To mediate its survival and virulence the malaria parasite Plasmodium falciparum exports hundreds of proteins into the host erythrocyte 1 . In order to enter the host cell, exported proteins must cross the parasitophorous vacuolar membrane (PVM) within which the parasite resides, but the mechanism remains unclear. A putative Plasmodium translocon of exported proteins (PTEX) has been suggested to be involved for at least one class of exported proteins; however, direct functional evidence for this has been elusive 2–4 . Here we show that export across the PVM requires heat shock protein 101 (HSP101), a ClpB-like AAA+ ATPase component of PTEX. Using a chaperone auto-inhibition strategy, we achieved rapid, reversible ablation of HSP101 function, resulting in a nearly complete block in export with substrates accumulating in the vacuole in both asexual and sexual parasites. Surprisingly, this block extended to all classes of exported proteins, revealing HSP101-dependent translocation across the PVM as a convergent step in the multi-pathway export process. Under export-blocked conditions, association between HSP101 and other components of the PTEX complex was lost indicating that the integrity of the complex is required for efficient protein export. Our results demonstrate an essential and universal role for HSP101 in protein export and provide strong evidence for PTEX function in protein translocation into the host cell. PTEX is a protein complex found in the PVM, where a translocon responsible for export into the host erythrocyte is expected 2 . The timing of its synthesis is appropriate for a putative translocon and it has a component (EXP2) with weak homology to bacterial hemolysins, which could form a membrane-spanning channel. PTEX contains two additional core components: the novel protein PTEX150 and HSP101, a member of the Clp/HSP100 family of AAA+ ATPases. To directly explore the role of HSP101/PTEX in protein export, we employed a conditional auto-inhibition approach by fusing a DHFR-based destabilization domain (DDD) to the endogenous HSP101 C-terminus (Fig. 1a). While this fusion strategy was originally developed to mediate conditional protein degradation via the proteasome 5,6 , we have found that the destabilized tag can conditionally interfere with chaperone function without protein degradation 7 . Following transfection, two independent clones were isolated which had undergone the intended recombination event (Extended Data Fig. 1). When the DDD-stabilizing small molecule trimethoprim (TMP) was removed from asynchronous HSP101DDD cultures, a complete block in growth was observed, with parasites accumulating as late ring-stage forms (Fig. 1b, c). While growth inhibition was TMP concentration dependent, this effect was not a result of HSP101DDD degradation (Extended Data Fig. 2a, b). To determine sensitivity to HSP101DDD destabilization across the asexual developmental cycle, TMP was removed at regular intervals in synchronized HSP101DDD parasites. Removal of TMP in early ring-stages resulted in growth arrest. In contrast, when TMP was removed at or after the beginning of the trophozoite stage (18–24 hours post-invasion), parasite development and re-invasion proceeded normally followed by arrest at the subsequent ring stage (Fig. 1d and Extended Data Fig. 2c, d). Parasite growth was rescued by adding back TMP as late as 48 hours after withdrawal, indicating that arrested ring forms remained viable. As protein export and PTEX expression also occur in parasite sexual stages, we further evaluated the ability of HSP101DDD parasites to form gametocytes required for malaria transmission 8,9 . In gametocytogenesis induction experiments, gametocyte development was severely inhibited when TMP was removed during stage I (Fig. 1e, f and Extended Data Fig. 3). Development was partially rescued by re-introducing TMP after 24 hours. Consistent with PTEX degradation in later gametocyte stages 9 , gametocyte formation was largely unaffected by later TMP removal (stage II onwards, 48 and 72 hrs), although a slight delay in development was observed. Collectively, these results demonstrate that HSP101 serves an essential function during asexual and sexual blood-stage development. To determine the impact of HSP101DDD inhibition on protein