Deletion of the Plasmodium falciparum exported protein PTP7 leads to Maurer’s clefts vesiculation, host cell remodeling defects, and loss of surface presentation of EMP1

Presentation of the variant antigen, Plasmodium falciparum erythrocyte membrane protein 1 (EMP1), at knob-like protrusions on the surface of infected red blood cells, underpins the parasite’s pathogenicity. Here we describe a protein PF3D7_0301700 (PTP7), that functions at the nexus between the intermediate trafficking organelle, the Maurer’s cleft, and the infected red blood cell surface. Genetic disruption of PTP7 leads to accumulation of vesicles at the Maurer’s clefts, grossly aberrant knob morphology, and failure to deliver EMP1 to the red blood cell surface. We show that an expanded low complexity sequence in the C-terminal region of PTP7, identified only in the Laverania clade of Plasmodium, is critical for efficient virulence protein trafficking.


Introduction
Plasmodium falciparum causes more than 200 million malaria infections every year, killing more than 600,000 people [1]. Central to the ability of P. falciparum to maintain an infection and cause disease, is the invasion and remodelling of host red blood cells (RBCs). Maturation of the parasite inside the RBC is accompanied by striking changes in the surface topology of the infected RBC and a marked loss of cellular deformability [2,3]. One key modification is the assembly of~90 nm diameter structures, called knobs, at the infected RBC surface [4]. The knob structure acts as an elevated platform at the RBC surface for presentation of the major virulence protein, P. falciparum erythrocyte membrane protein-1 (EMP1), which mediates binding of infected RBCs to endothelial ligands [5][6][7][8].
The trafficking of EMP1 and other virulence determinants beyond the confines of the parasite is mediated by the Plasmodium translocon of exported proteins (PTEX) that is present in the parasitophorous vacuole membrane (PVM), with help from a second complex termed the exported protein interacting complex (EPIC) [9][10][11][12]. Once exported across the PVM, EMP1 is thought to be trafficked across the RBC cytoplasm as a soluble, chaperoned complex; and inserted into the membrane bilayer of the Maurer's clefts-an intermediate trafficking compartment in the RBC cytoplasm [12][13][14]. The mechanisms controlling EMP1 insertion into the Maurer's clefts and its subsequent forward trafficking and delivery to the RBC membrane remain unclear. Of interest, coated electron-dense vesicles (EDVs) and uncoated vesicle-like structures have been observed in the host cytoplasm and have been shown to contain EMP1 [15][16][17]. These EMP1-containing vesicles have been reported to associate with remodeled RBC actin filaments and membranous structures, called tethers, that decorate the Maurer's clefts in the later stages of intraerythrocytic development [15,16,18].
An interesting feature of exported Plasmodium proteins is that they are enriched in repetitive sequences with low sequence complexity [19,20]. For example, glutamine and asparagine repeats are present in about 30% of P. falciparum proteins [21,22]. Proteins with low complexity sequences have been shown to play critical roles in vesicle trafficking in other eukaryotes [23,24]. Thus, it is possible that low complexity regions in exported Plasmodium proteins help regulate protein trafficking and host cell remodeling.
We recently reported a method for enrichment of Maurer's clefts that enabled profiling of the protein composition and identification of protein networks within this organelle [25]. That work identified several key protein clusters that play an important role in EMP1 trafficking through the Maurer's clefts. Among the clusters was an uncharacterised protein, PF3D7_0301700, that interacts with the Maurer's cleft resident protein, PfEMP1 trafficking protein 6 (PTP6). PTP6 has been previously implicated in EMP1 trafficking and was shown to locate at the periphery of the Maurer's clefts [25,26].
Here we provide a detailed characterization of PF3D7_0301700, showing that it associates with the Maurer's clefts as well as chaperone-containing structures known as J-dots, and EMP1-containing electron dense vesicles (EDVs), by light and electron microscopy and immuno-precipitation. Upon deletion of PF3D7_0301700, vesicles accumulate at the Maurer's clefts; the knobs become larger and fewer in number; EMP1 delivery to the RBC membrane fails; and cytoadhesion to an endothelial cell ligand is ablated. Removing the low complexity C-terminal region of PF3D7_0301700 leads to an accumulation of vesicles at the Maurer's clefts and a defect in EMP1 trafficking. Taken together, the data show that PF3D7_0301700 plays an important role in vesiclemediated trafficking of EMP1 to the RBC membrane. We have given it the name, P. falciparum EMP1 trafficking protein-7 (PTP7), to reflect its role in EMP1 trafficking.

