Extracellular vesicles participate in proteostasis and heat shock adaptation in Plasmodium falciparum Authors: Yunuen Avalos-Padilla, Lucía Román-Álamo, Inés Bouzón-Arnáiz, Elsa M. Arce, Diego Muñoz-Torrero, Xavier Fernàndez-Busquets Published: 2026 ABSTRACT ABSTRACTHeat shock is a hallmark of clinical malaria, where Plasmodium falciparum parasites are exposed to recurrent febrile episodes exceeding 40 °C, which lead to acute proteotoxic stress. Parasite survival under these conditions relies on efficient proteostasis mechanisms and molecular chaperones, yet how stress resilience is coordinated beyond chaperone responses remains poorly understood.Here, we identify a stress-associated role for extracellular vesicles (EVs) in parasite heat shock adaptation linked to vesicular trafficking mediated by PfVps60, an Endosomal Sorting Complex Required for Transport (ESCRT) protein. Using a PfVps60 knockout (PfVps60KO) line, we show that disruption of ESCRT-dependent vesicular trafficking affects EV cargo composition during thermal stress. Proteomic profiling revealed that 44.8% of EV-associated proteins from P. falciparum 3D7 overlapped with a previously defined set of aggregation-prone proteins. Loss of PfVps60 impaired EV-mediated export of the chaperones PfHsp70-x and PfHsp110, altered aggregation dynamics and induced the redistribution of protein aggregates near the parasitophorous vacuole, reduced induction of the cytosolic chaperone PfHsp70-1, and resulted in early loss of parasite viability following heat shock. Supplementation of PfVps60KO parasites with EVs derived from heat-stressed 3D7 parasites partially rescued heat shock tolerance in a dose-dependent manner. EVs released shortly after thermal stress were enriched in aggregation-prone proteins and associated with neighbouring uninfected erythrocytes, suggesting EV-mediated intercellular communication during febrile episodes. Together, these findings support a role for EV-associated cargo as a previously unexplored component of P. falciparum proteostasis during heat shock adaptation, identifying stress-induced EVs as a potential parasite vulnerability for malaria intervention. Introduction In 2024, malaria remained a major global health burden, with an estimated 270 million cases and 610,000 deaths [1]. The disease is caused by protozoan parasites of the genus Plasmodium, among which Plasmodium falciparum is the most virulent and the most widespread species in sub-Saharan Africa [2]. The parasite’s life cycle alternates between human and Anopheles mosquito hosts [3], exposing it to abrupt thermal fluctuations, ranging from around 25 °C in the insect to 37 °C in the human, and up to 41 °C during febrile episodes [4], a hallmark of clinical malaria which also influences parasite biology. It has been demonstrated that elevated temperatures can enhance cytoadherence of early intraerythrocytic stages [5,6], facilitate synchronization of parasite populations in vivo [7], induce stiffening of infected red blood cells (RBCs), contributing to microcirculatory blockages [8], affect proliferation within RBCs [4], and impact on artemisinin susceptibility [9], thereby influencing disease severity. Therefore, understanding how P. falciparum withstands and adapts to such thermal fluctuations is essential to unravelling malaria pathogenesis. In eukaryotic cells, heat stress disrupts protein homeostasis, increasing the risk of proteins to misfold, which potentially leads to the formation of toxic aggregates [10]. To counteract this, cells trigger a response characterized by the transcriptional upregulation of heat shock proteins (Hsp) [11,12]. These molecular chaperones maintain proteostasis by assisting proper protein folding, preventing aggregation, and facilitating degradation of damaged proteins [13,14]. While Hsp expression is primarily induced by elevated temperatures, it can also respond to metabolic and oxidative stress [15,16]. Under physiological conditions, transcription of inducible hsp genes remains inactive [17], reflecting the transient, protective nature of this process. In P. falciparum, proteostasis under febrile stress relies on a specialized chaperone network, including exported and cytosolic proteins such as PfHsp70-x and PfHsp110, which play key roles in protein folding, trafficking and stress tolerance [14,18,19]. However, how these chaperone systems are coordinated with other adaptive mechanisms remains poorly understood. Beyond the cellular response, stress adaptation can also occur through intercellular communication. The bystander effect describes how naïve cells acquire stress-associated traits after receiving molecular signals from stressed neighbouring cells [20]. This phenomenon often involves the release of extracellular vesicles (EVs) whose production is triggered by adverse conditions, including heat shock [21]. EVs can carry proteins, lipids, and nucleic acids that mediate signalling between cells. Similarly, misfolded or aggregation-prone proteins can be cleared via EVs under proteotoxic stress [22–24], suggesting a role for vesicle-mediated secretion in proteostasis maintenance [25,26]. EVs are heterogeneous in origin and size, being categorized into three groups: apoptotic bodies, microvesicles, and exosomes [27]. The biogenesis of EVs can occur through different pathways, one of which is orchestrated by the sequential recruitment of the Endosomal Sorting Complex Required for Transport (ESCRT), a machinery divided into 4 different complexes (ESCRT-0, -I, -II, and -III) and the accessory Vps4 complex [28]. Although the different ESCRT complexes are well conserved across the eukaryotic lineage, Plasmodium lacks most components of ESCRT-0, -I and -II [29], relying instead on a non-canonical ESCRT-III-based pathway. We previously showed that ESCRT-III activation in P. falciparum is mediated by the Bro1 domain-containing protein PfBro1 [30], which binds and induces the recruitment of the Snf7 domain-containing proteins, PfVps32 and PfVps60, triggering membrane budding in the model of giant unilamellar vesicles [30,31]. Moreover, deletion of Pfvps60 significantly reduces EV production [30], demonstrating the importance of this pathway in vesicle biogenesis. As occurs in other organisms, P. falciparum-derived EVs play key roles in parasite-parasite communication, carrying diverse cargoes, such as proteins, mRNA, miRNA, and other small regulatory RNAs [32–34]. These vesicles contribute to essential processes such as RBC remodelling and invasion [35]. However, despite the well-studied participation of EVs during heat shock response in other eukaryotic cells, their role in the malaria parasite under febrile, proteotoxic conditions remains poorly understood. In particular, whether EVs participate in maintaining proteostasis or transmitting stress signals during heat shock in P. falciparum has been scarcely explored. In this work, by combining proteomic, imaging, and functional analyses, we investigated the contribution of EV-associated cargoes to stress adaptation and parasite survival. Our results confirm that parasite-derived EVs enriched in aggregative proteins and chaperones are essential for recovery from conditions mimicking febrile stress, suggesting that this mechanism plays a protective role during malaria infection. Material and methods Unless otherwise stated, reagents were sourced from Sigma-Aldrich Corporation (St. Louis, MO, USA). P. falciparum cultures Intraerythrocytic stage P. falciparum 3D7 and PfVps60KO parasites were cultured in 3% hematocrit group B erythrocytes in Roswell Park Memorial Institute (RPMI) medium containing 0.5% Albumax II (Life Technology, Auckland, New Zealand) and 2 mM L-glutamine. Experimental procedures followed previously published protocols [30]. Heat shock assays P. falciparum cultures of the 3D7 wild type and PfVps60KO strains were synchronized to ring stages with a 5% sorbitol lysis and grown at 37 °C for 20–25 h, when parasitemia was adjusted to 2%. Cultures were then incubated at 41 °C for 3 h and placed back at 37 °C, following reported protocols [36]. For Western blot assays, samples were collected before heat shock (0 h), at an intermediate point during heat shock (1.5 h), and at 5, 7, and 9 h after the start of the experiment, corresponding to approximately 2, 4 and 6 h after completion of the heat shock pulse. These time points capture mid trophozoite, late trophozoite/early nuclear division, and late trophozoite/schizont stages, respectively. A final time point at 24 h post heat shock was included to analyze longer-term effects on parasite development and progression to the next