, more HSP101 is resistant to proteolytic degradation in saponin suggesting a large pool of HSP101 resides inside the parasite. Data were extracted from two independent biological replicates. Plotted data represents the mean ±SEM.

Fig 4. Brefeldin A treatment induces accumulation of HSP101 in the ER.

(A) Representative IFA images of HSP101-HAglmS parasites treated with 0, 2, and 10 μM Brefeldin A (BFA) for 24 hours starting from the ring stage. Cells were fixed and stained with anti-HA (HSP101) and anti-ERC antibodies (ER marker). (B) Quantification of the degree of co-localisation (Pearson’s coefficient) between HA and ERC fluorescence signal indicates that HSP101-HA becomes progressively more trapped within the ER at higher BFA concentrations. The Z-stack images of at least 50 cells were analysed for each treatment. Statistical significances were determined using ordinary one-way ANOVA. ****, p-value<0.0001. (C) Older ring stage parasites (16–20 hours post invasion) were treated with 0 and 18 μM BFA for 5 hours and stained with anti-HA (HSP101) and anti-PTEX150 (rabbit serum; r942) and anti-EXP2 (monoclonal antibody). The images indicate that HSP101 becomes more noticeably trapped within the parasite and depleted from the PV than EXP2 and PTEX150 although there is some accumulation of the latter two proteins within parasite after BFA treatment (white arrow, perinuclear staining of EXP2 and PTEX150). Scale bar, 5 μm. (D) Quantification of the Internal/PV ratio between the internal parasite and PV signal for the anti-HA, anti-PTEX150, and anti-EXP2 staining with or without 18 μM BFA treatment. > 20 cells were quantified for each treatment. Statistical significances were determined using an ordinary one-way ANOVA. ****, p-value<0.0001. ns, non-significant (p-value>0.05).

Fig 5. Intra-parasite HSP101 co-localises with P5 Lys Hyp1-Nluc-mDHFR-3xFLAG reporter construct in the endoplasmic reticulum of the parasite.

(A) Representative IFA images (n = 3 independent replicates) of parasites expressing Hyp1-Nluc-mDHFR-3xFLAG constructs with WT P5 glutamate reporter exported into the erythrocyte and the mutant P5 Lys Hyp1-Nluc-mDHFR-3xFLAG reporter trapped in the parasite ER. To visualise translocon components, cells were probed with anti-EXP2 and anti-HA (to visualise HA-tagged HSP101). Anti-PfERC was used to visualise the parasite’s ER. Hyp1-Nluc-mDHFR-3xFLAG reporter proteins were localised using either anti-Nluc or anti-FLAG antibodies. The red bar in the schematic picture of the construct indicates the PEXEL motif and variations thereof. The blue bar indicates the transmembrane signal peptide. Scale bars, 5 μm. DIC, Differential Interference Contrast. DAPI was used to stain parasite nuclei. (B) Representative Z-stacks of 3D-SIM images of the HA-tagged HSP101 parasite line expressing P5 Lys Hyp1-Nluc-mDHFR-3xFLAG and probed with either anti-HA for HSP101 (top, red) or anti-EXP2 (bottom, red), and anti-Nluc (green) to probe for the cargo. Scale bars represent 5 μm. (C) Degree of co-localisation between individual PTEX components to the P5 Lys Hyp1-Nluc-mDHFR-3xFLAG was calculated by measuring Pearson’s coefficients of the merged Z-stack images of >20 cells. Statistical significances were determined using an unpaired t-test with Welch’s correction. ****, p-value<0.0001). (D) Representative Western blot (n>3) of lysates made from mid-stage trophozoites expressing WT and P5 Lys Hyp1-Nluc-mDHFR-3xFLAG constructs showing full-length and cleaved forms of the WT Hyp1-Nluc-mDHFR-3xFLAG reporter. The P5 Lys Hyp1-Nluc-mDHFR-3xFLAG reporter appears to be miscleaved after the signal peptide (58.1 ± 1.2 kDa, n = 10). The blot was probed with anti-PTEX150 antibodies as a loading control.

Fig 6. ER-located HSP101 interacts with Hyp1-Nluc-mDHFR-3xFLAG irrespective of Hyp1 PEXEL processing.