PTP7 is exported to the host RBC cytoplasm and Maurer's clefts
The PTP7 (PF3D7_0301700) sequence contains a predicted Plasmodium export element (PEXEL) (KSLAE), a recessed signal sequence (SS), an acidic N-terminal region, a central transmembrane domain and a basic C-terminal region containing 34 consecutive asparagine residues ( Fig 1A). The mature protein has a predicted molecular mass of 26,954 Da.
We used selection linked integration (S1 Fig) to generate transfectants endogenously expressing PTP7 tagged with a 2xFK506-binding protein (FKBP)-GFP-2xFKBP tag (3D7-PTP7-GFP sand ) [27]. Correct integration of the tag was verified by PCR (S1 Fig). Immunoblotting, using anti-FKBP antibodies, revealed a band with an apparent molecular mass of 130 kDa (S1 Fig). This molecular mass is larger than the predicted 108 kDa, which may be due to the large number of basic residues in PTP7, or alternatively, due to non-skipping of the 2TA peptide [27,28,29]. A trypsin cleavage assay confirmed that EMP1 is still trafficked to the RBC surface in this cell line, albeit with fainter full length and trypsinized EMP1 fragments, possibly due to a switch in EMP1 variant expressed (S1 Fig). Likewise, electron micrographs of the parasite infected RBC confirmed that the addition of the GFP sand tag had no effect on the Maurer's cleft ultrastructure (S1 Fig). Live cell microscopy of the transfectants expressing PTP7-GFP sand revealed a population of punctate structures in the host cell cytoplasm (Fig 1B). The location of PTP7-GFP sand was also assessed by immunofluorescence microscopy at four different windows from 16-30 hours post-invasion (hpi). Throughout the time-course, the PTP7-GFP sand fluorescence signal partly overlaps with that of the Maurer's cleft resident protein, ring exported protein 1 (REX1) (Figs 1C and 1D and S2), consistent with the Maurer's cleft location reported in a previous study [30].
We also tagged endogenous PTP7 with a 3xHA tag (S1 Fig) and made use of the improved resolution afforded by Airyscan super-resolution immunofluorescence microscopy. Both GFP sand and HA tagged PTP7 are revealed to associate with distinct puncta at the periphery of REX1-labeled Maurer's clefts, as well as with structures distinct from the Maurer's clefts (Figs 1E and S2). The equivalent location patterns of PTP7 tagged HA and GFP sand further provides assurance that the GFP sand tag does not alter the location of PTP7. To further investigate the sub-cellular location of PTP7-GFP sand at the ultrastructural level, trophozoite stage-infected RBCs were permeabilized with Equinatoxin II and probed with anti-FKBP antibodies. Gold particles are observed at the Maurer's clefts and at vesicle-like structures, located between the Maurer's clefts and the RBC membrane ( Fig 1F, white arrows).
To confirm the presence of PTP7 in different locations in the RBC cytoplasm, dual labelled immunofluorescence microscopy was performed using antibodies recognizing GFP or HA and antibodies for antigens located in different infected RBC compartments. The co-occurrence at each location was analysed at 4 timepoints across development (S2 Fig). Image analysis reveals~40% co-occurrence with the Maurer's cleft markers, MAHRP1 and REX1, as well as with EMP1; but also reveals~30% co-occurrence with the Maurer's cleft/EDV-associated protein PTP2 (S2 Fig). Interestingly, the J-dot protein, 0801, shows an increasing level of cooccurrence with PTP7 as the parasite matures (S2 Fig). Immunofluorescence microscopy of ΔPTP7 trophozoite-infected RBCs revealed no obvious change in the profiles of the Maurer's cleft markers REX1 and MAHRP1; the EDV marker PTP2 (Fig 2D and 2E); or the RBC membrane skeleton associated proteins PfEMP3, MESA and RESA (S4 Fig). In contrast, the knob protein, knob associated histidine rich protein (KAHRP), exhibited a more punctate profile when compared to wild type CS2, suggesting a disruption in knob distribution (Fig 2E). The same phenotype was observed in an independently generated CS2ΔPTP7 cell line (S4 Fig). To investigate this further, we performed scanning electron microscopy of the external surface of infected RBCs and the cytoplasmic face of
We employed electron tomography to investigate the organization of the spiral structure that underlies the knobs [39]. The spiral is evident, but fragmented, in ΔPTP7-infected RBCs (Figs 2I and S6), compared to more complete spirals in the CS2 samples. Taken together, these data show that both the distribution and the structure of knobs are severely compromised in the absence of PTP7. Thin section transmission electron microscopy (TEM) confirmed the deformed nature of the knob structures in ΔPTP7 parasites (S7 Fig).

Forward trafficking of EMP1 is PTP7-dependent
We examined whether genetic disruption of PTP7 affects EMP1 trafficking and surface expression. We assessed the ability of infected RBCs to adhere to the syncytiotrophoblast ligand, chondroitin sulphate-A (CSA), under physiologically relevant flow conditions. The parent CS2 line shows efficient binding to CSA (54 ± 7 infected RBCs/field of view), however, binding to CSA is completely ablated in the ΔPTP7-infected RBCs ( Fig 3A). To confirm the loss of surface displayed EMP1, we examined the ability of an antibody recognizing the ectodomain of var2CSA to label EMP1 in intact infected RBCs. A complete loss of surface labeling was observed in ΔPTP7-infected RBCs (Figs 3B and S4). Lastly, we assessed the surface accessibility of the EMP1 ectodomain to trypsin cleavage. After trypsinization of intact infected RBCs, characteristic cleavage fragments are observed, confirming the presence of surface displayed EMP1 (Figs 3C and S4). In contrast, no cleavage products are observed in the ΔPTP7-infected RBCs (Figs 3C and S4). The presence of full-length skeleton binding protein 1 (SBP1), a Maurer's cleft resident, confirms that the RBC membrane was not breached during trypsin treatment (Figs 3C and S4).
To determine where in the export pathway EMP1 was being trapped, we performed immunofluorescence microscopy using an antiserum recognizing the conserved ATS region of EMP1. This revealed Maurer's cleft labelling in both the CS2 and ΔPTP7-infected RBCs, suggesting that EMP1 can still be trafficked to the Maurer's clefts ( Fig 3D, left panels). To determine if the amounts of EMP1 being trafficked to the Maurer's clefts had changed we performed a quantitative analysis of the images. We looked at the co-occurrence of EMP1 and the Maurer's clefts marker, REX1. A significant increase in the number of EMP1-containing clefts (