intraerythrocytic cycle. In parallel, Giemsa-stained thin blood smears were prepared at each time point using standard protocols to analyze parasite morphology and stage distribution. Only well-preserved smears with homogeneous staining were included in the analysis, and parasite stages were assigned based on established morphological criteria. Images were taken with an Olympus IX51 microscope equipped with a cooled CCD Digital Camera (Olympus, Tokyo, Japan). Parasite survival was quantified by flow cytometry using Syto11 (Thermo Fisher Scientific, Inc., OR, USA) in a 5-laser Cytoflex flow cytometer (Beckman Coulter Life Sciences, CA, USA), with the side- and forward-scatter on a logarithmic scale to determine the cell population, and 10,000 events acquired for each sample. Syto11 fluorescence signal was gated using the B525 channel and survival percentages were calculated as the total parasitemia at the next generation of heat shock-exposed cultures (60–65 h after sorbitol treatment) relative to control cultures and expressed as percentage. This time frame ensured that all parasites had completed the cycle, including those affected by heat shock, which usually showed delayed growth [36]. All experiments were performed in three biological replicates. In all cases, control cultures were included (from 3D7 and PfVps60KO strains), which were maintained at 37 °C throughout the whole experiment. EVs purification EVs isolation was performed using sequential centrifugation steps followed by size-exclusion chromatography (SEC). This approach has been shown to reduce contamination from soluble proteins compared to other isolation methods. In addition, the combination of ultrafiltration and SEC has been reported to improve EV yield relative to ultracentrifugation-based protocols [37], while SEC itself allows efficient separation of vesicles from free protein components, thereby enhancing sample purity [38]. EV isolation was performed as in [30,39] from conditioned media collected either immediately after the heat shock exposure or following a 24 h recovery period. Parasite cultures were first cleared of cells and debris through sequential low-speed centrifugation steps (400× g, 10 min), followed by two rounds at 2,000× g for 10 min each. The clarified supernatant was concentrated in 100 kDa cut-off AmiconTM Ultra-15 centrifugal filters (Millipore-Merck, Cork, Ireland), and one mL of the concentrated sample was subjected to SEC on Sepharose CL-4B columns equilibrated in 1× phosphate-buffered saline (PBS). Fractions were collected by gravity flow, and EV-enriched fractions were further concentrated to a final volume of 200 µL for downstream analyses. EV size and abundance measurements Particle size distribution and relative particle abundance were analyzed by dynamic light scattering (DLS) as described before with minor modifications [30]. Purified EVs were diluted in 0.22 µm-filtered 1× PBS and gently mixed. Measurements were performed at room temperature using a Zetasizer Nano S instrument (Malvern Instruments, Ltd., Malvern, UK), and each sample was analyzed in triplicate. The hydrodynamic diameter and derived count rate (kilo counts per second, kcps) were recorded to estimate vesicle size and distribution and relative particle concentration. Cryogenic transmission electron microscopy (cryoTEM) For EV imaging, a 1.5 µL aliquot of purified EVs sample was applied onto the carbon side of a glow-discharged Lacey Carbon 300 mesh copper grid (Ted Pella Inc. Redding, CA, USA). Grids were maintained at 100% humidity in a Vitrobot Mark III chamber (FEI Company, Eindhoven, The Netherlands), where excess liquid was automatically blotted with filter paper prior to cryo-immobilization by plunge freezing in liquid ethane. Samples were transferred to a Tecnai Spirit transmission electron microscope (FEI Company, Eindhoven, The Netherlands) using a cryo-holder (Gatan, Pleasanton, USA). Imaging was performed under cryogenic conditions at 120 kV using low-dose settings. Micrographs were acquired with a 1k × 1k CCD Megaview III camera. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) protein identification EV-associated proteins were extracted under denaturing conditions (8 M urea, 50 mM ammonium bicarbonate, pH 8.5) supplemented with protease inhibitors and disrupted by brief sonication. LC-MS/MS analyses were performed using nano-flow liquid chromatography coupled to an Orbitrap mass spectrometer operating in data-dependent acquisition mode. Database searching and peptide identification were carried out using Protein Discoverer with the Sequest HT algorithm and Percolator-based false discovery rate filtering, following workflows as previously described [40]. Search was conducted against a composite database containing the UniProt P. falciparum 3D7 proteome, human protein entries, and common laboratory contaminants. Proteins were retained for analysis if detected in at least 2 out of 3 biological replicates for PfVps60KO samples and at least 3 out of 4 biological replicates for wild type (WT) samples. Heatmaps were generated using ClustVis (https://biit.cs.ut.ee/clustvis/). Data were log2-transformed (log2(x + 1)), mean-centered and scaled by rows (unit variance scaling). Hierarchical clustering was performed using Euclidean distance and average linkage for both rows and columns. Analysis of aggregation-prone proteins from EV-derived extracts Aggregation tendency was calculated as previously described [41]. Briefly, the TANGO algorithm [42] was run using standard parameters (pH 7.4, 37 °C, ionic strength = 0.25 mM) against the entire P. falciparum 3D7 proteome. The resulting β-aggregation propensity (“AGG” value) for each protein was normalized using a min–max scaling approach, where each value was rescaled according to the minimum and maximum AGG values obtained across the full proteome, yielding a normalized scale from 0 (lowest aggregation propensity) to 1 (highest aggregation propensity). These normalized AGG values were the ones assigned to the proteins identified in the EV-derived extracts. Based on commonly used empirical interpretations, values below 0.01 were considered low- or non-aggregative, whereas values above 0.05 were indicative of aggregation-prone proteins [42]. Protein extraction and Western blot assays Parasitized RBCs (pRBCs) from the 3D7 or PfVps60KO strains were washed twice with 1× PBS supplemented with 1× cOmpleteTM. Parasite pellets were then resuspended in six volumes of 0.15% (w/v) saponin and incubated for 10 min at 4 °C to lyse the host RBC membrane, followed by centrifugation at 10,000× g for 15 min at 4 °C. The resulting pellet was washed thoroughly with 1× PBS containing protease inhibitors and subsequently extracted in RIPA buffer (150 mM NaCl, 2 mM EDTA, 10% glycerol, 1% Triton X-100, 0.5% sodium deoxycholate, 0.2% SDS, 1× cOmplete™, 40 mM tris-HCl, pH 7.4) under continuous agitation at 4 °C. Afterwards, cell lysates were sonicated for 2 s at 40% amplitude and incubated on ice for 15 min to ensure complete protein solubilization. Finally, debris were removed by centrifugation at 20,000× g for 15 min at 4 °C. For Western blot assays, 20 µg of protein extracts from both parasite lines were resolved by 12% SDS-PAGE. After electrophoresis, proteins were transferred onto nitrocellulose membranes (Cytiva Life Sciences, Marlborough, MA, USA) and blocked overnight at 4 °C with 5% (w/v) skim milk prepared in tris-buffered saline (TBS) containing 0.5% Tween 20 (TBS-T). Membranes were then washed and incubated for 3 h with either rabbit anti-PfHsp70-1 antibodies (StressMarq Biosciences, Victoria, BC, Canada; 1:10,000) or mouse monoclonal anti-spectrin antibodies (α and β, Sigma-Aldrich; 1:10,000) diluted in TBS-T. Afterwards, membranes were washed 5× with TBS-T and incubated for 1 h with either goat anti-rabbit (Abcam, Cambridge, UK; 1:10,000) or sheep anti-mouse IgG horseradish peroxidase-labelled secondary antibodies (Cytiva Life Sciences), dissolved in TBS-T. Detection was performed using the ECL Prime Western blotting detection reagent (Cytiva Life Sciences). Band intensity was quantified by densitometric analysis using the Fiji software. The relative expression of PfHsp70-1 in each timepoint was calculated by normalizing the integrated density values to spectrin signal from the same membrane. No apparent changes in spectrin signal were observed across heat shock and control conditions in any of the biological replicates, indicating that spectrin is not affected under the experimental conditions used. Confocal fluorescence microscopy P. falciparum parasites from the 3D7 and PfVps60KO strains were seeded