export, we examined a panel of proteins representative of the diversity of solubility states, targeting motifs, host cell destinations and expression timing observed among the P. falciparum repertoire of exported effectors. Most exported proteins contain a pentameric Plasmodium export element (PEXEL) 10,11 motif that is cleaved by the aspartic protease plasmepsin V in the parasite ER to license these proteins for export 12,13 . We first examined histidine-rich protein II (HRP2) 14 , a soluble PEXEL protein that is exported into the erythrocyte cytosol at the ring stage. To this end, TMP was removed from synchronized HSP101DDD schizonts, which were then allowed to develop for 18–24 hours into rings. Remarkably, export of HRP2 into the host cytosol was completely blocked in nearly all infected RBCs, with the protein accumulating in the parasitophorous vacuole (PV) surrounding the parasite (Fig. 2a, b). Similar results were obtained with the soluble, PEXEL-containing protein REX3 15 and the membrane-associated, PEXEL-containing protein KAHRP 16 (Extended Data Fig. 4a, b and g). Among PEXEL-containing proteins, ring-infected erythrocyte surface antigen (RESA) is uniquely stored in secretory organelles of invading merozoites and discharged into the PV along with PTEX immediately following invasion 3,4,17 . Robust block in RESA export was seen in recently invaded parasites (≤1 hour post invasion), indicating that in the absence of TMP, HSP101DDD function is impaired from the time of invasion (Extended Data Fig. 4c, g). PEXEL-negative exported proteins (PNEPs) are not cleaved by plasmepsin V 18 but appear to contain export signals in their N-termini 19 . As PTEX has thus far been implicated only in translocation of PEXEL proteins 2–4 , we examined the PNEP REX1, a protein associated with the cytosolic face of parasite-induced structures in the RBC cytosol called Maurer’s clefts, which serve as platforms for sorting of exported proteins 20 . We again observed a block in export with REX1 accumulating in a ring around the parasite where it colocalized with HSP101DDD (Fig. 2c and Extended Data Fig. 4g). A similar block was observed for SBP1 21 and REX2 15 , integral membrane PNEPs exported to the Maurer’s clefts (Extended Data Fig. 4d–g). Export was readily restored when TMP was added back to blocked cultures and restoration was largely insensitive to cycloheximide, indicating that previously synthesized HSP101DDD and exported proteins in the PV are sufficient to reactivate export (Extended Data Fig. 5). As later stages of parasite growth and development are not sensitive to TMP removal (Fig. 1d), we wondered if an export block could be induced after the ring stage. To this end, TMP was removed from synchronous HSP101DDD parasites at the late ring stage and export of trophozoite/schizont-specific proteins was assessed 12–24 hours later. We first examined MSRP6, a soluble PNEP peripherally associated with Maurer’s clefts 22 . Parasites lacking TMP displayed a dramatic block of MSRP6 export while SBP1, exported before the block, was found in the erythrocyte (Fig. 2d and Extended Data Fig. 4g). Similar results were seen for the PNEP PfEMP1, a variable surface antigen and key virulence determinant for P. falciparum malaria 23 (Fig. 2e and Extended Data Fig. 4g). Importantly, PfEMP1 trafficking to the RBC surface is dependent on several additional exported proteins 1 . Thus, while these results indicate delivery of PfEMP1 to the RBC surface is HSP101-dependent, we cannot exclude the possibility that this block is an indirect result of a failure to export other proteins required for PfEMP1 trafficking rather than a direct effect of HSP101 inactivation. Finally, we examined PFA660, a PEXEL-containing HSP40 found in parasite-induced membrane structures in the RBC called J-dots 24 . Again, washout of TMP efficiently blocked export as visualized by expression of a PFA660-GFP fusion protein in live parasites (Fig. 2f and Extended Data Fig. 4g). Together, these data show that HSP101 function is necessary for export but not parasite development in the trophozoite and schizont stages. Formation of a normal digestive vacuole containing hemozoin (Fig. 2d–f, DIC panels) indicates that hemoglobin uptake and degradation proceeded normally in these parasites, a