Nuclear import of malaria RNA rewires splicing in host immune cells

TL;DR

The malaria parasite uses its RNA to enter host immune cells' nuclei and disrupt RNA splicing, altering immune responses.

Contribution

Discovery of an RNA-based mechanism by malaria parasites to interfere with host splicing machinery and immune signaling.

Findings

Malaria mRNAs are imported into host nuclei and bind to splicing proteins ACIN1 and PNN.

This binding disrupts host RNA splicing, leading to altered immune protein expression.

The mechanism reveals a novel strategy for pathogen manipulation of host immunity.

Abstract

Eukaryotic pathogens deploy diverse strategies to manipulate host immunity, yet RNA-based virulence mechanisms remain poorly understood. We describe a mechanism by which the malaria parasite Plasmodiumfalciparum governs host immune responses through direct interference with host nuclear RNA processing. Malaria-encoded mRNAs of the early transcribed membrane protein family, exported from infected red blood cells, evade cytoplasmic degradation within host immune cells and are imported into their highly secured nuclei. Inside the nucleus, parasite transcripts bind the host RNA-binding proteins ACIN1 and PNN, key components of the splicing machinery that associate with the exon junction complex. This interaction disrupts host splicing regulation, leading to widespread misprocessing of transcripts and altered expression of proteins involved in immune function. Our findings uncover an RNA-based strategy by which pathogen-derived transcripts exploit the host splicing machinery, reshaping transcript isoform landscapes and immune signaling.

•Plasmodium falciparum delivers three parasitic ETRAMP mRNAs into host monocytes via EVs•Delivered RNAs evade degradation, enter the protected host nucleus, and bind ACIN1/PNN•Host splicing dysregulation causes transcript misprocessing and altered immune response•An extracellular RNA-based mechanism shows how malaria reshapes host immune signaling

Plasmodium falciparum delivers three parasitic ETRAMP mRNAs into host monocytes via EVs

Delivered RNAs evade degradation, enter the protected host nucleus, and bind ACIN1/PNN

Host splicing dysregulation causes transcript misprocessing and altered immune response

An extracellular RNA-based mechanism shows how malaria reshapes host immune signaling

Abou Karam et al. identify a cell communication mechanism used by the malaria parasite Plasmodium falciparum. The parasite delivers its mRNAs into monocytes and imports them into their protected nuclei, where they perturb the splicing. This extracellular RNA-based strategy disrupts host transcript processing and immune signaling, revealing another layer of host interference.

Figures

Splicing is altered in monocytes after treatment with *Pf*-EVs as revealed by RNA-seq analysis(A) Volcano plot of all AS events between *Pf*-EV-treated and untreated THP-1 cells. Shown are significant results based on Δ percent spliced in (PSI) (inclusion level difference), and significance (−log *p*) determined using rMATS-turbo.(B) Volcano plots of individual AS events (skipped exons [SE], retained introns [RI], alternative 3′ splice sites [A3SS], alternative 5′ splice sites [A5SS], and mutually exclusive exons [MXE]) between *Pf*-EV-treated and untreated THP-1 cells. Shown are significant r

Full text

Introduction

Malaria is the deadliest mosquito-borne parasitic disease, caused by various Plasmodium species. Plasmodium falciparum (Pf) is the most virulent species of Plasmodium and accounts for the majority of malaria cases.1 The Plasmodium parasite has a complex life cycle with three major developmental stages: the mosquito, the host liver, and the host blood stages.1 The blood stage is responsible for the clinical symptoms of malaria diseases.2 Pf parasites mature and replicate within the human red blood cells (RBCs). During the blood stage of the infection, monocytes control the parasite burden and contribute to host protection through phagocytosis, cytokine production, and antigen presentation.3^,^4 While numerous immune cell subsets participate in the response to malaria, recent clinical and experimental studies have highlighted the importance of monocytes in both protective and pathogenic responses, particularly in severe malaria syndromes such as cerebral and placental malaria.3^,^5 Notably, nonclassical monocytes have been linked to disease severity and poor outcomes in pediatric cerebral malaria,6^,^7 and transcriptional profiling has revealed persistent immune reprogramming in monocytes following Plasmodium infection.8 These findings underscore the relevance of monocytes as both effector and regulatory cells, making them a critical target of parasite-mediated immune modulation. Like many other pathogens, malaria parasites modify the functions of diverse host cells,9^,^10^,^11^,^12 including monocytes.13^,^14