in an 8-well µ-slide chamber (ibidi GmbH, Germany). Briefly, pRBCs were washed twice with warm 1× PBS and deposited into wells pre-coated with 5 mg/mL concanavalin A. Cells were incubated for 10 min at 37 °C to allow adhesion, after which unbound RBCs were gently removed by washing with 1× PBS at 37 °C. For fixed preparations, adhered pRBCs were incubated with 4% paraformaldehyde for 20 min at 37 °C and thoroughly washed with 1× PBS. Live or fixed cells were then incubated for 30 min at 37 °C with 0.5 µM YAT2150 and 5 µg/mL Hoechst 33342. After staining, samples were rinsed with 1× PBS to remove excess dyes. Fluorescence imaging was performed using a Leica TCS SP5 confocal microscope (Leica Microsystems GbmH, Germany) equipped with a 63×/1.4 NA oil immersion objective. Hoechst 33342 was excited with a 405 nm diode laser, and YAT2150 with a 488 nm diode-pumped solid-state laser. Emission was collected between 415–460 nm for Hoechst 33342 and 490–590 nm for YAT2150. Quantitative analysis of protein aggregation levels in live P. falciparum cultures Aggregation tendency was calculated as previously described [43]. For the quantitative analysis of protein aggregation, P. falciparum 3D7 and PfVps60KO parasites were synchronized to the trophozoite stage (20–25 h post infection, hpi) via sorbitol lysis. The resulting cultures were adjusted to 3% hematocrit and 2% parasitemia in complete RPMI medium and one sample from each line was subjected to heat shock at 41 °C for 2 h in a water bath. This shorter exposure was selected to minimize excessive parasite death during EV supplementation, thereby allowing the determination of functional rescue during the recovery phase. Corresponding control samples were maintained at 37 °C under identical conditions. Immediately after this treatment (t0), 5 mL of each culture sample were harvested by centrifugation (1,500× g, 5 min), and the resulting pellet was resuspended in 300 µL of osmotic lysis buffer (4.5 mg/mL NaCl, supplemented with 1× cOmpleteTM EDTA-free protease inhibitor). An additional 5-mL aliquot from each condition (37 °C and heat shock) was collected after further incubation at 37 °C for 4 h. All samples were incubated overnight at 4 °C under constant stirring to ensure complete lysis. Lysates were clarified at 14,000× g for 5 min, and protein content was determined using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA). For the aggregation assay, 30 µg of total protein per sample were adjusted to 100 µL with 1× PBS and loaded in triplicate into a black 96-well plate (Greiner Bio-One, Madrid, Spain). Thioflavin T (ThT) was added to a final concentration of 25 µM from a freshly prepared stock in 1× PBS. Plates were incubated for 15 min at room temperature in the dark under gentle stirring. Fluorescence emission spectra (500–700 nm) were acquired using an excitation wavelength of 450 nm in an Infinite® M Plex multimode microplate reader (Tecan Trading AG, Männedorf, Switzerland). Fluorescence values were corrected by subtracting a ThT blank prepared in 1× PBS under identical conditions. Data was analyzed by calculating the area under the curve (AUC) from each emission spectrum using GraphPad Prism software (v10.4.2). Statistical analysis was performed using a two-way ANOVA to evaluate the effects of strain and treatment at each time point. Extracellular vesicle supplementation during heat shock survival assays For the heat shock survival assays, parasites from the 3D7 and PfVps60KO lines were synchronized to the trophozoite stage (24–28 hpi) using a 70% Percoll gradient followed by sorbitol lysis. Synchronized cultures were adjusted to 3% hematocrit and 2% parasitemia and divided into 1 mL aliquots in complete RPMI. Heat shock was performed by incubating parasite aliquots in a water bath at 41 °C for 2 h, while control samples were kept at 37 °C under identical conditions. For EV supplementation experiments, PfVps60KO parasites were incubated during heat shock in the presence of four different EV preparations derived from 3D7-parasitized erythrocytes: three obtained from heat shocked cultures (s1, s2, and s3) and one obtained from a culture maintained at 37 °C (s4). All EV preparations were added at a fixed volume (25 µL per mL of culture, corresponding to 2.5% of total volume), but containing decreasing EV concentrations: s1 (7437.7 ± 47.9 kcps/mL), s4 (4153.0 ± 36.4 kcps/mL), s2 (2781.0 ± 51.2 kcps/mL) and s3 (1017.0 ± 28.9 kcps/mL). Additionally, a 1× PBS control (2.5% v/v) was included. After heat exposure, cultures were returned to standard culturing conditions and allowed to recover for 24 h. Parasite survival was quantified by flow cytometry using Syto11 (Thermo Fisher Scientific, Inc.) in a 5-laser Cytoflex flow cytometer (Beckman Coulter Life Sciences, CA, USA), as previously described. Statistical analyses Statistical analyses were performed using GraphPad Prism (version 10.4.2, GraphPad Software). Data are presented as mean ± standard error of the mean (SEM) unless otherwise indicated. The specific statistical tests applied for each experiment are described in the corresponding figure legends. For comparisons involving more than two groups, one-way or two-way analysis of variance (ANOVA) was used as appropriate, followed by either Tukey’s or Šídák’s multiple comparison tests to analyze pairwise differences. For comparisons between two groups, unpaired Student’s t-test was applied. Statistical significance was defined as p < 0.05. Results The disruption of PfVps60 in P. falciparum leads to an impairment in the export of the heat shock proteins PfHsp70-x and PfHsp110 Given the central role of molecular chaperones in parasite proteostasis under stress conditions, we first investigated whether disruption of the ESCRT-III protein PfVps60 affects the EV-associated export of stress-related proteins in P. falciparum. According to previous findings from our group, knockout (KO) of the Pfvps60 gene resulted in a significant reduction in EVs production compared to the wild type (WT) 3D7 strain, and impaired export of other ESCRT subunits, including PfVps32 and PfBro1, to the host cell cytoplasm [30]. Despite the reduced EV yield, cryoTEM analysis of PfVps60 preparations confirmed the presence of nanosized vesicular structures with morphology consistent with EVs, including heterogeneous populations with occasional electron-dense vesicular material (Figure S1). These observations indicate that PfVps60 disruption does not abolish vesicle formation but rather affects EV abundance and/or cargo composition. WT 3D7-derived EV morphology was previously characterized in our group [39]. To further investigate whether PfVps60 loss broadly affects vesicular trafficking of proteins relevant to proteostasis, we performed a comparative LC-MS/MS proteomic analysis of EVs isolated from WT and PfVps60KO pRBCs. A total of 111 P. falciparum proteins were initially identified in four WT and three PfVps60KO biological replicates. After excluding proteins not consistently detected across biological replicates, 69 were retained for analysis (Table S1, see Material and Methods). A heatmap representation of protein abundance across samples revealed distinct clustering patterns between WT- and PfVps60KO-derived EVs (Figure S2). Among the analyzed proteins, 19 were absent in EVs from the PfVps60KO strain (Figure 1A). In particular, several of the missing proteins have been previously reported in EVs isolated from clinical samples, including MSA180, GAC, SERA5, MAHRP1, Alba 4, EXP1 and EBA175 [35]. Additionally, the export of EV-associated proteins such as RACK1, ADF1, PHISTc and ARF1, previously characterized in pRBC cultures [32,33], was also disrupted in the PfVps60KO strain. Interestingly, EV-associated export of the molecular chaperones PfHsp70-x (K7NTP5) and PfHsp110 (Q8IC01) was not found in the KO line (Figure 1B). Figure 1.Comparative proteomic analysis of EVs from P. falciparum 3D7 and PfVps60KO strains. (A) Venn diagram showing the distribution of proteins identified in EVs derived from the WT 3D7 strain and from the PfVps60KO mutant, with the overlapping region indicating proteins shared between both samples. (B) List of the 19 proteins found in 3D7 EVs and absent in EVs derived from the PfVps60KO strain.Two diagrams: a Venn diagram and a table comparing 3D7 and PfVps60KO vesicle proteins, with 19 unique proteins listed for 3D7.The figure shows two visuals comparing proteins in extracellular vesicles from Plasmodium falciparum 3D7 and PfVps60 knockout strains. On the left, a Venn diagram displays two overlapping circles labeled 3D7 and PfVps60KO. The 3D7 circle contains