process requiring proteases that are secreted into the PV before entering the digestive vacuole where they are activated 25 . Indeed, maturation of plasmepsin II was unaffected by TMP removal, indicating normal trafficking through the PV to the digestive vacuole (Extended Data Fig. 6). Additionally, HSP101 inactivation did not affect egress, a process mediated by secretion of the PfSUB1 protease into the PV 26 . Since PV processes unrelated to export proceed normally when HSP101 function is interrupted, HSP101 appears to serve a specific role in export. Furthermore, parasites expressing a PV-targeted GFPDDD fusion protein showed no defect in growth or export, indicating the HSP101DDD phenotype is not an indirect result of destabilizing the DDD in the PV compartment (Extended Data Fig. 7). Finally, we examined PfGECO, a PEXEL-containing HSP40 that is specifically expressed in gametocytes 27 . Again, export of PfGECO across the PVM was dramatically blocked in stage I gametocytes, showing an HSP101 requirement for protein export in sexual stages (Extended Data Fig. 4h). Parasite nutrient acquisition is supported by establishment of a Plasmodium surface anion channel (PSAC or new permeability pathways) at the late ring stage. This channel modifies the sensitivity of the infected RBC to osmotic lysis upon solute uptake. Following TMP removal prior to RBC invasion, arrested ring-stage HSP101DDD parasites failed to lyse in the presence of sorbitol (Fig. 3a). In contrast, sensitivity to osmotic lysis developed normally when TMP was removed at the end of the ring stage (Fig. 3b). CLAG3 proteins localize to the infected RBC periphery and are the only parasite proteins known to be involved in PSAC activation 28 . Interestingly, we observed no change in CLAG3 localization to the host periphery when export was blocked, suggesting that additional exported factors are required to activate PSAC (Fig. 3c). CLAG3 may be directly secreted into the RBC membrane during invasion similar to other rhoptry proteins 29 , or could traffic from the PV by an HSP101-independent export pathway. Nutrient uptake via PSAC is an essential process and block of its formation at the late ring stage may be the lethal event that follows HSP101 inhibition. We reasoned that export block might arise from the disordered DDD interfering with interactions between HSP101DDD and other PTEX components. To test this, HSP101DDD was immunoprecipitated (IP) from parasite lysates. Remarkably, while localization of HSP101DDD and EXP2 was unchanged +/− TMP (Extended Data Fig. 8a–c), their interaction decreased by more than 90% in the absence of TMP (Fig. 4a, f). A similar loss of HSP101-EXP2 interaction (70% decrease) was observed in reciprocal IP experiments (Fig. 4b, f). In contrast, association of HSP101DDD with RESA increased by more than two-fold (Fig. 4c, f and Extended Data Fig. 9a). To analyze PTEX150 interactions, we generated an HSP101DDD strain expressing a second copy of PTEX150 with a FLAG tag. The fusion protein localized to the PVM and distribution was unchanged +/− TMP (Extended Data Fig. 8d, e). Interaction between HSP101DDD and PTEX150 was substantially decreased in the absence of TMP (84% decrease, Fig. 4d, f) while interactions between PTEX150 and EXP2 were not affected (Fig. 4e, f and Extended Data Fig. 9b). We conclude that destabilization of the DHFR-domain fold results in HSP101DDD dissociation from the PTEX complex, producing a block in translocation of exported proteins out of the PV (Fig. 4g). Increased association with RESA under these conditions shows that interaction with exported substrates is not impaired. Interestingly, we also observed increased interaction between EXP2 and RESA in the absence of TMP (Fig. 4c, f and Extended Data Fig. 9a), suggesting that PTEX components can interact with substrates independent of HSP101 but that HSP101 is needed to drive export, likely through substrate unfolding and/or translocation (although we cannot exclude the possibility that association of RESA and HSP101DDD in the absence of TMP results from “stickiness” of the destabilized DDD). These data demonstrate that HSP101 is required to mediate protein export in P. falciparum. As HSP101 is a major component of PTEX, our results provide strong functional evidence for