Pathogens use extracellular vesicles (EVs) in a virus-like manner as an attack tool against the immune system.15^,^16^,^17 Pf secrete EVs to deliver parasite genomic DNA into host monocytes, activating the stimulator of interferon genes (STING) DNA sensing pathway13 and inducing the type I interferon (IFN) response. Parasitic EVs have also been shown to contain various types of small RNA molecules.10^,^13^,^18

mRNA splicing is a highly coordinated process that involves the precise assembly of ribonucleoprotein complexes to define exon-intron boundaries, ensuring accurate inclusion or exclusion of exons in the mature mRNA.19^,^20 Alternative splicing (AS) is tightly regulated and plays a pivotal role in expanding transcriptomic diversity, enabling physiological cell function and dynamic adaptation to environmental cues.21^,^22 Disruptions in splicing regulation can lead to aberrant transcript structures and have been implicated in a broad spectrum of diseases, including cancer and immune dysfunction.23 Emerging evidence indicates that pathogens can hijack the host splicing machinery as a mechanism to modulate host gene expression and subvert immune responses,24^,^25^,^26 highlighting AS as a critical interface in host-pathogen interactions.

Here, we discovered that Pf parasites deliver into host monocytes three mRNAs that encode members of the early transcribed membrane protein (ETRAMP) family, via EVs. Upon cell internalization, these parasitic transcripts are rapidly imported into the host nucleus, where they strongly bind two human splicing proteins ACIN1 and PNN. These two proteins act as a scaffold for a complex that regulates AS events in a manner that is both dependent on and independent of the exon junction complex (EJC).27 Sequestration of these proteins alters the splicing of host mRNAs, resulting in the downregulation of protein expression in the recipient monocytes and impairing their immune functions. Our work reveals a previously unknown host target of pathogens, the EJC components, which the parasite exploits to enhance its virulence.

Results

Parasitic mRNAs are encapsulated within Pf-EVs and internalized into human monocytes

It was previously demonstrated that the malaria parasite secretes EVs that contain RNA cargo,13^,^28 including small non-coding RNAs.18 Using fluorescent RNA labeling with spectral flow cytometer analysis, we demonstrated that Pf-EVs contain a higher amount of RNA than uninfected (u) RBC-EVs (Figures 1A, 1B, S1A, and S1B), and that external Pf-RNA was not detected upon EV pelleting (Figure S1C). One Pf transcript, the ETRAMP11.2 mRNA (PF3D7_1102800), was previously detected as a full-length transcript encapsulated within secreted Pf-EVs.13^,^14 This transcript encodes one of the 13 members of the ETRAMP family, which are unique to Plasmodium species.29 The ETRAMPs are structurally conserved membrane proteins, abundantly expressed at various stages of the blood cycle, that have unknown functions.29 RNA expression profiling showed stage-specific ETRAMP transcripts across the three blood stages (ring, trophozoites, and late trophozoites), although the majority of the seven ETRAMP transcripts examined were expressed at the earliest ring stage (Figure S2).Figure 1Pf-EVs contain mRNAs that are associated with ribosomes in host monocytes(A) Concentrations of lipophilic dye (R18) and RNA cargo (SytoRNA) for *Pf-*EVs and uRBC-EVs (n = 3, color coded, ^∗^p < 0.05, two-tailed, unpaired t test); the dotted lines connect uRBC-EVs and Pf-EV samples from the same experiment. The data are also shown as fold changes (Pf/uRBC EVs), with uRBC-EVs values normalized to an average of one.(B) Spectral signatures of unlabeled and labeled Pf-EVs acquired on spectral flow cytometer.(C) Illustration demonstrating the three enriched ETRAMPs encapsulated within Pf-EVs created with BioRender.(D) qPCR quantification of eleven ETRAMP mRNAs in Pf-EVs. Displayed are the Ct values; dots represent the six biological replicates. The dotted line indicates the threshold above which gene expression is considered undetectable (Ct = 37).(E) qPCR quantification of ETRAMP11.2, ETRAMP2, and ETRAMP14.1 in THP-1 cells treated with Pf-EVs for 1, 4, 16, and 24 h. Expression levels were normalized to the HPRT housekeeping gene and to the 1 h time point. Dots represent the three biological replicates. The dashed line indicates the baseline expression at 1 h, used as the reference for comparison across time points.(F) Polysome profiles of untreated THP-1 cells and THP-1 cells treated with Pf-EVs for 4 h. Shown are data from one experiment representative of three biological repeats.(G) qPCR analysis of the HPRT and ETRAMP11.2 mRNAs in the free (untranslated transcripts), light (transcripts translated at an intermediate level), and heavy (highly translated transcripts) polysomal fractions of THP-1 cells treated with Pf-EVs or untreated (UT) for 4 h (n = 3), ^∗^p < 0.05. The housekeeping gene HPRT served as a control. Error bars indicate SEM between the biological replicates.