Vesicles accumulate at the Maurer's clefts in PTP7-disrupted parasites
Ultrastructural analysis of the Maurer's clefts was performed in infected RBCs that had been permeabilized with Equinatoxin II to allow introduction of antibodies [16]. In wildtype parasites, the Maurer's clefts are observed as single slender cisternae with an electron-lucent lumen and an electron-dense coat. Upon immunolabelling with anti-REX1, gold particles are observed at the cleft periphery (Fig 4A top,   . Infected CS2 and CS2ΔPTP7 cell lines were passed through a channel slide coated with CSA under physiological flow conditions. The number of infected RBCs in 10 fields of view were recorded. Samples were run in technical triplicate and the experiment was repeated 3 times. Data displayed are mean ± SD of each technical repeat (CS2 n = 3; ΔPTP7 n = 3). (B) Flow cytometry analysis of infected RBCs labeled with antibodies recognizing the ectodomain of var2CSA, followed by secondary antibodies and tertiary antibodies conjugated to Alexa Fluor 488. Samples were run in triplicate and the experiment was repeated 3 times. The relative mean fluorescence of events was calculated as the (FITC geometric mean × number of events × 10 5 ) and averaged per cell line for each biological repeat. Data displayed are mean ± SD of each technical repeat (CS2 n = 3; ΔPTP7 n = 3, symbols indicate different biological repeats). (C) Trypsin cleavage assay. Membranes are probed with αATS and αSBP1 as a control. P: PBS mock treatment; T: Trypsin treated samples; asterisk: spectrin cross-reactivity band; arrows: trypsin cleaved EMP1 products. Loading control and experimental control, αSBP1, expected molecular weight of~50 kDa [28]. (D) Indirect immunofluorescence assays of cells fixed in acetone/methanol then probed with antibodies to the acidic terminal segment (ATS, green) and the Maurer's cleft protein REX1 (magenta) and stained for DNA using DAPI (blue). Example merged (left) and mask (right) images are shown, where EMP1 only (green), REX1 only (magenta) and both (white) objects are distinguished. White arrows indicate structures containing both REX1 and EMP1 (ATS). Scale bar, 2 μm.

PLOS PATHOGENS
Deletion of P. falciparum exported protein affects adhesin trafficking and host cell remodeling 4A bottom, zoom, white arrows). However, the clefts are decorated with numerous membrane-bound vesicle-like structures (Fig 4A bottom, zoom), with an average diameter of 44 ± 17 nm. The budding structures are morphologically similar to previously described vesicle-like structures [16] but are much more numerous than in wildtype parasites and remain in close proximity to the Maurer's clefts (Fig 4A-4C and S1-S4 Movies). The clouds of REX1 labeled vesicles around the clefts are consistent with our observation of fewer, larger clefts in ΔPTP7 (Fig 3G and 3H).
Immunolabelling with an antibody recognizing the ATS region of EMP1 [40] revealed gold particles on the body of the clefts in wildtype parasites (Fig 4B top, white arrows); and on both the body of the clefts and the associated vesicles in the ΔPTP7-infected RBCs (Fig 4B bottom, white and magenta arrows). These data suggest that these budding structures may be EMP1-trafficking vesicles that fail to separate from the clefts in the absence of PTP7.

The C-terminal asparagine repeats are needed for PTP7 function
The majority of the PTP7 sequence is highly conserved across Plasmodium species, however in P. falciparum and P. praefalciparum there has been an insertion of up to 34 asparagines near the C-terminal end of the protein (see multiple sequence alignment in S9 Fig). To test the functional significance of the C-terminal extension, we generated transfectants endogenously expressing one of three truncated PTP7 species fused to a GFP reporter (Figs 5A and S10). The first truncation has amino acids 265-317 deleted, leaving the transmembrane domain and four amino acids as a linker. The second cell line contains an internal deletion removing the asparagine repeats (Δ278-310), leaving the C-terminal poly basic motif 'KKSKKN'. The final cell line expresses a truncation that removes the poly basic motif (Δ300-317). These cell lines were validated by immunoblot, confirming the presence of PTP7-GFP fusion proteins of expected

PLOS PATHOGENS
Deletion of P. falciparum exported protein affects adhesin trafficking and host cell remodeling sizes, whereas no signal is observed in the parent CS2 cell line ( Fig 5B). Detection of REX1 and KAHRP products by indirect immunofluorescence microscopy confirmed the conservation of these subtelomeric genes and PCR validation confirmed genomic integration of the constructs (S10 Fig).
Scanning electron microscopy was performed on the truncations to determine if the knob morphology was affected. Truncation of the C-terminal 17 amino acids (PTP7 Δ300-317 ) had a minor effect on knob density and knob diameter (Fig 5C-5E). By contrast, deletion of the entire C-terminal region (PTP7 Δ265-317 ) was associated with fewer, larger knobs ( Fig 5C-5E), recapitulating the phenotype observed in the PTP7 knock-out line. Interestingly, internal deletion of the C-terminal asparagine repeats, while keeping the C-terminal 'KKSKKN' motif We next investigated if delivery of EMP1 to the RBC surface was affected in the truncations. Flow cytometry analysis revealed complete ablation of EMP1 at the surface of the PTP7 Δ265-317 infected RBCs (Fig 6A). The PTP7 Δ278-310 infected RBCs had a 63% reduction in surface EMP1 while the PTP7 Δ300-317 infected RBCs had a 32% reduction (Fig 6A). A trypsin cleavage assay was performed to confirm these results. Complete loss of EMP1 from the surface is observed in the PTP7 Δ265-317 infected RBCs, while faint trypsinized EMP1 bands are observed in both PTP7 Δ278-310 and PTP7 Δ300-317 parasites, consistent with the reduction of surfaceexposed EMP1 detected in the flow cytometry experiments (Fig 6B). In a similar trend, the number of GFP-positive puncta is significantly reduced in PTP7 Δ265-317 and PTP7 Δ278-310 parasites (S10 Fig). These puncta counts suggest that the length of the C-terminal region of PTP7 affects its distribution.
To determine the effect that truncation of regions in the C-terminal domain of PTP7 has on vesicle accumulation at the clefts, thin section TEM was performed. Truncation of the entire C-terminal region leads to an increase in vesicles at the Maurer's clefts as revealed in the ultrastructural analysis (Fig 6C). The Maurer's clefts appear less affected when at least some of

PLOS PATHOGENS
Deletion of P. falciparum exported protein affects adhesin trafficking and host cell remodeling the PTP7 C-terminal domain is retained (PTP7 Δ278-310 and PTP7 Δ300-317 ; Fig 6C). The vesicle diameters are not significantly affected by PTP7 mutations (Fig 6D); however, significantly more vesicles are observed within 100 nm of the PTP7 Δ265-317 clefts, while the PTP7 Δ278-310 and PTP7 Δ300-317 parasites exhibited a non-significant increase, relative to the full-length control ( Fig 6E). The accumulation of vesicles at the clefts is similar in the PTP7 Δ265-317 and ΔPTP7 lines.