a PTEX role in protein translocation (Fig. 4g), though we cannot formally rule out the possibility that these proteins perform some other critical activity upstream of actual translocation across the PVM. The essential role of HSP101 in parasite survival and export of key virulence determinants provides impetus for pursuing PTEX as a drug target. Full Methods Cell Fractionation TMP was removed from synchronous 13F10 parasites at the late schizont stage and development was allowed to proceed +/− 10 μM TMP for 18 hours to the ring stage. Ring-stage parasites were enriched to >95% parasitemia by an initial tetanolysin treatment of 2.5 hemolytic units (HU) for 20 minutes as described 30 . Samples were then fractionated by sequential treatment with tetanolysin (1 HU for 20 min) and saponin (0.035% for 5 min) as described 19,30 . Parasite strains, culture and genetic modification Parasite culture and flow cytometry analysis was performed as described 7 . Cloning was performed using the In-Fusion system (Clontech) unless otherwise noted. For generation of HSP101DDD parasites, the DDD cassette was PCR amplified from plasmid pGDB 6 with primers 5′-GAGCAGCTGACCTAGGTACCCATACGATGTTCCAGATTACGCTTACCCATACGATGTTCCAGATTACGCTTACCCATACGATGTTCCAGATTACGCTATCAGTCTGATTGCGGCGTTAGCGGTAGATCACGTTATCGGC-3′ and 5′-TAACTCGACGCGGCCGTCATCGCCGCTCCAGAATCTCAAAGCAATAGCTGTGAGAG-3′ to incorporate a 5′ 3xHA tag. The amplicon was inserted into plasmid pPM2GT 25 between restriction sites AvrII/EagI. The BSD cassette was then PCR amplified from pGDB with primers 5′-ATGCCTGCAGGTCGATTGCATGCTTAGCTAATTCGCTTG-3′ and 5′-GAGAAATCTAGAGGTACCGAGCTCGATCTGCCGGTCTCC-3′ and inserted between restriction sites SalI/BglII. Finally, a portion of the 3′ genomic locus of HSP101 up to but not including the stop codon was PCR amplified from P. falciparum genomic DNA using primers 5′-ACGATTTTTTCTCGAGCGAAAACTTTTATGGTATTAATATAACAGATAAAGCTTTAGTAGCAGCAGC-3′ and 5′-CTGCACCTGGCCTAGGGGTCTTAGATAAGTTTATAACCAAGTTTTTAGCTTTACTATTATAATCAACAAATACATCC-3′ and inserted between restriction sites XhoI/AvrII. The resulting plasmid pHSP101-HDB was transfected into the 3D7-derived parental strain PM1KO 31 , which contains an hDHFR expression cassette conferring resistance to TMP. Selection, cycling and cloning were performed as described 7 . TMP was maintained in the medium throughout. Two clones from independent transfections were isolated and designated 13F10 and 14G11. Integration was detected by PCR using the primers 5′-CCTCCTTCAGTAGATATGACCG-3′ and 5′-CTAACGCCGCAATCAGACTG-3′ and confirmed by Southern blot using a probe to the 3′ end of the HSP101 gene, generated with the primers 5′-CGAAAACTTTTATGGTATTAATATAACAG-3′ and 5′-GGTCTTAGATAAGTTTATAACCAAG-3′. For gene expression, the episomal expression plasmid pyEOE 7 containing a yeast dihydroorotate dehydrogenase selection cassette was modified for transposase-mediated genomic integration by inserting a piggyBac element containing inverted terminal repeats amplified from plasmid pXL-BACII-DHFR 32 with primers 5′-GGAATTTCCTTATAAGATCTTAATACGACTCACTATAGGGCGAATTGGG-3′ and 5′-ATAATGGTTTCTTAGACGTCGATAAAAGTTTTGTTACTTTATAGAAGAAATTTTGAG-3′ between the restriction sites BglII/AatII, resulting in the plasmid pTyEOE. For expression of PFA660-GFP, the PFA660 coding sequence was PCR amplified from parasite cDNA with primers 5′-ACGATTTTTTCTCGAGATGGCAACCTTAAGGAAAAGC-3′ and 5′-CTGCACCTGGCCTAGGATAACTCTCTTTAAATATCTCTTTTATC-3′ and inserted into plasmid pTyEOE between restriction sites XhoI/AvrII, placing the gene under the control of the HSP86 promoter and generating a C-terminal GFP fusion. For expression of the PV-targeted GFP-DDD fusion protein, sequence encoding GFP fused to DDD with a C-terminal HA tag was amplified from plasmid pGDB with the forward primer 5′-ACCCCGGGATCTCGAGATGACAAGAAGATATTTAAAGTATTATATTTTTGTTACTTTATTGTTTTTTGTTCAAGTTATTAATAATGTATTGTGTGCTCCTAGGGCAGCAAGTAAAGGAGAAGAACTTTTCACTGGAG-3′ (encoding the HSP101 signal peptide, amino acids 1-27) and the reverse primer 5′-AACTCGACGCGGCCGTCAAGCGTAATCTGGAACATCGTATGGG-3′. This amplicon was inserted in plasmid pTyEOE between the restriction sites XhoI/EagI. For expression of the PTEX150-FLAG fusion protein, sequence encoding a 6xFLAG epitope was generated with the primer 