the number 19, the overlap contains 47, and the PfVps60KO circle contains 3, indicating counts of proteins unique or shared between vesicles from each strain. On the right, a table lists the 19 proteins present in 3D7 vesicles and absent from PfVps60KO vesicles. The table has two columns labeled Uniprot ID and Description. The Uniprot IDs and descriptions are: C0H5F4, Reticulocyte binding protein 2 homolog b; A0A5K1K8J3, Glideosome associated connector, GAC; Q8IK15, EMP1 trafficking protein, PTP5; Q8IJ49, Merozoites associated armadillo repeats protein, MAAP; Q8IBE8, Erythrocyte binding antigen 175, EBA175; Q8IJQ4, Merozoite surface protein MSA180; K7NTP5, Heat shock protein 70, HSP70 x; Q9TY95, Serine repeat antigen protein 5, SERA5; C0H5L9, Membrane associated histidine rich protein 1, MAHRP1; Q8IIF0, Exported protein 1, EXP1; Q8IBA0, Receptor for activated c kinase, RACK1; Q8IC01, Heat shock protein 110, HSP110; Q8IJS2, Duffy binding like merozoite surface protein, MSPDBL1; Q8IM16, Plasmepsin IV; Q03400, S antigen protein; A0A5K1K8Y8, DNA and RNA binding protein Alba 4; Q8I467, Cofilin and actin depolymerizing factor homolog 1, ADF1; Q8I2F2, Plasmodium exported protein, PHISTc; Q7KQL3, ADP ribosylation factor 1, ARF1. Comparative proteomic analysis of EVs from P. falciparum 3D7 and PfVps60KO strains. (A) Venn diagram showing the distribution of proteins identified in EVs derived from the WT 3D7 strain and from the PfVps60KO mutant, with the overlapping region indicating proteins shared between both samples. (B) List of the 19 proteins found in 3D7 EVs and absent in EVs derived from the PfVps60KO strain. Together, these results indicate that the ESCRT-III machinery is required to maintain the protein composition of EVs released by infected erythrocytes. Given the selective loss of EV-associated chaperones observed in the PfVps60KO line, we next examined whether disruption of ESCRT-mediated vesicular trafficking compromised the parasite’s ability to adapt to febrile heat stress. Heat shock induces a protective response in the P. falciparum 3D7 WT line but not in the PfVps60-deficient strain During the blood stage of malaria infection, recurrent febrile episodes expose P. falciparum to periodic thermal stress. As in other eukaryotic cells, these temperature fluctuations promote protein unfolding and aggregation, thereby activating protective chaperone-mediated mechanisms [19,44]. Among the most studied parasite molecular chaperones, PfHsp70-1 (Q8IB24) is a cytosolic protein essential for parasite survival that is upregulated in response to stress [45]. Other chaperones, including PfHsp90 and a large family of exported Hsp40 co-chaperones, cooperate with PfHsp70-1 to maintain proteostasis under fluctuating host temperatures [18,46]. PfHsp70-x, a dispensable exported chaperone unique to P. falciparum [47], and PfHsp110, which prevents protein aggregation in vitro and protects the parasite’s asparagine-rich proteome during febrile episodes [19], are also part of this network. To analyze whether PfVps60 disruption impacted the response of P. falciparum to heat shock, trophozoite stage parasites were incubated for 3 h at 41 °C, and samples were collected before, during, and after heat shock exposure. Subsequently, expression levels of PfHsp70-1 (whose export is not affected in the KO strain, see Table S1) were analyzed by Western blot and compared with control parasites maintained at 37 °C (Figure S3). Under normal conditions, mostly non-significant differences in PfHps70-1 expression were detected between the WT 3D7 and PfVps60KO lines (Figure 2A). However, during heat shock treatment, PfHsp70-1 expression in the KO strain failed to increase following thermal stress, in contrast to the robust induction observed in WT parasites (Figure 2B). It is important to point out that the peak of expression detected at 7 h matches with schizont formation, a stage likely requiring enhanced chaperone activity to support cell division. Figure 2.PfHsp70-1 expression in P. falciparum WT and PfVps60KO lines under heat shock stress. Protein expression levels of PfHsp70-1 in parasites synchronized at the trophozoite stage (initial), either (A) maintained at 37 °C or (B) subjected for 3 h to heat shock at 41 °C, followed by recovery at 37 °C. The indicated time points represent hours elapsed since the start of the heat shock treatment. Protein levels were normalized to spectrin. Quantification was obtained by densitometric analysis of Western blot signals (shown in Figure S3 and Supplementary S1 Data). Data represents the mean ± SEM of three independent experiments. a.u., arbitrary units. Statistical analysis was performed using a two-way ANOVA comparing each condition to the corresponding 3D7 control. **p ≤ 0.01, ****p ≤ 0.0001.Two bar charts of PfHsp70-1 relative expression in WT 3D7 and PfVps60KO parasites at 37 degree C and 41 degree C over time.The figure shows two grouped bar charts comparing PfHsp70-1 protein expression in Plasmodium falciparum WT 3D7 and PfVps60KO lines over time. The left chart is labeled 37 degree C and the right chart is labeled 41 degree C. In both charts, the x axis shows sampling times labeled Initial, 1.5 h, 5 h, 7 h, 9 h, and 24 h. The y axis in both charts is labeled Relative expression in arbitrary units, ranging from 0 to 6 in the 37 degree C chart and from 0 to 15 in the 41 degree C chart, with tick marks at whole numbers. Each time point has a pair of vertical bars with error bars, a dark bar for WT 3D7 and a light bar for PfVps60KO, as indicated in a legend on each chart. At 37 degree C, both strains show modest changes, with WT 3D7 slightly higher at 1.5 h and 7 h but remaining below 6 units. At 41 degree C, WT 3D7 bars rise from a low initial value to a prominent peak near 7 h and stay elevated through 9 h and 24 h, while PfVps60KO bars stay comparatively low at all times. Curly brackets above several paired bars are marked with 2 or 4 asterisks. All data are approximate. PfHsp70-1 expression in P. falciparum WT and PfVps60KO lines under heat shock stress. Protein expression levels of PfHsp70-1 in parasites synchronized at the trophozoite stage (initial), either (A) maintained at 37 °C or (B) subjected for 3 h to heat shock at 41 °C, followed by recovery at 37 °C. The indicated time points represent hours elapsed since the start of the heat shock treatment. Protein levels were normalized to spectrin. Quantification was obtained by densitometric analysis of Western blot signals (shown in Figure S3 and Supplementary S1 Data). Data represents the mean ± SEM of three independent experiments. a.u., arbitrary units. Statistical analysis was performed using a two-way ANOVA comparing each condition to the corresponding 3D7 control. **p ≤ 0.01, ****p ≤ 0.0001. Giemsa-stained preparations revealed that, 24 h after heat shock treatment, WT parasites progressed through the intraerythrocytic cycle and reached ring stage at the expected time (Figure 3A). In contrast, PfVps60KO parasites exposed to the same conditions exhibited a marked loss of fitness, evidenced by the presence of ruptured cells and pyknotic nuclei (Figure 3, arrowheads). To quantitatively measure parasite recovery following heat shock, survival was independently evaluated by flow cytometry using Syto11 staining. Consistent with morphological observations, flow cytometry analysis revealed that approximately 80% of WT parasites survived heat shock, whereas survival of PfVps60KO parasites dropped to ∼15% 24 h post-treatment (Figure 3B). Together, these results indicated that PfVps60-deficient parasites fail to efficiently recover from febrile stress, leading to loss of viability and impaired progression through the intraerythrocytic cycle. It is important to point out that flow cytometry does not distinguish between actively replicating and dormant parasites, which may retain viability while exhibiting reduced metabolic activity. Therefore, we cannot exclude that a fraction of the surviving population corresponds to transiently growth-arrested parasites. Figure 3.Heat shock impairs recovery of PfVps60KO parasites. (A) Representative Giemsa-stained images of P. falciparum cultures at the indicated times after heat shock exposure (41 °C, 3 h) of synchronized trophozoite stage parasites from the 3D7 or PfVps60KO lines. Arrowheads indicate ruptured pRBCs. Scale bar = 10 µm. (B) Parasite survival following heat shock quantified 24 h post-treatment for 3D7 and PfVps60KO lines, expressed as parasitemia determined by flow cytometry using Syto11 staining. Data represent mean ± SD of 3 independent biological replicates. Statistical analysis was