Discussion
The canonical protein trafficking system that is used by most eukaryotic cells to transport proteins to the plasma membrane is not present in mature human RBCs [41]. Thus, intraerythrocytic P. falciparum faces the challenge of exporting integral membrane proteins, such as the virulence antigen, EMP1, to the RBC membrane; and inserting proteins into the lipid bilayer, in the absence of host cell machinery. To meet this challenge, the parasite establishes organelles called Maurer's clefts that are thought to function as intermediate trafficking compartments in the transport of EMP1 to the surface [13,14]. Moreover, several studies have described vesiclelike structures in the infected RBC cytoplasm [13,16,17,37,38,42,43]. However, the finding that EMP1 trafficking is GTP-independent [44] suggests that the molecular machinery that the parasite employs to traffic virulence proteins is divergent from that found in other cell types.
Here we characterized PTP7, a protein that was previously reported as a Maurer's cleft resident and prospective J-dot protein [30,45]. Our analysis of the location of PTP7 relative to a range of other exported P. falciparum proteins suggests that PTP7 localizes to different compartments in the RBC cytoplasm. Immunofluorescence labelling shows partial co-occurrence of PTP7 with the Maurer's cleft protein, REX1, throughout the intraerythrocytic cycle; with super-resolution imaging revealing that PTP7 is located at distinct puncta at the periphery of the Maurer's clefts. PTP7 also partly co-occurs with the EDV marker, PTP2, and the J-dot marker, 0801. Interestingly, an increase in the co-occurrence of PTP7 and 0801, is observed as the parasite matures.
To assess the function PTP7 plays in virulence protein trafficking, we generated two independent PTP7 knockout cell lines. Analysis of these cell lines showed that PTP7 is not essential for growth of parasites in culture, in agreement with a previous report [48]. However, in the absence of PTP7, EMP1 accumulates at the Maurer's clefts and is not presented at the surface of the infected RBC. Interestingly, upon deletion of PTP7, the Maurer's clefts became decorated with numerous vesicle-like structures that we speculate are in the process of budding from the cleft surface. Quantitative imaging revealed an accumulation of EMP1 at the Maurer's clefts in the ΔPTP7 transfectant-infected RBCs. Immuno-EM, using an antibody to EMP1, reveals labeling of both vesicles and the Maurer's clefts. Taken together, this suggests that the increase in vesicles at the clefts of the knockout infected RBCs may be linked to the defect in forward trafficking of EMP1 to the RBC membrane.
In addition to accumulation of vesicles at the clefts, the PTP7 knockout parasites also exhibit altered knob morphology, with fewer, larger knobs being assembled at the RBC surface. The main structural component of the knob complex, KAHRP, has been shown to traffic

PLOS PATHOGENS
Deletion of P. falciparum exported protein affects adhesin trafficking and host cell remodeling directly to the RBC membrane, without passage through the Maurer's clefts [46]. Our data for the PTP7 knockout suggests convergence of the KAHRP and EMP1 trafficking pathways during the final step, where EMP1 is loaded into the knobs at the RBC membrane. One possibility is that the vesicles also transport a component needed for correct assembly of knobs.
It is interesting to consider how PTP7, which has no ATP or GTP binding domain, could provide the driving force for vesicle fission. In this context, the unusual asparagine repeat sequence in the C-terminal region of PTP7 may be of functional significance. While P. falciparum asparagine repeat sequences are dispensable in some cases [49,50], recent studies suggest that membrane-associated proteins with low complexity regions can induce membrane curvature (reviewed in [24]). In a planar membrane, disordered regions of protein are constrained, thereby reducing the conformational entropy of the protein. As a membrane curves, this constraint is relaxed and the gain in conformational entropy enables curvature sensing [51], which can support vesicle budding [52,53].
To examine the role of the PTP7 asparagine repeats in the molecular architecture of the infected RBC and in the EMP1 trafficking process, we generated transfectants in which PTP7 was truncated. Complete removal of the C-terminal domain recapitulated the phenotype seen in ΔPTP7; that is, numerous vesicles accumulate at the Maurer's clefts, while the knobs showed aberrant morphology and EMP1 trafficking is ablated. The Maurer's cleft phenotype was less dramatic if the C-terminal polybasic motif was removed (PTP7 Δ300-317 ) or if the asparagine repeats were removed but the C-terminal polybasic motif was maintained (PTP7 Δ278-310 ). In all truncation mutants, the trafficking of EMP1 to the surface was decreased, suggesting that the length of the C-terminal region is a factor in the efficiency of EMP1 trafficking.
Conservation of PTP7 in the Laverania subgenus may be linked to the expansion of virulence protein trafficking and/or host cell remodeling machinery in these species. An analysis of the currently available Plasmodium genome assemblies showed that the N-terminal domain of PTP7 is well conserved in P. praefalicparum, P. reichenowi, P. gaboni, P. billcollinsi, P. alderi, and P. blackloci orthologues, but the C-terminal asparagine repeats are found only in P. falciparum and P. praefalciparum [54,55] (S9 Fig). We cannot currently explain the evolution of the C-terminus in the light of what we know about the evolution of var genes in the Laverania, but it may suggest that the expansion of the PTP7 sequence preceded the relatively recent speciation of P. falciparum and P. praefalicparum [56]. One hypothesis is that the expanded C-terminus of PTP7 interacts with the ATS domain of EMP1, which is conserved in P. falciparum and P. praefalicparum [56], a suggestion that is supported by our truncation data showing that this region is required for EMP1 trafficking; however, additional experiments would be required to test whether the interaction of PTP7 with ATS is enhanced in the expanded PTP7 sequence.
Taken together, our data provide strong evidence that PTP7 and/or its interacting proteins are required for the recruitment or formation of EMP1-containing vesicles at the Maurer's clefts and their subsequent transfer to the host RBC membrane. Our data shows that deletion of the C-terminus of PTP7 coincides with an accumulation of vesicles at the Maurer's clefts; however, the molecular basis for this phenomenon remains unclear. One possibility is that the low complexity asparagine repeats increase the hydrodynamic radius of the C-terminal domain of PTP7 [51,52]. This increased hydrodynamic radius, in turn, may drive a preferential location at highly curved membrane structures, like the edges of Maurer's clefts and vesicles, possibly crowding these microenvironments and/or recruiting other proteins to induce vesicle fission. However, this needs to be formally tested and alternative explanations, such as a role of the C-terminus in facilitating protein-protein interactions that are required for vesicle budding are equally likely.
The interaction of PTP7 with exported proteins at several locations within the RBC cytoplasm suggest that the routes of exported protein trafficking may be more integrated and