5′-TTGATAGACCTAGGGACTACAAGGACGACGACGACAAGGATTATAAAGATGATGATGATAAAGATTATAAAGATGATGATGATAAAGATTATAAAGATGATGATGATAAAGATTATAAAGATGATGATGATAAAGATTATAAAGATGATGATGATAAATAACGGCCGCGTCGAGTT-3′ and its reverse complement. The primers were annealed and inserted between restriction sites AvrII/EagI in plasmid pTyEOE, resulting in plasmid pTyEOE-6xFLAG. The PTEX150 gene (which contains no introns) was then PCR amplified from parasite genomic DNA with its endogenous 5′ UTR up to but not including the stop codon using primers 5′-GATCGAGACGTCCTCTTTGTGGTCAAAATAAGTAAAATTTTATAAATTC-3′ and 5′-GTCACCTAGGATTGTCGTCCTCTTCTTCGTCC-3′ and inserted between restriction sites AatII/AvrII by standard ligation. pTyEOE-derived plasmids were co-transfected with plasmid pHTH for transient expression of the piggyBac transposase to mediate genomic integration 32 . Growth assays For asynchronous growth assays, parasites were subcultured 1:2 each day to avoid confounding effects from high parasite density in making growth rate comparisons between different samples. Parasitemia at each time points was back calculated based on the subculturing schedule. Data were fit to exponential growth equations using Prism (Graphpad Software, Inc.). In synchronous growth assays, daily media changes were performed without subculture. Parasitemia was determined by flow cytometry. Gametocyte induction Synchronous 13F10 parasites at 5–7% parasitemia were stressed by increasing hematocrit to 4% at the mid-trophozoite stage for ~12 hours. Hematocrit was returned to 2% at the schizont stage and TMP washout was performed at late schizont stage shortly before rupture and at 24 hr intervals thereafter. Asexual parasites were killed by treatment with 50 mM N-acetylglucosamine in the following cycle. Antibodies The following antibodies were used for IFA and Western blot: rabbit polyclonal anti-HA (Life Technologies) (IFA: 1:100); rabbit polyclonal anti-HA (Sigma) (WB 1:1000); rat anti-HA mAb clone 3F10 (Roche) (IFA: 1:100, WB: 1:1000); mouse anti-FLAG mAb clone M2 (Sigma) (IFA: 1:100, WB 1:500); mouse anti-EXP2 mAb clone 7.7 33 (IFA: 1:1000, WB: 1:1000); mouse anti-HRP2 mAb 2G12 14 (IFA 1:1000, WB: 1:500); mouse anti-RESA mAb clone 28/2 34 (IFA: 1:1000, WB: 1:1000); rabbit polyclonal anti-BiP 35 (WB: 1:1000); rabbit polyclonal anti-REX1 36 (IFA: 1:1000); mouse polyclonal anti-REX2 15 (IFA: 1:500); mouse polyclonal anti-REX3 15 (IFA: 1:500); rabbit polyclonal anti-SBP1 21 (IFA: 1:500); mouse polyclonal anti-MSRP6 22 (IFA: 1:500); mouse polyclonal anti-PfGECO 27 (IFA: 1:300); rabbit polyclonal anti-Pfs16 37 (IFA: 1:500); rabbit polyclonal anti-PfEMP1 ATS 38 (IFA: 1:200); mouse polyclonal anti-CLAG3 28 (IFA: 1:300); rabbit polyclonal anti-PM2 antibody 737 39 (WB 1:2000); rabbit polyclonal anti-KAHRP 16 (IFA: 1:500); rabbit polyclonal anti-SERP 40 (IFA: 1:500, WB: 1:1000). Light microscopy and image processing For indirect IFA detecting HRP2, REX3, PfGECO and Pfs16, cells were allowed to settle on coverslips coated with concanavalin A and fixed with a mixture of 4% paraformaldehyde and 0.0075% glutaraldehyde in PBS. For indirect IFA detecting HA and FLAG epitopes, EXP2, RESA, REX1, REX2, SBP1, MSRP6, PfEMP1, and KAHRP, thin smears were air dried and fixed with room temperature acetone. For indirect IFA detecting CLAG3 and SBP1, thin smears were air dried and fixed with ice-cold 90% acetone/10% methanol. Primary antibodies were detected by Alexa Fluor 488 or 594 secondary IgG antibodies (Life Technologies) used at 1:2000. Image stacks were collected at z-increments of 0.2 μm with an ORCA-ER CCD camera (Hamamatsu) and AxioVision software on an Axio Imager.M1 microscope (Zeiss) using a 100x oil immersion objective. Deconvolved images were generated using manufacturer specified point-spread functions and displayed as maximum intensity projections. Adjustments to brightness and contrast were made for display purposes. Live imaging was performed as previously described 41 . All IFA and live imaging experiments were repeated at least twice. Immuno electron microscopy For immunolocalization at the ultrastructural level, infected RBCs were fixed in 4% paraformaldehyde/0.05% glutaraldehyde (Polysciences Inc.) in 100mM PIPES/0.5mM MgCl2, pH 7.2 for 1 hour at 4°C. Samples were then embedded in 10% gelatin and infiltrated overnight with 2.3M sucrose/20% polyvinyl pyrrolidone in PIPES/MgCl2 at 4°C. Samples were trimmed, frozen in liquid nitrogen, and sectioned with a Leica Ultracut UCT cryo-ultramicrotome (Leica Microsystems Inc.). 