performed using a Student’s t-test comparing each condition to the corresponding untreated control. **p ≤ 0.01, ****p ≤ 0.0001.Two microscopy grids and one bar chart comparing 3D7 and PfVps60KO survival after 41 degree heat shock over 24 hours.The figure shows two grouped microscopy grids above a bar chart. The microscopy grids on the top are labeled A. Each grid contains side by side Giemsa stained parasite preparations for lines labeled 3D7 and PfVps60KO. The left grid is labeled Initial, 1.5 hours, and 5 hours along the left edge. The right grid is labeled 7 hours, 9 hours, and 24 hours along the left edge. Within each microscopic field, circular red blood cells surround darker stained Plasmodium falciparum parasites at trophozoite or later stages. In the 3D7 images, intact infected cells appear at each time point. In the PfVps60KO images at 24 hours, several infected cells are disrupted with arrowheads pointing to ruptured red blood cells and condensed parasite nuclei. The lower visual, labeled B, is a vertical bar chart titled percent survival. The y axis label reads percent survival, with a scale from 0 to 150 at intervals of 50. The x axis categories, written at an angle, are 3D7, 3D7 HS, PfVps60KO, and PfVps60KO HS. Bars for 3D7 and PfVps60KO reach near the 100 mark, 3D7 HS is slightly lower, and PfVps60KO HS is near 15. Brackets with double * and quadruple * appear above the paired conditions. All data are approximate. Heat shock impairs recovery of PfVps60KO parasites. (A) Representative Giemsa-stained images of P. falciparum cultures at the indicated times after heat shock exposure (41 °C, 3 h) of synchronized trophozoite stage parasites from the 3D7 or PfVps60KO lines. Arrowheads indicate ruptured pRBCs. Scale bar = 10 µm. (B) Parasite survival following heat shock quantified 24 h post-treatment for 3D7 and PfVps60KO lines, expressed as parasitemia determined by flow cytometry using Syto11 staining. Data represent mean ± SD of 3 independent biological replicates. Statistical analysis was performed using a Student’s t-test comparing each condition to the corresponding untreated control. **p ≤ 0.01, ****p ≤ 0.0001. Heat shock triggers a rapid but transient increase in EV production To further explore the reduced survival observed in the PfVps60KO line after heat shock, we examined EV production under conditions of thermal stress. Synchronized parasites from the WT and KO lines were exposed to heat shock (41 °C, 3 h) and subsequently returned to 37 °C; a control sample was maintained at 37 °C throughout the assay. When EVs were quantified 24 h after heat shock, the number of released EVs from heat-stressed WT cultures was significantly lower than that from cultures maintained continuously at 37 °C (Figure 4A). As expected, PfVps60KO parasites produced fewer EVs under normal conditions, as observed in [30], and their EV abundance remained low following heat shock (Figure 4A). Figure 4.Relative EV abundance in P. falciparum 3D7 and PfVps60KO lines after heat shock. Derived count rate of purified EVs from three independent replicates, expressed in kilo counts per second (kcps), measured at (A) 24 h and (B) 2 h post-heat shock treatment. Data represent mean ± SD of 3 independent biological replicates. Statistical analysis was performed using one-way ANOVA followed by Tukey’s multiple comparisons test comparing each condition to the corresponding control. *p ≤ 0.05, ***p ≤ 0.001, ****p ≤ 0.0001.Two bar charts comparing extracellular vesicle count rate for 3D7 and PfVps60KO at 37 and 41 degrees Celsius after heat shock.The figure shows two bar charts of extracellular vesicle measurements in Plasmodium falciparum 3D7 and PfVps60 knockout lines after heat shock. In the first bar chart, titled 24 hours after heat shock, the x axis label is Temperature with tick marks at 37 degrees Celsius and 41 degrees Celsius. The y axis label is Count rate in kilo counts per second with a scale from 0 to 5000 in intervals of 1000. For 3D7 at 37 degrees Celsius, the bar is near 4100. For PfVps60 knockout at 37 degrees Celsius, the bar is near 500. For 3D7 at 41 degrees Celsius, the bar is near 1800. For PfVps60 knockout at 41 degrees Celsius, the bar is near 700. Error bars extend above each bar and horizontal brackets with asterisks appear over group comparisons. In the second bar chart, titled 2 hours after heat shock, the x axis label is Temperature with tick marks at 37 degrees Celsius and 41 degrees Celsius. The y axis label is Count rate in kilo counts per second with a scale from 0 to 14000 in intervals of 2000. For 3D7 at 37 degrees Celsius, the bar is near 200. For PfVps60 knockout at 37 degrees Celsius, the bar is near 300. For 3D7 at 41 degrees Celsius, the bar is near 9500 with a long upper error bar. For PfVps60 knockout at 41 degrees Celsius, the bar is near 1500. A legend identifies 3D7 and PfVps60 knockout bars. All data are approximate. Relative EV abundance in P. falciparum 3D7 and PfVps60KO lines after heat shock. Derived count rate of purified EVs from three independent replicates, expressed in kilo counts per second (kcps), measured at (A) 24 h and (B) 2 h post-heat shock treatment. Data represent mean ± SD of 3 independent biological replicates. Statistical analysis was performed using one-way ANOVA followed by Tukey’s multiple comparisons test comparing each condition to the corresponding control. *p ≤ 0.05, ***p ≤ 0.001, ****p ≤ 0.0001. Previous studies in other cell types have reported increased EV release after heat stress [21,48], contrasting with our observations. These differences suggest that thermal stress responses may differ among organisms, reflecting distinct survival strategies. Given the combination of reduced EV release in WT parasites and the markedly lower viability of the EV-deficient line under heat stress, we considered an alternative explanation: in P. falciparum, rather than a sustained increase in EV secretion, febrile stress might trigger rapid EV turnover through enhanced release followed by internalization. Such mechanisms could confer a selective advantage by facilitating the uptake of EV-associated protective or stress-response factors by other neighbouring parasite cells. To test this hypothesis, EVs were isolated at earlier time points after heat shock exposure. When EV harvesting was performed 2 h after the heat shock, a significant increase in EV abundance was detected in both 3D7 and PfVps60KO cultures compared with untreated controls (Figure 4B). These results indicate that heat shock transiently promotes EV release in P. falciparum, supporting the notion that EV secretion is part of an immediate adaptive response to febrile stress. P. falciparum-derived EVs contain aggregation-prone proteins Building on the observed stress sensitivity upon PfVps60 disruption, we next investigated whether EVs released by P. falciparum 3D7 contained aggregation-prone or chaperone-related proteins that could contribute to the parasite’s stress response. Previously, we demonstrated the presence of aggregation-prone proteins in live P. falciparum cultures, which accumulated within the parasite cytoplasm throughout its life cycle [49]. However, the export of these protein aggregates via EVs and their potential activity in recipient cells has remained unexplored. To address this, proteins isolated from 3D7-derived EVs were extracted and compared with a previously defined set of aggregative proteins identified in whole-parasite lysates [41]. Across three independent biological replicates, a total of 67 EV-associated proteins were identified, of which 30 (44.8%) overlapped with the aggregation-prone set (Figure 5A, Table S2). Figure 5.LC-MS/MS data analysis of aggregation-prone proteins found in pRBC lysates and in EVs of the 3D7 line. (A) Venn diagram showing the number of proteins identified in 0.1% SDS-resistant pellets purified from trophozoite and ring stages as well as in P. falciparum-derived EVs, and the proteins shared between the two samples. (B) List of aggregation-prone proteins identified in P. falciparum-derived EVs, according to TANGO score. Data represents hits detected in at least 3 of 4 biological replicates.Two visuals: a Venn diagram and a table showing overlap between 0.1 percent SDS resistant proteins and exported proteins in 3D7 line.The figure shows two visuals comparing aggregation prone proteins from Plasmodium falciparum 3D7 parasites and extracellular vesicles. The first visual is a Venn diagram with two overlapping circles. The left circle is labeled 0.1 percent sodium dodecyl sulfate resistant proteins, trophozoites and rings, and contains the number 349. The