PLOS PATHOGENS
Deletion of P. falciparum exported protein affects adhesin trafficking and host cell remodeling interdependent than previously thought. This raises the alternative possibility that PTP7 functions indirectly, or in combination with other exported proteins, to facilitate vesiculation and EMP1 trafficking. Indeed, further characterization of the PTP7 interacting proteins identified in our co-immunoprecipitation experiments may reveal other exported proteins responsible for these phenotypes. An increased understanding of the processes for trafficking virulence proteins could lead to new therapies to tackle malaria pathogenesis.

Ethics statement
Red blood cells and serum were acquired from the Australian Red Cross Lifeblood blood service. All blood products were anonymous and individual donors could not be identified. This work was approved with the written consent of the University of Melbourne Human Research Ethics Committee (approval number 1750526.3).

Vector construction and generation of transgenic parasites
For endogenous 3' tagging, the 3' 710 bp of the ptp7 was amplified using SLI-sand-PTP7_fw and SLI-sand-PTP7_rv primers (S1 Table) and directionally cloned into the NotI/AvrII restriction sites in the pSLI-sandwich plasmid (2xFKBP-GFP-2xFKBP) [27]. The PTP7-HA plasmid was provided by P. Gilson [45]. Ring-stage parasites were transfected with 50-100 μg of plasmid as previously described [61]. Briefly, precipitated plasmid was resuspended in sterile TE buffer and Cytomix [61]. 5-10% ring-stage infected RBCs were resuspended in the DNA mix and transferred to a 2 mm electroporation cuvette (BTX). Cells were electroporated at 310 V, 950 μF, and 1 resistance (Gene Pulser Xcell Electroporation System, Bio-Rad), then washed in warm RPMI media and transferred to culturing dishes. After neomycin selection, correct integration of the plasmid was verified by PCR using the primers as described in S1 Fig and S1 Table. To generate the plasmid for CRISPR/Cas9 mediated disruption of ptp7, 5' and 3' homology regions (HR;~500 bp) were PCR-amplified from genomic DNA using the primers PTP7-HR1_fw, PTP7-HR1_rv for the HR1 and PTP7-HR2_fw, PTP7-HR2_rv for HR2 (S1 Table). The HR1 fragment was cloned into the AvrII/NcoI sites and the HR2 into the SpeI/ SacII sites of the pUF-TK plasmid [62]. The pUF-TK vector was linearized by digestion with

PLOS PATHOGENS
Deletion of P. falciparum exported protein affects adhesin trafficking and host cell remodeling AvrII and used as the repair template. CRISPR/Cas9-mediated double-stranded breaks were guided by a single guide RNA and Cas9. The Cas9 target 'ggttccaacacagtcacacg' was selected using CHOPCHOP and PTP7-sgRNA_top and PTP7-sgRNA_bottom oligonucleotides were annealed and cloned into the BsrgI site of pAIO using Infusion cloning (Takara Bio) [62]. Both linearized repair template and guide plasmids were transfected simultaneously into CS2 parasites. Gene disruption was confirmed by PCR using primers yDHODH_ScrF and yDHODH_ScrR, and primers in the native 5' and 3' UTRs of the ptp7 locus, PTP7-KO_ScrF and PTP7-KO_ScrR (Fig 2 and S1 Table).

Protein sample preparation, gel electrophoresis and immunoblotting
Whole parasite samples were prepared by lysis with saponin (Sigma). Cells were harvested and resuspended in 10 volumes of ice cold 0.03% saponin in PBS and incubated on ice for 10 min. Lysates were centrifuged at 500 × g for 10 min at 4˚C, the pellet was retained, and the supernatant discarded. The pellet was washed 4 times in PBS containing cOmplete EDTA-free Protease Inhibitor Cocktail (Roche). Protein samples were mixed with 4X LDS buffer containing DTT (NuPAGE Thermofisher) and heated at 70˚C for 10 min prior to loading. Samples were separated by SDS-PAGE on either 4-12% Bis-Tris or 3-8% Tris-Acetate gradient gels (NuPAGE) run in MES, MOPS, or Tris-Acetate SDS running buffers (NuPAGE) following the manufacturers recommendations.