To examine whether Pf-EVs carry other transcripts of the ETRAMP family, EVs derived from Pf-infected RBCs were harvested as previously described.13 EVs were of the expected purity, concentration (Figure S3A), size, and morphology (Figure S3B). The RNA cargo was extracted from the EVs and subjected to a qPCR analysis for the eleven ETRAMP genes. Five ETRAMP transcripts, ETRAMP2, ETRAMP14.1, ETRAMP11.2, ETRAMP11.1, and ETRAMP4 were detected within the Pf-EVs (Figures 1C and 1D). The three most abundant ETRAMPs, ETRAMP11.2, ETRAMP2, and ETRAMP14.1 (PF3D7_1102800, PF3D7_0202500, and PF3D7_1401400, respectively) were chosen for further analysis.

Since Pf-EVs are taken up by human monocytes,13^,^28^,^30 we sought to determine whether these three ETRAMP mRNAs are detectable within host immune cells and, if so, how stable they are post-EV internalization. THP-1 cells, which are a human monocyte line, were treated with Pf-EVs, and the vesicle uptake was measured using imaging flow cytometry as previously described.28^,^31 This experiment showed that more than 95% of cells internalized Pf-EVs (Figure S4). Total RNA was extracted from the recipient cells, and ETRAMP mRNAs were quantified using qPCR. The three mRNAs ETRAMP11.2, ETRAMP2, and ETRAMP14.1 were most abundant at 1 h and showed a modest decrease by 4 h but could not be detected by 16 h post-internalization (Figure 1E). The fact that these transcripts remained stable for several hours in the host monocytes suggests that they may have a functional role within the host cells or might be associated with host machinery, such as ribosomes.

Pf mRNAs associate with polysome fractions in host monocytes

The translation machinery is one of the primary targets by which pathogens can affect the physiology of their host.32 We thus set out to examine whether Pf-EV internalization alters translation in monocytes. We used a ribosome pull-down assay to detect mRNA association with ribosomes.33^,^34 The assay uses sucrose density fractionation to separate “free” RNA and monosomes from polysomes, and the relative distributions of mRNAs are determined in these fractions.35^,^36 Cell lysates derived from Pf-EV-treated and untreated THP-1 cells were fractionated according to the ribosome densities using a sucrose gradient. Similar ribosome profile patterns were obtained for the EV-treated and untreated monocytes (Figure 1F), indicating that the parasitic EVs did not significantly alter global translation in recipient cells. We reasoned that the translation of only a subset of proteins might be affected, and therefore, we examined the association of the three ETRAMP transcripts with the different polysome fractions. We found that more than 50% of ETRAMP2, ETRAMP14.1, and ETRAMP11.2 mRNA transcripts were associated with light and heavy fractions (Figures 1G and S5), suggesting that these parasitic transcripts might be translated within host monocytes.