right circle is labeled Exported proteins and contains the number 37. The overlap between the circles contains the number 30, indicating shared proteins between sodium dodecyl sulfate resistant pellets and exported proteins. All data are approximate. The second visual is a two column table titled Uniprot ID and Description that lists aggregation prone proteins detected in exported vesicles. The Uniprot ID column contains entries such as Q76NM3, Q7K6A4, C0H4V6, Q8I4X0, Q8IKK7, Q9TY99, Q8IIV2, Q8IOP6, Q8I492, Q8IKF0, Q8IB24, C6KSV0, C6KT18, Q8IIV1, Q8IAX8, Q8I6U8, C0H571, Q8I467, Q8IDQ9, Q8IKW5, Q7KQK6, C6KT34, Q7KQL5, Q8IC05, Q6LFH8, C6KTA4, Q8IDG8, Q8I2G1, Q8I2F2, and A0A5K1K8J3. The Description column provides each protein name and function, including enzymes, histones, heat shock proteins, cytoskeletal proteins, exported surface antigens, and binding proteins. LC-MS/MS data analysis of aggregation-prone proteins found in pRBC lysates and in EVs of the 3D7 line. (A) Venn diagram showing the number of proteins identified in 0.1% SDS-resistant pellets purified from trophozoite and ring stages as well as in P. falciparum-derived EVs, and the proteins shared between the two samples. (B) List of aggregation-prone proteins identified in P. falciparum-derived EVs, according to TANGO score. Data represents hits detected in at least 3 of 4 biological replicates. Among the EV-associated proteins were the molecular chaperones PfHsp70-1 (Q8IB24) and PfHsp90 (Q8IC05), which form a functional complex involved in protein folding and protection against aggregation [50,51]. In addition, EVs contained aggregation-prone proteins associated with host cell invasion (C0H4V6, Q8I4X0, Q8I6U8), cytoskeletal remodelling (Q7KQL5, Q8IDG8), and gene regulation (C6KSV0, C6KT18, Q8IIV1, Q7KQK6), along with other proteins that may participate in multiprotein complex assemblies (Figure 5B). These findings suggested that P. falciparum-derived EVs contain aggregation-prone and proteostasis-related proteins, supporting a role for EVs during proteotoxic stress. PfVps60-deficient parasites accumulate aggregative proteins in the vicinity of the parasitophorous vacuole Given that PfVps60 disruption selectively impaired the EV-associated export of chaperones and other proteins, we next examined whether loss of this protein affects global proteostasis within the parasite. As discussed earlier, PfHsp110 contributes to P. falciparum proteostasis through its ability to bind unfolded proteins via its substrate-binding domain [19]. Since this chaperone was absent from EVs derived from the PfVps60KO line, we next assessed whether disruption of ESCRT-dependent EV trafficking impacted intracellular protein aggregation. To this end, we performed confocal fluorescence microscopy using YAT2150, a fluorescent probe previously characterized by our group that detects intracellular protein aggregates in Plasmodium cells [41,49]. Fixed (Figure 6A) and live (Figure 6B) P. falciparum cultures of the WT and the PfVps60KO strains were analyzed and compared. In WT parasites, protein aggregates displayed a homogenous cytoplasmic distribution. In contrast, PfVps60-deficient parasites exhibited a distinct localization of the aggregated protein pool toward the periphery of the parasite, accumulating near the parasitophorous vacuole (see zoom panels in Figure 6), a differential distribution particularly evident in live preparations. Importantly, because Pfvps60 knockout does not alter the parasite’s intraerythrocytic development [30], the observed spatial redistribution of aggregative proteins likely reflected an altered proteostasis response rather than a general defect in the asexual cycle under basal conditions. Figure 6.Distribution of aggregative proteins in P. falciparum 3D7 and PfVps60KO strains. Confocal fluorescence microscopy analysis of (A) fixed or (B) live P. falciparum blood stages of the 3D7 WT and PfVps60KO strains stained for 30 min with 0.5 µM YAT2150 (red); nuclei were counterstained with Hoechst 33342 (blue). The bottom panels show zooms of the indicated regions in the merge panels, to allow better visualization. (C) Quantification of aggregation by ThT assay in synchronized trophozoite-stage parasite lysates. Parasites were maintained at 37 °C or exposed to heat shock (HS). Samples were collected immediately after heat shock (t0) or after an additional 4 h recovery period at 37 °C (4 h). Aggregation rates were estimated by calculating the area under the curve (AUC) of the fluorescence emission spectra (500–700 nm). Each symbol represents a biological replicate, and horizontal bars indicate the mean. Statistical analysis was performed using two-way ANOVA with Šídák’s multiple comparisons test. *p ≤ 0.05, **p ≤ 0.01, ****p ≤ 0.0001.Three visuals: two microscopy montages and one scatter plot chart comparing protein aggregation in 3D7 and PfVps60KO parasites.The figure shows two microscopy montages and one scatter plot chart comparing protein aggregation in Plasmodium falciparum 3D7 and PfVps60 knockout strains. The first montage labeled A displays fixed infected erythrocytes. Each column compares 3D7 and PfVps60 knockout parasites through bright field views, nuclei staining, YAT2150 staining, merged fluorescence, merged fluorescence overlaid on bright field, and a magnified view. In 3D7 cells, YAT2150 labeling appears spread through the parasite cytoplasm, whereas in PfVps60 knockout cells the labeling appears concentrated near the edge of the parasite. The second montage labeled B shows the same sequence for live parasites, again contrasting the more uniform distribution in 3D7 with peripheral YAT2150 enrichment in PfVps60 knockout parasites. The third visual labeled C is a scatter plot chart with the x axis labeled t0 and 4 h and the y axis labeled area under the curve of thioflavin T emission 500 to 700 nanometers from 0 to 20000 in steps of 5000. Four conditions are plotted: 3D7 at 37 degrees Celsius, 3D7 after heat shock, PfVps60 knockout at 37 degrees Celsius, and PfVps60 knockout after heat shock. PfVps60 knockout heat shock values at t0 cluster highest, while other conditions form lower clusters that converge by 4 h. Asterisks mark group comparisons. All data are approximate. Distribution of aggregative proteins in P. falciparum 3D7 and PfVps60KO strains. Confocal fluorescence microscopy analysis of (A) fixed or (B) live P. falciparum blood stages of the 3D7 WT and PfVps60KO strains stained for 30 min with 0.5 µM YAT2150 (red); nuclei were counterstained with Hoechst 33342 (blue). The bottom panels show zooms of the indicated regions in the merge panels, to allow better visualization. (C) Quantification of aggregation by ThT assay in synchronized trophozoite-stage parasite lysates. Parasites were maintained at 37 °C or exposed to heat shock (HS). Samples were collected immediately after heat shock (t0) or after an additional 4 h recovery period at 37 °C (4 h). Aggregation rates were estimated by calculating the area under the curve (AUC) of the fluorescence emission spectra (500–700 nm). Each symbol represents a biological replicate, and horizontal bars indicate the mean. Statistical analysis was performed using two-way ANOVA with Šídák’s multiple comparisons test. *p ≤ 0.05, **p ≤ 0.01, ****p ≤ 0.0001. To further determine whether this altered spatial distribution of aggregation-prone proteins translated into changes in global proteostasis dynamics, we performed a quantitative analysis of protein aggregation using ThT fluorescence assays, which enables the detection of β-sheet-enriched proteins in the parasite lysate. As observed in Figure 6C, analysis of ThT fluorescence revealed significant differences in protein aggregation between conditions at early time points. Immediately following heat shock (t0), PfVps60-deficient parasites displayed a marked increase in ThT signal compared to WT and control conditions, indicating an early perturbation of proteostasis. In contrast, 4 h following initial thermal treatment (4 h), WT parasites exhibited a significant reduction in ThT signal under heat shock relative to 37 °C controls, suggesting an active remodelling or clearance of aggregation-prone species. In contrast, PfVps60KO parasites maintained aggregation levels similar to the control. The marked decrease of ThT signal observed in heat-shocked PfVps60KO parasites between t0 and 4 h may reflect a combination of stress-induced remodelling of ThT-reactive species and disposal of misfolded proteins during early recovery. Because ThT fluorescence reports β-sheet-enriched species rather than total protein aggregation, this decrease should not be