PLOS PATHOGENS
Deletion of P. falciparum exported protein affects adhesin trafficking and host cell remodeling

Immunoprecipitation and mass spectrometric analysis
Mid-trophozoite stage-infected RBCs were enriched by Percoll or gelatin floatation and washed twice in RPMI media. The cell pellet was solubilized on ice for 30 min with 10 volumes of immunoprecipitation buffer (1% Triton X-100 in 50 mM Tris-HCl, 150 mM NaCl, 2 mM EDTA and cOmplete EDTA-free Protease Inhibitor Cocktail (Roche)). The sample was mixed every 10 min. The samples were centrifuged at 16 000 × g for 10 min at 4˚C and the supernatants were transferred to new tubes. The supernatant was again centrifuged to pellet any remaining insoluble material. The lysates were pre-incubated with Protein-A Sepharose beads (Novagen) and allowed to mix for 30 min at 4˚C. The lysate was again centrifuged, and the supernatant transferred into a microcentrifuge tube containing GFP-Trap beads and mixed at 4˚C for 4 h or overnight. Beads were washed 5 times in immunoprecipitation buffer and transferred to fresh tubes during the last wash. Samples were then either prepared for immunoblotting or mass spectrometry. In preparation for SDS PAGE analysis, the beads were resuspended in 4X LDS buffer containing DTT (NuPAGE Thermofisher) and warmed to 70˚C for 10 min. For mass spectrometry the beads were washed twice in 1 mM Tris-HCL pH 7.4. Formic acid (0.04% final) then 2,2,2-Trifluoroethanol (7% final) were added and the samples and incubated at 50˚C for 5 min. The sample was centrifuged, and the supernatant transferred to a fresh tube. Tetraethylammonium bromide (50 μM final) and Tris(2-carboxyethyl)phosphine hydrochloride (50 μM final) were added to the sample and incubated for 10 min at 70˚C. Trypsin (0.25 μg) was added and the samples were incubated overnight at 37˚C. Samples were subjected to tandem mass spectrometry with electrospray ionization (ESI) LC-MS/MS on a Q-Exactive mass spectrometer and analyzed as previously described [25]. Identified proteins were included if � 2 unique peptide sequences were detected and were either absent or 3-fold enriched over the control.

EMP1 trypsin cleavage assay
Cultures containing at least 5% mid-trophozoite stage-infected RBCs were washed in PBS and divided into 2 equal fractions. Cells were harvested at 300 × g for 5 min and resuspended in 10 volumes of warm PBS (P) or PBS containing 1 mg/ml TPCK-Treated Trypsin (Sigma) (T). Samples were incubated at 37˚C for 1 h, inverting occasionally to prevent sedimentation. Trypsin inhibitor from Glycine max (soybean) (Sigma) was added to each sample to 5 mg/ml final concentration and incubated at room temperature for 15 min. Cells were harvested and the supernatant discarded. The pellet fraction was solubilized in 10 volumes of ice cold 1% Tri-tonX-100 (Sigma) in PBS containing cOmplete EDTA-free Protease Inhibitor Cocktail (Roche) and incubated on ice for 30 min. All subsequent steps contained 1x cOmplete EDTAfree Protease Inhibitor Cocktail (Roche). Samples were centrifuged at 16 000 × g for 10 min at 4˚C, then the pellets were washed three times in the ice cold TritonX-100 solution. The pellet was solubilized in 20 volumes of 2% SDS/PBS and mixed at room temperature for 30 min. The samples were centrifuged at 16 000 × g for 10 min, the supernatant transferred to a new tube and centrifuged again. The resulting supernatant was transferred to a new tube and prepared for immunoblotting.

Antibody generation
Polyclonal antibodies against PTP7 were generated by Genscript. Briefly, the mature N-terminal domain of PTP7 (89-317 aa) was expressed in E. coli as a His tagged recombinant protein, which was used to immunize rabbits. Total IgG was purified from the rabbit serum using protein A.

PLOS PATHOGENS
Deletion of P. falciparum exported protein affects adhesin trafficking and host cell remodeling

Live cell and immunofluorescence microscopy
For live-cell fluorescence microscopy,~2 μl of P. falciparum culture was dispensed on a slide and flattened with a coverslip for immediate imaging. For indirect immunofluorescence assays, cells were either fixed in ice-cold acetone/methanol or formaldehyde/glutaraldehyde. Thin blood smears were fixed with ice-cold acetone/methanol at a ratio of either 1:1 or 9:1 for 10 min and allowed to air dry. For formaldehyde-glutaraldehyde fixation, coverslips were coated with 0.05 mg/ml Phaseolus vulgaris erythroagglutinin (PHA-E) lectin (Sigma) and incubated in a humid environment for 15 min at 37˚C. All subsequent incubations were conducted in a humid environment. Infected RBCs were washed in PBS, resuspended at 2% hematocrit and overlaid on the PHA-E coated slide. Cells were allowed to adhere for 15 min, then washed with PBS until a monolayer of cells was obtained. The slides were fixed in 4% formaldehyde 0.0065% glutaraldehyde in PBS for 10-20 min. The slides were washed 3 times in PBS and permeabilized with 0.1% (v/v) TritonX-100 for 10 min. The slides were washed prior to the addition of the primary antibodies diluted in PBS containing 3% (w/v) bovine serum albumin. The antibodies were incubated on the slide for 1 h at room temperature. Primary antibodies used in this study include mouse αGFP were added to the wells for 1 h at room temperature and then the wells were washed 3 times with PBS. Each well was incubated with 2 μg/ml 4',6-diamidino-2-phenylindole (DAPI) for 15 min, then washed and mounted with p-phenylenediamine (PPD) antifade in 90% glycerol prior to sealing. Samples were imaged on a DeltaVision Elite restorative widefield deconvolution imaging system (GE Healthcare) as detailed in [25]. Briefly, samples were excited with solid state illumination (Insight SSI; Lumencor). The following filter sets (excitation [Ex] and emission [Em] wavelengths) were used: DAPI (Ex390/18 nm, Em435/48 nm), fluorescein isothiocyanate (FITC) (Ex475/28, Em523/26 nm), tetramethyl rhodamine isocyanate (TRITC) (Ex542/27, Em594/45 nm), and Cy5 (Ex632/22, Em676/34 nm). A 100X UPLS Apo (Olympus, 1.4 NA) oil immersion lens objective was used for imaging. Super-resolution confocal microscopy was performed on the LSM880 with Airyscan (Zeiss). Samples were imaged using a 63x Plan-Apochromat (Zeiss, 1.4 NA) oil immersion lens. The following Elyra laser emission and band pass filter sets were used: GFP (Ex488, Em bandpass 495-550 + lowpass 570 nm), RFP (Ex561, Em bandpass 420-480 + bandpass 495-620 nm), Cy5 (Ex642, Em bandpass 570-620 + lowpass 645 nm). Images were processed using ImageJ software [71] and deconvolved with Huygens image analysis software (Professional version 19.04, Scientific Volume Imaging) using the CMLE algorithm with 50-100 iterations and 12-15 signal to noise ratio thresholds. Image segmentation and quantification was performed using custom macros which are available from the open-access Zenodo repository (https://zenodo.org/record/6747921). Where object segmentation was required, image/object validation was performed blinded.