As an unbiased complementary approach, we used pulsed stable isotope labeling with amino acids in cell culture (SILAC) proteomics. SILAC is a high-throughput approach that distinguishes existing from newly synthesized proteins through metabolic labeling with lysine and arginine that contain ^13^C and ^15^N isotopes.37 Proteins that are synthesized post the exposure are labeled with “heavy” atoms, whereas proteins synthesized prior to that time contained only “light” amino acids. The SILAC analysis showed that only a few Pf proteins were identified, each by only a single peptide, and no ETRAMP proteins were detected. Thus, although ETRAMP mRNAs were associated with ribosomes, definite proof of translation was not obtained.

Pf mRNAs are imported into host monocyte nuclei

As no ETRAMP proteins could be detected post EV internalization, we investigated the fates of the ETRAMP mRNAs within the host monocytes. To explore the subcellular destinations of the three parasitic ETRAMP transcripts within monocytes, we performed a nuclear and cytosolic biochemical fractionation assay, as previously described.38 The purities of subcellular fractions were evaluated by western blot analysis using five known markers: nuclear lamin B2,39 cytosolic GAPDH,40 early endosomal Rab5, late endosomal Rab7,41 and endoplasmic reticulum calnexin (Figure 2A).42 RNA was extracted from the two fractions, and ETRAMP11.2, ETRAMP2, and ETRAMP14.1 were quantified in each fraction. The three ETRAMP mRNAs were detected in both the nuclear and cytoplasmic fractions (Figures 2B and S6). These findings suggest that malaria transcripts are actively internalized into the nuclei of the host human monocytes.Figure 2Pf mRNAs are rapidly internalized into host cell nuclei and are bound by endogenous RNA-binding proteins(A) Western blot analyses of nuclear and cytoplasmic fractions from Pf-EV-treated and untreated THP-1 cells for markers of the nucleus (lamin B2), cytoplasm (GAPDH), early endosomes (Rab 5), late endosomes (Rab 7), and endoplasmic reticulum (calnexin).(B) Representative qPCR analysis of nuclear and cytoplasmic fractions from Pf-EV-treated and untreated THP-1 cells at the indicated time points. Relative transcript levels were calculated using the standard curve method and normalized to the levels of U2 small nuclear RNA for the nuclear fraction, RPS11 for the cytoplasmic fraction, and to the 0.5 h time point. Shown is a qPCR analysis representative of three biological replicates.(C) Representative bright field (BF) and fluorescence confocal microscopy images of Pf-EV-treated and untreated THP-1 cells using HCR smFISH probes designed to identify ETRAMP11.2 (red) and ETRAMP2 (yellow). Hoechst was used to stain nuclei (blue), and carboxyfluorescein succinimidyl ester (CFSE) (green) was used to stain the cells. Scale bars represent 10 μm.(D) Quantification of the HCR smFISH probe signals (i.e., number of dots) using semiautomated Imaris software. Following segmentation of nuclei, cytoplasm, and FISH spots, the proportion of nuclear signal was calculated, with the cytoplasmic proportion defined as one minus the nucleus proportion. Statistical significance was assessed by comparing the nuclear proportions to an expected mean of 0.5 using a one-sample t test. Data represent the mean of three independent biological experiments; (ETRAMP11.2, p = 2.82 × 10^−7 and^ ETRAMP2, p = 6.26 × 10^−4^); ^∗∗∗^p < 0.001 and ^∗∗∗∗^p < 0.0001.(E) Representative confocal microscopy images of Pf-EV-treated and untreated THP-1 cells using HCR smFISH probes designed to identify human β-actin mRNA, which served as positive host RNA control (green) and ETRAMP11.2 mRNA (red). Yellow arrows indicate ETRAMP11.2 in the nuclei of the monocyte. Hoechst was used to stain nuclei (blue). Scale bars represent 10 μm.(F) THP-1 cells were treated with Pf-EVs for 1 h. After UV crosslinking or not, biotinylated probes against ETRAMP11.2 mRNA were added to lysates, and the probes and associated mRNA and protein were precipitated with streptavidin-coated magnetic beads. Proteins were subjected to label-free quantitative mass spectrometry. Volcano plot of differentially expressed proteins precipitated in cells subjected to UV-mediated crosslinking and unirradiated cells. The results presented were significant based on a t test with permutation-based FDR calculation by log2 fold change (difference) and significance (−log p) using the Perseus software. Protein quantification was based on four independent biological experiments.