interpreted solely as aggregate clearance. Aggregation-prone enriched EVs associate with the surface of naïve RBCs following heat shock To investigate whether stress-induced vesicle release influenced neighbouring cells, P. falciparum 3D7 parasites were exposed to heat shock and subsequently labelled with YAT2150. Nuclear staining allowed discrimination between infected and uninfected RBCs. In untreated control cultures, no YAT2150-positive vesicles were observed on the surface of neighbouring RBCs (Figure 7A). In contrast, following heat shock treatment, EVs enriched in aggregative proteins (as indicated by YAT2150 fluorescence) accumulated on the surface of adjacent uninfected RBCs (Figure 7B). In some cases, vesicles were detected at the periphery of pRBCs, apparently ready for release into the extracellular space (Figure 7C, arrowheads). High magnification images revealed that vesicles remained attached to the membrane of neighbouring naïve RBCs (Figure 7D), and in certain cells, YAT2150 fluorescence was diffusely distributed throughout the RBC cytoplasm (Figure 7D, lower row), an observation consistent with a possible cargo transfer into the target cell, although direct uptake was not demonstrated. These results indicated that EVs released from heat-stressed parasites can associate with neighbouring erythrocytes and are in agreement with a potential role of EVs in intercellular communication under stress conditions. Figure 7.Heat shock-induced EVs associate with neighbouring naïve RBCs. Representative confocal fluorescence microscopy images of (A) 3D7 parasites maintained at 37 °C or (B) subjected to heat shock at 41 °C for 30 min. (C) Detail of an infected RBC showing small vesicles adjacent to the plasma membrane. (D) Representative images showing YAT2150-labelled EVs associated with the surface of naïve RBCs. In all cases, cells were stained with 0.5 µM YAT2150 (red) and nuclei counterstained with Hoechst 33342 (blue).Four microscopy montages showing 3D7 parasite infected and naive red blood cells with YAT2150 labelled vesicles and nuclear stain.The figure shows four microscopy montages labeled A, B, C, and D, each arranged in five columns titled Bright field, Nuclei, YAT2150, Merge, and Composite. Each montage depicts Plasmodium falciparum 3D7 parasite infected red blood cells and surrounding uninfected red blood cells. In montage A, the Bright field view shows a uniform field of red blood cells. The Nuclei view shows a few bright puncta inside a subset of cells, while the YAT2150 view shows small puncta overlapping those nuclei. The Merge view combines the nuclear and YAT2150 signals within the same infected cells. The Composite overlay places these signals on the Bright field background, with a scale bar labeled 5 micrometers. Montage B shows a Bright field view with clustered infected cells. The Nuclei view again highlights nuclear puncta in the infected cluster. The YAT2150 view now shows numerous small puncta both within infected cells and scattered near neighboring cells. The Merge view shows overlapping nuclear and YAT2150 signals in infected cells, and the Composite overlay reveals extracellular vesicle like puncta near adjacent red blood cells, also with a 5 micrometers scale bar. Montage C focuses on a single infected red blood cell. The YAT2150 view shows several puncta near the periphery of the cell. In the Composite overlay, arrowheads mark vesicle like structures close to the inner edge of the cell membrane, and a 5 micrometers scale bar is present. Montage D contains two rows, each centered on one red blood cell. In the upper row, YAT2150 puncta lie at the outer surface of an apparently uninfected neighboring cell and remain visible in the Merge and Composite overlays. In the lower row, the YAT2150 signal outlines multiple puncta along the inside edge of the cell, suggesting vesicle like material closely associated with the membrane, with another 5 micrometers scale bar shown. Heat shock-induced EVs associate with neighbouring naïve RBCs. Representative confocal fluorescence microscopy images of (A) 3D7 parasites maintained at 37 °C or (B) subjected to heat shock at 41 °C for 30 min. (C) Detail of an infected RBC showing small vesicles adjacent to the plasma membrane. (D) Representative images showing YAT2150-labelled EVs associated with the surface of naïve RBCs. In all cases, cells were stained with 0.5 µM YAT2150 (red) and nuclei counterstained with Hoechst 33342 (blue). Wild type-derived EVs confer protection to PfVps60KO parasites under heat shock stress Based on the marked heat shock sensitivity of PfVps60KO parasites detailed in Figure 3, and on the association of EVs enriched in aggregation-prone proteins with neighbouring erythrocytes following thermal stress, we next examined whether extracellular vesicles could functionally restore parasite survival under febrile conditions. For this, PfVps60KO parasites were exposed to heat shock in the presence of EVs derived from heat shocked 3D7 cultures, which were added immediately prior to and maintained throughout the heat shock period. Three independent EV preparations (s1, s2, and s3), containing decreasing EV concentrations (s1 > s2 > s3) but added at a fixed volume were tested. Supplementation with the most concentrated preparations (s1 and s2) significantly increased survival relative to the heat shocked PfVps60KO untreated control, whereas the least concentrated sample (s3) did not provide detectable protection (Figure 8). Moreover, an additional EV preparation (s4) derived from 3D7 parasites maintained at 37 °C, despite having high EV abundance, failed to confer a significantly protective effect under heat shock conditions. These results suggested that vesicles released by WT parasites under heat shock conditions contain factors that can partially restore parasite viability in a dose-dependent manner. 1× PBS used as a control in equal tested volumes did not exhibit any significant effect. Figure 8.Extracellular vesicles from the 3D7 line rescue PfVps60KO parasites from heat shock-induced death in a dose-dependent manner. Parasite survival following heat shock (41 °C, 2 h) was quantified by flow cytometry using Syto11. During heat stress, PfVps60KO parasites were supplemented with three EV preparations derived from heat shocked 3D7 parasites: s1, s2 and s3, as well as one derived from 3D7 parasites maintained at 37 °C (s4). Average EV concentrations are indicated in the x-axis labels (kcps/mL). All EV preparations were added at a fixed volume but contained different EV concentrations (s1 > s4 > s2 > s3). A control containing only PBS was included. Bars represent mean ± SD of three biological replicates. HS: heat shock. Statistical significance was calculated using one-way ANOVA. *p ≤ 0.05, ****p ≤ 0.0001.A bar chart showing PfVps60 knockout parasite survival after heat shock with PBS or 3D7 extracellular vesicle samples s1 to s4.The figure shows a vertical bar chart summarizing PfVps60 knockout parasite survival after heat shock with different extracellular vesicle treatments. The y axis label is percent survival, scaled from 0 to 100 with tick marks at 20, 40, 60, 80, and 100. The x axis label lists 6 conditions, written diagonally: KO HS, KO HS + PBS, KO HS + s1 (7.4k), KO HS + s2 (2.8k), KO HS + s3 (1.0k), and KO HS + s4 (4.1k). All bars include error bars showing standard deviation. Approximate bar heights are: KO HS 18 percent, KO HS + PBS 14 percent, KO HS + s1 62 percent, KO HS + s2 32 percent, KO HS + s3 20 percent, and KO HS + s4 28 percent. A horizontal bracket above the KO HS + s1 bar and its neighboring lower bars is annotated with four * symbols. A longer bracket spanning from KO HS to KO HS + s1 is annotated with one * symbol. All data are approximate. Extracellular vesicles from the 3D7 line rescue PfVps60KO parasites from heat shock-induced death in a dose-dependent manner. Parasite survival following heat shock (41 °C, 2 h) was quantified by flow cytometry using Syto11. During heat stress, PfVps60KO parasites were supplemented with three EV preparations derived from heat shocked 3D7 parasites: s1, s2 and s3, as well as one derived from 3D7 parasites maintained at 37 °C (s4). Average EV concentrations are indicated in the x-axis labels (kcps/mL). All EV preparations were added at a fixed volume but contained different EV concentrations (s1 > s4 > s2 > s3). A control containing only PBS was included. Bars represent mean ± SD of three biological replicates. HS: heat shock. Statistical significance was calculated using one-way ANOVA. *p ≤ 0.05, ****p ≤ 0.0001. Discussion In P. falciparum, proteostasis represents