Scanning electron microscopy
Whole infected RBCs were fixed and imaged as previously described [25]. Scanning electron microscopy (SEM) imaging of the cytoplasmic surface of the host RBC membranes was

PLOS PATHOGENS
Deletion of P. falciparum exported protein affects adhesin trafficking and host cell remodeling performed on sheared infected RBC membranes. The membranes were prepared as previously described [25]. Briefly, glass coverslips were cleaned with acetone and 50% methanol and functionalized with 3-aminopropyl triethoxysilane (APTES), bis-sulfosuccimidyl suberate, and erythroagglutinating phytohemagglutinin (PHAE). Mature infected-RBCs were immobilized on the functionalized glass slides and sheared by forceful application of a hypotonic buffer (5 mM Na 2 HPO 4 /NaH 2 PO 4 , 10 mM NaCl, pH 8). For CS2 and ΔPTP7 cells, coverslips were coated with gold at 25 mA for 40 s and 75 s using a Dynavac sputter coating instrument to thicknesses of~0.2 nm and~0.4 nm for the sheared and whole cells, respectively. Otherwise, whole cells were gold-coated with a Safematic CCU-010 sputter coater to a thickness of 5 nm. Images were acquired with the T1 (A+B, for CS2 and ΔPTP7 whole cells) or the Everhart-Thornley detector in 'Optiplan' mode (for whole cells and sheared membranes) of an FEI Teneo SEM using a working distance of 5 mm, a beam current level of 50 pA, and 2 kV accelerating voltage.

Immunoelectron tomography
Infected RBCs (20-32 hpi) were magnet purified, washed in PBS, and fixed in 10 pellet volumes of 2% (v/v) paraformaldehyde (PFA)/PBS for 20 min at room temperature. Cells were washed and then permeabilized in 10 pellet volumes of PBS with 1 hemolytic unit (HU) of Equinatoxin II for 6 min then washed again. Cells were then lightly fixed in 2% PFA/PBS for 5 min, washed, and incubated for 1 h with 3% BSA/PBS. The cells were incubated in one volume of the primary antibody (1:10) or no-primary control (3% BSA/PBS) for 2 h, washed, and then incubated with one volume of the gold secondary antibody for 1 h (1:15; Aurion protein A EM-grade 6-nm-diameter gold; catalog no. JA806-111). Cells were washed in 3% (w/v) BSA/ PBS and then in PBS to remove the BSA.
Cell pellets were resuspended and fixed in 2.5% glutaraldehyde at 4˚C for long term storage prior to embedding in agarose, fixed in 2% osmium tetroxide (w/v) then resin embedded, sliced, and imaged as described previously [25]. Alternatively, stored cells were post-fixed in 1% osmium tetroxide (w/v) and 1.5% potassium ferrocyanide (w/v) in 0.1 M SPB (Sorensen's Phosphate Buffer, Na 2 HPO 4 /NaH 2 PO 4 , pH 7.4) for 30 min at room temperature in darkness. Cells were washed with 0.1 M SPB and incubated in 1% tannic acid (w/v) in 0.1 M SPB for 20 min. Cells were washed in double-distilled H 2 O and dehydrated via sequential 5 min incubations in 30%, 50%, 70%, 90%, (2x) absolute ethanol, and (2x) absolute acetone for 10 min each. Samples were progressively infiltrated with Procure 812 Epon-substitute resin and polymerised at 60˚C for 48 hours. Ultrathin (70 nm) and semi-thin (300 nm) sections were generated using a Leica EM Ultracut 7 ultramicrotome (Leica, Heerbrugg, Switzerland) and sections were post-stained using 2% uranyl acetate and lead citrate respectively. Imaging and electron tomography were performed on an FEI Tecnai F30 electron microscope (FEI Company, Hillsboro, OR) at an accelerating voltage of 300 kV. The tilt series were acquired for every 2˚in the range between -70˚and 70˚. Virtual sections were reconstructed from the raw tilt series in IMOD using a weighted back-projection algorithm [72]. Model contours were sculpted and visualized using IMOD and ImageJ.

Infected RBC binding assay under flow conditions
Cytoadherence of infected RBCs was performed as described previously [25]. Briefly, Ibidi μ-Slide 0.2 channel slides were incubated with 100 μl chondroitin sulfate A (100 μg/ml; Sigma) in 1% BSA/PBS overnight at 37˚C. Channel slides were blocked with 1% BSA/PBS for 1 h at room temperature prior to gentle washes with 37˚C bicarbonate-free RPMI 1640 (Invitrogen). Mature infected cultures (3% parasitemia and 1% hematocrit) were harvested and resuspended

PLOS PATHOGENS
Deletion of P. falciparum exported protein affects adhesin trafficking and host cell remodeling in warm bicarbonate-free RPMI 1640. Samples were pulled through the channel at 100 μl/min using a syringe pump (Harvard Apparatus) for 10 min at 37˚C to allow cytoadherence, then washed for 10 min to remove unbound cells. Adhered cells were counted at 10 points along the axis of the channel. The microscopy was performed on a DeltaVision Elite widefield imaging system (GE Healthcare).