(A) Schematic of parasite treatment and the subsequent co-immunoprecipitation to assess the interaction between Hyp1 reporters and HSP101. (B) Western blot of anti-Nluc immunoprecipitation performed with HSP101-HAglms parasites expressing WT or P5 Lys (K) Hyp1-Nluc-mDHFR-3xFLAG. Where indicated, parasites samples were crosslinked with 0.5 mM DSP and lysed with either 1% TX-100 buffer or RIPA buffer. The data indicate that the mutant PEXEL reporter engages with HSP101 despite residing in the ER and not being cleaved by plasmepsin V. (C) Reciprocal anti-HA immunoprecipitation was performed using the same samples as (B) to confirm that HSP101 was interacting with both the PEXEL-cleaved (WT) and non PEXEL-cleaved (P5 Lys) Hyp1-mDHFR-3xFLAG reporters. Parasites that expressed Hyp1-mDHFR-3xFLAG without a HA-tagged version of HSP101 were used as a negative control (Neg). For both immunoprecipitations, inputs (2%) and eluates (100%) were fractionated by SDS-PAGE. The result was reproducible in three independent experiments. (D) Schematic of parasite treatment with proteinase K and the subsequent co-immunoprecipitation experiment to examine if ER-resident HSP101 binds to Hyp1 reporter proteins. (E) Western blot analysis of co-immunoprecipitation of WT and P5 Lys Hyp1-mDHFR-3xFLAG and parental HSP101-HAglmS line (negative control, Neg) captured with anti-Nluc IgG. In situ crosslinking was performed using 0.5 mM DSP. Input (2%) and eluate (100%) were fractionated by SDS-PAGE. Immunoblots were stained with anti-FLAG antibody to visualise cargo, anti-HA to visualise HSP101 and anti-EXP2 to control for PVM disruption. The data indicate that both WT and P5 Lys Hyp1-mDHFR-3xFLAG bind to the proteolytically resistant ER-pool of HSP101. The result was reproducible in two independent experiments.

Fig 7. Plasmepsin V is not directly associated with HSP101 in the parasite ER.

(A) Western blot analysis of WT and P5 Lys Hyp1-Nluc-mDHFR-3xFLAG (K) proteins captured with anti-FLAG M2 agarose beads indicate that the ER-trapped P5 Lys Hyp1-Nluc-mDHFR-3xFLAG associates with both HSP101 and plasmepsin V. Infected erythrocytes were magnetically purified then crosslinked with 0.5 mM DSP, lysed, and incubated with anti-FLAG M2 agarose beads. As a negative control, the HSP101-HAglmS parental parasites were treated similarly (Neg). Input (1.5%) and eluate (37.5%) were fractionated by SDS-PAGE. Immunoblot detection was performed using an anti-HA antibody to detect HSP101 as well as antibodies to plasmepsin V (PMV). Antibodies to GAPDH were used as a negative control. The result was reproducible in three independent experiments. (B) Western blot of immunoprecipitated HSP101-HA from parasites expressing P5 Lys Hyp1-mDHFR-3xFLAG and probed with anti-plasmepsin V indicate the protease is not directly associated with HSP101-HA. Input (2%) and eluate (50%) were fractionated by SDS-PAGE. The result was reproducible in two independent experiments.

Fig 8. Model of PEXEL protein targeting to the vacuolar translocon by HSP101.

(A) PEXEL-containing proteins are imported after translation into the ER via a Sec61/62/63-SPC25-PMV translocon [26] and are recognised by HSP101, either during import or after PEXEL processing, prior to their release into the ER lumen. HSP101-bound cargo is then trafficked to the parasite membrane via the vesicular trafficking system and released into the PV. HSP101 affinity for the PTEX150-EXP2 subcomplex at the vacuolar membrane drives reconstitution of the full PTEX complex capable of translocating the cargo into the host cell. The model depicted here is described for soluble PEXEL proteins, but it is possible that the same mechanism governs the trafficking of transmembrane proteins. (B) Proteins with a classical N-terminal signal sequence destined for the PV such as SERA5, are imported into the ER by the Sec61-SPC25-SPC21 complex [26]. After signal sequence cleavage, the released proteins are folded and secreted into the PV. Since these proteins are not bound to HSP101, they cannot translocate via PTEX.