50 nm sections were blocked with 5% fetal bovine serum/5% normal goat serum for 30 min and subsequently incubated with primary antibodies followed by secondary anti-rabbit conjugated to 18 nm colloidal gold (Jackson ImmunoResearch Laboratories, Inc.). Sections were washed in PIPES buffer followed by a water rinse, and stained with 0.3% uranyl acetate/2% methyl cellulose. Samples were viewed with a JEOL 1200EX transmission electron microscope (JEOL USA Inc.) equipped with an AMT 8 megapixel digital camera (Advanced Microscopy Techniques). All labeling experiments were conducted in parallel with controls omitting the primary antibody. These controls were consistently negative at the concentration of colloidal gold conjugated secondary antibodies used in these studies. Assessment of export in recently invaded parasites To monitor RESA export immediately after invasion, synchronous late schizonts were purified over Percoll and then mixed with fresh RBCs in pre-warmed media. Cultures were shaken at 37° C for one hour. After lysis of unruptured schizonts by sorbitol treatment, samples were fixed for IFA and stained with RESA antibodies as described above. Osmotic Lysis assays Osmotic lysis assays were performed as described by incubating cells in 290 mM D-sorbitol and measuring hemoglobin release by absorbance at 405 nm 42,43 . Samples were prepared as follows: TMP was removed or not from synchronous, late schizont-stage 13F10 parasites and development was allowed to proceed for 24 hours, resulting in 25% parasitemia cultures of early trophozoites (+TMP) or arrested late ring-forms (−TMP). Alternatively, TMP was removed or not from synchronous 13F10 parasites at the late ring-stage and development was allowed to proceed 24 hours before magnet purification to >95% parasitemia. In either case, the final concentration of cells in each sample was 1×108/ml. At the end of the assay, samples were lysed with saponin and data were normalized to these values (100% lysis). Immunoprecipitation For immunoprecipitation, TMP was removed or not from synchronous parasites at the early trophozoite-stage and parasites were allowed to develop for 24 hours to late schizonts and early rings. Parasites were then harvested with 0.035% saponin in PBS and lysed in 0.5% triton X-100 in PBS. Following sonication, lysates were cleared by centrifugation at 16,000 × g for 5 min and supernatants were nutated for 3 hours at 4°C with indicated antibodies and dynabeads coupled to protein G or protein A (or dynabeads alone as a negative control when necessary) (Life Technologies). Immune complexes were purified on magnets with extensive washing with 0.5% triton X-100 in PBS and then lysed in sample buffer. Western blots Western blot analysis was performed using an Odyssey infrared imaging system (LI-COR Biosciences). Primary antibodies were detected by IRDye 680 and 800-conjugated secondary antibodies (LI-COR Biosciences) used at 1:15,000. Quantitative measurements were made with Image Studio software (LI-COR Biosciences). Extended Data Extended Data Figure 1 Generation of HSP101DDD strains a, Schematic of strategy to generate HSP101DDD parasites. 3′F, 3′ flank for homologous recombination; BSD, blasticidin S deaminase; TMP, trimethoprim; HA, hemagglutinin tag; DDD, DHFR destabilization domain. b, Diagnostic PCR showing integration of the DDD fusion in the two independent clones 13F10 and 14G11. Primers shown as black arrows in a. Image is representative of two independent experiments. c, Southern blot showing integration of plasmid pHSP101-HDB occurred at the intended genomic locus. Expected NcoI digestion products and sizes are indicated in blue, red and orange in a. The 7.3 kb band in 13F10 and 14G11 indicates the presence of concatamers commonly observed in P. falciparum. Image is representative of one experiment. d, Western blot with anti-HA antibodies detects a 124 kD band in clones 13F10 and 14G11. Image is representative of two independent experiments. e, IFA of acetone-fixed 13F10 parasites showing colocalization of HSP101DDD and EXP2 at the PVM. Scale bar = 5μm. Images are representative of two