a major challenge even in the absence of external stress. The parasite’s atypical genome composition (approximately 81% of adenine (A) and thymine (T) residues [52]) results in an extreme A/T bias leading to an overrepresentation of asparagine and glutamine repeats, which constitute nearly 30% of the proteome [53] and are intrinsically prone to misfolding and aggregation [54,55]. This aggregation-prone nature has been confirmed through the isolation of 0.1% SDS-resistant aggregates from parasite cultures, as well as by staining with the commercial kit PROTEOSTAT®, which specifically detects protein aggregates in live cells [49]. In addition, the active ingredient of PROTEOSTAT®, YAT2150, synthesized in-house, stained aggregates formed in vitro by aggregation-prone peptides selected from the P. falciparum proteome [49,56], and in-cell throughout all the blood stages of the parasite [41]. These intrinsic features indicated that P. falciparum is particularly dependent on efficient protein quality control mechanisms. To cope with this aggregation-prone proteome and the recurrent temperature fluctuations experienced during infection, Plasmodium has evolved a highly effective heat shock defense system. More than 90 Hsp have been identified in the parasite, playing key roles during development, survival, and stress adaptation [57]. Unlike other eukaryotes, some Plasmodium Hsp are constitutively expressed at high levels [19,58,59], but also upregulated upon heat shock [58–60], hence combining house-keeping and stress response roles. Several Hsp are essential for parasite survival, especially members of the Hsp90, Hsp70, and Hsp40 families [61–63]. These chaperones can operate both in functional networks and independently with their corresponding co-chaperones [64–67]. Moreover, Hsp are widely distributed in all the parasite organelles and in the host, where they exhibit different functions. For instance, PfHsp70-1 (Q8IB24) is the major cytosolic chaperone expressed during the intraerythrocytic life cycle, it is upregulated upon heat shock and participates in the maintenance of proteostasis [45]. As occurs in higher eukaryotes, some Plasmodium Hsp are exported to the extracellular space through the endomembrane system, such as PfHsp70-x (K7NTP5) [68] and PfHsp40 (O96123) [69], which could participate in systemic stress responses. In this study, we show that disruption of PfVps60 impairs EV-associated export of the chaperones PfHsp70-x and PfHsp110. Both are known to assist in protein folding and to protect the parasite from the aggregation driven by its low-complexity proteome, as well as from temperature fluctuations experienced throughout its life cycle [19,66]. The absence of these chaperones from EVs derived from the PfVps60KO strain suggests that the ESCRT-mediated vesicular trafficking contributes to their export, either through EV biogenesis or cargo selection. This observation is consistent with our previous work showing that PfVps60 disruption reduces EV production [30]. The functional consequences of these defects are evident under thermal stress. In particular, induction of the major cytosolic chaperone PfHsp70-1, which peaked coinciding with nuclear division at the end of the trophozoite stage, was significantly weaker in heat-shocked PfVps60KO compared to heat-shocked WT parasites. Given that PfHsp70-1 is strongly transcriptionally induced during heat shock, with reported increases up to ∼16-fold mediated by the PfAP2-HS transcription factor [36], the reduced induction observed in PfVps60KO parasites may reflect both potential alterations in stress-responsive transcriptional programmes and impaired proteostasis. Consistent with this, YAT2150 staining showed an evident accumulation of aggregation-prone proteins near the parasitophorous vacuole in PfVps60-deficient parasites, which exhibited markedly impaired recovery from thermal stress. The redistribution of aggregates toward the parasitophorous vacuole suggests that defective vesicular export may contribute to local accumulation of misfolded proteins at sites normally involved in EV budding [70]. Complementary ThT measurements further revealed an early increase followed by a decrease in signal in PfVps60KO parasites after heat shock. Because ThT selectively detects β-sheet-rich species, this decline likely reflects remodelling of ThT-reactive aggregates and/or loss of severely damaged proteins, rather than efficient aggregate clearance. Previously, it has been proposed that EVs released in response to stress can trigger intercellular signalling in neighbouring cells while simultaneously enhancing their resilience to subsequent challenges (bystander effect) [21]. Although this specific condition was not directly tested here, our observations align with this model. We found that EV production is dynamically regulated in response to heat shock with a rapid increase shortly after stress exposure followed by a decrease 24 h post-stress, indicating that EV release constitutes an early and transient adaptive response rather than a sustained secretory process. It is important to note that EV quantification in this study is based on DLS-derived particle counts, which provide a relative rather than absolute measurement and do not directly assess vesicle purity. Similar stress-dependent modulation of EV cargo and abundance and cargo composition has been reported under nutrient starvation, where P. falciparum exhibited a rapid increase in EV release together with altered protein and small RNA composition, particularly in ring-stage cultures [71], supporting the notion that EV-mediated responses may represent a conserved mechanism in malaria parasites. Consistent with this model, EVs containing aggregation-prone cargoes were observed adhering to neighbouring uninfected erythrocytes following heat shock, indicating that stress-induced EVs can interact with surrounding cells. However, direct evidence of cargo transfer and functional uptake remains to be established, and the precise mechanisms governing EV trafficking, uptake, and cargo delivery between infected and uninfected erythrocytes require further investigation. Likewise, although our data reveal enrichment of aggregation-prone proteins and molecular chaperones in EVs, the specific contribution of individual cargo components is yet unresolved. Our findings suggest that the protective effect of EVs likely arises from the combined action of multiple stress-related factors, and dissecting the relative contribution of individual cargo components will require targeted functional approaches in future work. We provide functional evidence that EVs released by WT parasites under heat shock conditions can partially protect PfVps60-deficient parasites from severe thermal sensitivity in a dose-dependent manner when administered during the heat shock period. This experimental design aligns with a role for EVs in acute stress protection rather than preconditioning or post-stress recovery, since EVs obtained under non-stressed conditions fail to confer protection. These findings indicate that stress-induced EV cargoes are associated with improved parasite survival under heat shock conditions and may contribute to the restoration of proteostasis-related functions. The present study has some limitations that should be considered when interpreting the results. First, all experiments were performed using in vitro culture systems, which may not fully mimic the complexity of in vivo infection dynamics during clinical malaria. Second, our analyses were performed in a single parasite genetic background (3D7), and strain-specific differences in stress responses or EV biology cannot be excluded. Finally, while our findings support a role for EV-associated cargo in stress adaptation, in vivo validation will be required to establish the physiological relevance of this pathway during infection. Taken together, our results identify ESCRT-dependent EV trafficking as a key component of the heat shock response in P. falciparum, linking proteostasis, vesicular communication, and parasite survival under febrile conditions. EVs may contribute to the redistribution of aggregation-prone proteins and molecular chaperones during stress, as well as to coordinated adaptation within the parasite population. These findings suggest that EV biogenesis may represent a druggable vulnerability, raising the possibility that interfering with this pathway could sensitize parasites to febrile stress, and representing a novel strategy to disrupt Plasmodium survival during clinically relevant febrile episodes. Supplementary Material Supplemental Material