VAR2CSA ectodomain labeling and analysis via flow cytometry
Mid-trophozoite stage cultures were diluted to 3-5% parasitemia and 0.4% hematocrit and loaded into a 96-well plate in duplicate per condition (primary and no-primary controls). Cells were harvested at 528 × g for 90 sec and washed once in 1% BSA/PBS. Cells were incubated with 10 μl of rabbit polyclonal anti-VAR2CSA antibody for 30 min at 37˚C (1:100, R1945 [40]) or 1% BSA/PBS as a control. Following incubation, the cells were washed with 1% BSA/PBS, the supernatant was discarded, and the pellets were resuspended in 10 μl of mouse anti-rabbit IgG (Sigma RG-96; 1:100) and incubated for 30 min at 37˚C. The samples were again washed in 1% BSA/PBS before addition of goat anti-mouse antibody conjugated to a fluorophore, which was incubated for 30 min at 37˚C. For CS2/ΔPTP7 experiments, tertiary antibodies were conjugated to Alexa Fluor 488. For the GFP expressing PTP7 truncated cell lines, the antibodies were conjugated to Alexa Fluor 647. Where the tertiary antibody was conjugated to Alexa Fluor 488, cells were washed and stained with the nucleic acid stain SYTO-61 as described previously [73]. Where the tertiary antibody was conjugated to Alexa Fluor 647, cells were washed and incubated with the dsDNA stain Hoechst 33342 (1:2000) for 30 min at 37˚C. Following dsDNA Hoechst labeling, the cells were washed 2 times in 1% (w/v) BSA/PBS then once in PBS prior to analysis. Flow cytometry was performed on a BD FACSCanto II with an integrated high throughput sampler. The following filter sets were used for fluorophore detection: Alexa Fluor 647 and SYTO-61 (APC, 660/20 nm), Alexa Fluor 488 (FITC, 530/30 nm) and Hoechst 33342 (Pacific Blue, 450/50 nm). A total of 50,000 events were collected and doublet discriminated. Infected cells were gated on SYTO-61/Hoechst positivity. The mean fluorescence intensities of Alexa Fluor 488/Alexa Fluor 647 fluorescence (i.e., EMP1 labeling) in these populations was calculated. Non-specific events in the secondary-only antibody controls served as a guide for drawing of the FITC+/APC+ boundary.

Statistical analysis
Statistical tests and visualization were performed using GraphPad Prism 9.

PLOS PATHOGENS
Deletion of P. falciparum exported protein affects adhesin trafficking and host cell remodeling of surface trypsinized infected RBCs shown. 3D7: parent line. High molecular weight band is full length EMP1 recognized by αATS. Bands annotated with an asterisk indicate spectrin cross-reactivity expected at approximately 225 kDa for polypeptides [75] and 65 kDa for spectrin degradation products [76]. Arrows indicate cleaved EMP1 species where intracellular ATS regions were protected during trypsin incubation. Loading control, αSBP1, is also an experimental control, as breaching of the infected red blood cell membrane during trypsinization would cleave the Maurer's cleft protein SBP1. SBP1 expected molecular weight of~50 kDa [28].

PLOS PATHOGENS
Deletion of P. falciparum exported protein affects adhesin trafficking and host cell remodeling conjugated to Alexa Fluor 647. Samples were run in technical duplicates. Data displayed are mean fluorescence values ± SD for each biological repeat (n = 4 per cell line). (H) Trypsin cleavage assay of truncation lines. Membranes were probed with αATS and the loading/experimental control αSBP1. P: PBS mock treatment; T: Trypsin treated samples; asterisk: spectrin cross-reactivity band; arrows: EMP1 cleavage products. Loading control and experimental control, SBP1, expected molecular weight is~50 kDa. (TIF) S5 Fig. CS2 and CS2ΔPTP7 internal scanning EM images. Early to mid-trophozoite cultures were adhered to lectin functionalized slides and sheared off with hypotonic buffer to reveal the cytoplasmic face of the infected red blood cell footprint which was then fixed, dehydrated, coated with gold, and imaged using scanning electron microscopy. Knobs are indicated by inverted discs. Data displayed are mean ± SD (CS2 n = 14; ΔPTP7 n = 12). P-values deter-  (Fig 4). (A) IFA channels probed with αATS and αREX1 were Huygens deconvolved to resolve cytoplasmic αATS puncta and clefts indicated by αREX1 signal. A FIJI Macro script counts the αATS only, αREX1 only, and both particles to generate: a proof, including the merge image (color assignment as in (Fig 4A)) and results image with EMP1 only (green), REX1 only (magenta) and both (white) depicted. (B) Ratio of EMP1-only particles to clefts. Data displayed are mean ± SD (CS2 n = 44; CS2ΔPTP7 n = 63). (C-H) Quantitation of αATS signal distribution and cleft features. (C) Additional examples of the cleft and cytoplasm masks used to quantify the fluorescence intensities of these compartments. Bright field (BF) and DNA stain DAPI merged for reference. The 'Cleft mask' depicts where αATS signal was classified as 'cleft'. The 'Cytoplasm' mask illustrates i) the internal object used to determine αATS cytoplasmic signal and ii) the external space where αATS signal was measured as a background mean gray value control. The merge illustrates the merged IFA channels (color assignment as in (Fig 4A)). Mean gray values in the cleft and cytoplasm compartments were background subtracted. Data analysis performed is indicated in the y-axis.