independent experiments. Extended Data Figure 2 Characterization of HSP101DDD strains a, TMP dose-response for 13F10 and 14G11 parasites grown for 96 hours was measured. Error bars represent s. d. of three technical replicates. b, Western blot on lysates from asynchronous 13F10 parasites grown +/− TMP for 24 hours. BiP serves as a loading control. Blotting with anti-HA antibody shows no decrease in HSP101DDD protein levels relative to the BiP loading control. Images are representative of two independent experiments. c, Quantification of new ring parasitemia +/− TMP. In a parallel experiment to that shown in Fig 1d (but beginning with lower parasitemia cultures), new rings were counted in Giemsa-stained smears of synchronous 13F10 parasites where TMP was removed or not during the proceeding trophozoite stage. No significant difference in the resulting new ring parasitemia was observed +/− TMP, indicating no difference in re-invasion efficiency -TMP. Error bars represent s. d. of three technical replicates. Data are representative of two independent experiments. d, Giemsa-stained smears of synchronous 13F10 parasites grown +/− TMP from Fig. 1d. Reintroduction of TMP after 24 or 48 hours restored progression through the intraerythrocytic cycle. After 72 hours without TMP, most parasites appeared as dead, pyknotic forms. Images are representative of three independent experiments. Extended Data Figure 3 HSP101 function is critical for early stage gametocyte development Giemsa-stained smears from day nine post gametocyte induction as quantified in Fig. 1e. TMP was removed from parallel samples at 24-hour intervals beginning just prior to reinvasion (0 hrs) following gametocytogenesis induction. Data are representative of 4 independent experiments. While control (+TMP) gametocytemia varied between experiments, a similar effect on gametocyte formation was observed in each experiment following TMP removal. Extended Data Figure 4 Localization and quantification of exported proteins following HSP101DDD inactivation a–f, IFA and immuno electron microscopy (EM) of ring-stage 13F10 parasites +/− TMP. TMP was removed in late schizont stage and parasites were allowed to reinvade and grow 18–24 hours before fixation with paraformaldehyde (a) or acetone (b–d, f). Immuno EM fixation (e) is detailed in methods. A similar export block with accumulation of the exported protein at the parasite periphery was observed in each case when TMP was removed. The soluble PEXEL-containing protein REX3 (a) is normally exported into the host RBC cytosol. The PEXEL-containing KAHRP protein (b) is normally exported through Maurer’s clefts to knob structures at the cytoplasmic face of the infected RBC membrane. The PEXEL-containing protein RESA (c) is normally exported to the RBC periphery. SERP is a marker for the PV. SBP1 (d) is an integral membrane PNEP normally exported to the Maurer’s clefts. In the absence of TMP, blocked SBP1 colocalizes with EXP2. e, Immuno electron microscopy showing localization of SBP1 in ring-stage parasites +/− TMP. MC, Maurer’s cleft. Scale bars = 500nm. Images are representative of one experiment. REX2 (f) is an integral membrane PNEP normally exported to the Maurer’s clefts. g, Quantification of export block by IFA for exported proteins shown here and in Fig. 2c–f. Export was scored as complete (all signal in the host cell), partial (signal within the host cell but also within the PV) or no export (signal only seen within the PV and not in the host cell). Example images of each scoring scenario are given for REX1 in Extended Data Fig. 5a. In the case of PfEMP1, cells were scored as having PfEMP1 signal at the RBC periphery (complete export) or not (no export) due to the fact that some PfEMP1 signal is always seen within the PV under normal export conditions (see Fig. 2e, +TMP). Error bars represent s. d. of three technical replicates. Data are representative of at least two independent experiments. h, IFA of PfGECO in paraformaldehyde-fixed, stage I gametocytes 36 hours post invasion +/− TMP. Pfs16 is a gametocyte-specific PVM marker. All IFA scale bars = 5μm. All IFA images (a–d, f, h) are representative of two independent experiments. Extended Data Figure 5 Reactivation of export does not require new protein synthesis a–b, TMP was