Cells have barriers to block the transport of pathogen RNAs across the nuclear envelope.43 We therefore sought to confirm this subcellular localization of ETRAMP transcripts at a single-cell resolution. We established hybridization chain reaction (HCR), single-molecule fluorescence in situ hybridization (smFISH) assays for the two Pf mRNAs most enriched in EVs ETRAMP11.2 and ETRAMP2, as previously described for other RNAs.44^,^45^,^46 This smFISH technique provides exceptional signal amplification capabilities, crucial for detecting extremely low RNA signals.45 THP-1 cells were treated with Pf-EVs, and probes were introduced into the treated cells. To enable differential signal quantification, dyes to distinguish the nucleus and the cytoplasm were also introduced. Remarkably, both ETRAMP11.2 and ETRAMP2 mRNAs were detected in the nuclei of the host recipient cells at 1 h post EV internalization (Figure 2C). Quantification of the smFISH signal revealed that at least 60% of each Pf transcript localized to nuclei (Figure 2D). Probes designed against human β-actin were used as a positive control. In contrast to the ETRAMP mRNAs, β-actin mRNA was localized solely to the host cytoplasm in both untreated and Pf-EV-treated cells (Figure 2E). These data indicate that Pf mRNAs are rapidly imported into the host cell’s nucleus after EV internalization.

Pf ETRAMP mRNAs bind to host nuclear RNA-binding proteins

To study the function of the parasitic RNAs within the host nucleus, we sought to determine whether the parasitic ETRAMP transcripts interact with host proteins once they are imported into the nucleus. We therefore used RNA antisense purification (RAP) assay47^,^48 to examine the proteins bound to ETRAMP mRNAs within the recipient cells. This approach also includes an ultraviolet (UV) light to covalently crosslink protein and RNA molecules in close proximity.48 Biotinylated probes complementary to the ETRAMP11.2 mRNA—the most abundant ETRAMP transcript within Pf-EVs—were used to precipitate this mRNA and associated proteins. The control sample was not subjected to UV. Mass spectrometry analysis identified 239 proteins, most of which were, as expected, known RNA-binding proteins (Figure 2F). Remarkably, only two proteins were differentially precipitated in the sample subjected to crosslinking with UV irradiation: human ACIN1 and human PNN.

ACIN1 and PNN are nuclear RNA-binding proteins associated with the EJC during mRNA splicing.49 The EJC is a protein complex that assembles on pre-mRNA during splicing and plays a role in mRNA export, localization, translation, and nonsense-mediated mRNA decay.50 The EJC remains attached to the mature mRNA until it is removed by translocating ribosomes. These two human proteins act as scaffolds for two different apoptosis- and splicing-associated protein complexes: ACIN1 with the ASAP complex and PNN with the PSAP complex.51 Transcriptome-wide RNA sequencing (RNA-seq) studies have shown that the ASAP and PSAP complexes regulate AS events in both EJC-independent and EJC-dependent manner.27 The endogenous RNA targets of PNN are still unknown, but the targets of ACIN1 have been identified.52 ACIN1 preferentially binds two motifs: AT-rich sequence (TTTTTTTT) and a GAAGAA-like motif.52 However, currently, there is not much evidence that pathogens specifically target this particular host pathway. Notably, our computational analysis identified the two ACIN1 RNA-binding motifs within the three ETRAMP transcripts (Figures S7A and S7B).