removed in late schizonts and parasites were allowed to reinvade and develop for 18 hours before TMP add back with or without 10 μg/ml cycloheximide (CHX). Parasites were acetone-fixed 24 hours later and processed for IFA. Export was scored as complete (no REX1 retained within the EXP2-labeled PVM), partial (REX1 in the host cell and retained within the PVM) or no export (no REX1 signal beyond the PVM). Error bars represent s. d. of three technical replicates. Data are representative of two independent experiments. Scale bar = 5μm. c, Metabolic labeling with [35S]methionine/cysteine, performed as previously described 44 , confirms that CHX treatment conditions inhibit new protein synthesis. Parasite proteins were TCA-precipitated and incorporated radioactivity was determined through scintillation counting. Error bars represent s. d. of three technical replicates. Extended Data Figure 6 Maturation of plasmepsin 2 is not impacted by inactivation of HSP101 Western blot showing normal maturation of plasmepsin 2 (PM2) in asynchronous parasites after 24 hours −TMP. BiP serves as a loading control. Maturation requires proPM2 trafficking through the PV prior to internalization to the digestive vacuole where maturation occurs 25 . Images are representative of two independent experiments. Extended Data Figure 7 DDD targeted to the PV independent of HSP101 does not interfere with parasite growth or export a, Schematic of the PV-targeted GFP-DDD fusion protein consisting of a signal peptide appended to a GFP-DDD fusion with a C-terminal HA epitope tag and expressed under the control of the HSP86 promoter. The predicted size of the fusion protein after signal peptide cleavage is 46 kD. Live imaging of GFP demonstrates PV targeting of the fusion protein. Images are representative of two independent experiments. b, Western blot showing GFPDDD is not degraded in the absence of TMP. Synchronized GFPDDD parasites were grown 72 hours with or without TMP before purification over Percoll and treatment with tetanolysin to release the RBC cytosol but not the PV contents. The GFPDDD fusion was detected with anti-HA antibodies. BiP serves as a loading control. Images represent one experiment. c, Growth analysis of asynchronous GFPDDD parasites shows no growth defect -TMP. Error bars represent s. d. of three technical replicates. Data are representative of two independent experiments. d, IFA showing no defect in export in GFPDDD parasites in the absence of TMP. No difference was observed in export of ring-specific (REX1) and the trophozoite-specific (MSRP6) exported proteins in the presence or absence of TMP. All scale bars = 5μm. Images represent one experiment. Extended Data Figure 8 Localization of PTEX components is unchanged under export blocking conditions a, Immuno EM showing localization of HSP101DDD in ring-stage parasites +/− TMP. TMP was removed or not from synchronous late schizonts and parasites were allowed to re-invade and develop for 18 hours prior to fixation. Scale bars = 500nm. Images represent one experiment. b, c, IFA of 13F10 parasites showing co-localization between HSP101DDD or EXP2 and the PV marker SERP +/− TMP, indicating localization of these PTEX components is not altered under export blocking conditions. TMP TMP treatment was performed as in a prior to fixation with acetone and processing for IFA. d, e, IFA of ring-stage 13F10 parasites expressing a PTEX150-FLAG fusion. The upper panel (d) shows co-localization with HSP101DDD. The lower two panels (e) show PTEX150 remains at the PVM during export block and partially colocalizes with blocked SBP1. TMP treatment and fixation as in b,c. All IFA scale bars = 5μm. All IFA images (b–e) are representative of two independent experiments. Extended Data Figure 9 Immunoprecipitation bead controls indicate target-specific interactions Replicate IP experiments to those shown in Fig. 4c, e with bead controls included. Purification of (a) RESA by HSP101DDD or EXP2 and purification of (b) EXP2 by PTEX150FLAG is specific to the target antibodies. IP experiments were performed in parallel by incubating equivalent portions of lysate supernatant input with beads alone or beads and the indicated IP antibodies. S, supernatant input. E, elution. Data are representative of two independent experiments.