Pf-EVs induce alterations in the splicing of mRNAs in monocytes

We hypothesized that the association of ACIN1 protein with ETRAMP11.2 may lead to disruptions in the splicing or stability of host target mRNAs, subsequently resulting in interrupted translation of certain host proteins. To test this, we explored our data from SILAC analyses of Pf-EV-treated and untreated THP-1 cells (Figure 3A). This analysis demonstrated significant reductions in 152 host proteins following Pf-EV internalization as compared to untreated cells (Figures 3B, 3C, and S8A; Table S3). Analysis of protein-protein interaction networks showed that the proteins downregulated by incubation of monocytes with Pf-EVs were enriched in pathways that include nucleic acid metabolic process, RNA processing, ribosome biogenesis, and mitochondrion organization. The 33 most differentially downregulated proteins had enrichment in organelle organization, nitrogen compound metabolic processes, and metabolic processes.Figure 3Pf-EV treatment changes the proteomic profile of monocytes(A) Plot of differential protein expression in THP-1 cells treated with Pf-EVs for 16 h versus control untreated cells (UT). The x axis is the ratio between the amount of a protein detected in the existing (“light” isotope) pool and that in the newly synthesized (“heavy” isotope) pool (H/L) of the Pf-EV-treated samples and the H/L of the untreated sample (log2). The y axis represents the fold change between the total intensities of the two groups (log2). Data are the averages of the three repeats, and missing values were replaced by the minimal intensity in the experiment. In the upper right quadrant, the newly synthesized proteins elevated post Pf-EV treatment that were also elevated in the total intensity are marked in red. In the lower left quadrant, the newly synthesized proteins decreased post Pf-EV treatment, which were also decreased in the total intensity, are marked in blue.(B) Network interactions of proteins downregulated in THP-1 cells upon Pf-EV treatment as compared to untreated cells by search tool for the retrieval of interacting genes/proteins (STRING) analysis.(C) Gene Ontology (GO) annotation enrichments of human downregulated proteins in THP-1 cells upon Pf-EV treatment as compared to untreated cells. Count refers to the number of proteins that were annotated with a particular term.(D) Quantification of indicated mRNA levels in THP-1 cells treated with Pf-EVs for 1, 2, 3, 4, and 16 h. HPRT served as a control. Data are presented as the fold induction over untreated control (UT – 1 h) and are means and standard errors of the means of three biological replicates. ΔCT values per gene were compared with a linear model (two-way ANOVA) testing the effect of treatment, time, treatment and time interaction, and batch (RHOT2, p = 0.0049; TEX264, p = 0.0460; SCLY, p = 0.0176; and FN3KRP, p = 0.0411), ^∗^p < 0.05 and ^∗∗^p < 0.01.

Notably, approximately 30% of the most downregulated proteins (highlighted in red) identified through SILAC proteomics post Pf-EV internalization (Table S3, highlighted in orange) represent mRNAs that are known targets of ACIN1,52 including RHOT2, NHLRC2, CNOT3, PCYT2, TEX264, FN3KRP, ZWINT, TPX2, SCLY, and ATAD2.52 To test the possibility that levels of endogenous mRNA targets of ACIN1 are affected by Pf-EV internalization, we treated cells with Pf-EVs for various time points, extracted RNA, and performed qPCR analysis. Of the 12 mRNAs examined (ten ACIN1 targets and two controls), levels of RHOT2, TEX264, SCLY, and FN3KRP were significantly reduced at 3 h post Pf-EV treatment compared to the levels in control cells (Figure 3D and S8B). Previous work has shown that ACIN1 depletion increases intron retention and causes AS alterations53^,^54; therefore, it is plausible that ETRAMP11.2 acts as a decoy by binding ACIN1, thereby sequestering it from its endogenous mRNA targets. This interaction could suppress translation of the host proteins in a titration-like fashion or induce structural modifications in the mRNA without altering its overall expression levels.

To further investigate the impact of ACIN1/PNN-ETRAMP interaction on host AS patterns, THP-1 cells were treated with Pf-EVs or were left untreated for 3 h followed by RNA extraction and transcriptome-wide sequencing. Differential AS was analyzed using rMATS-turbo55 by comparing transcriptomic profiles from THP1-treated and untreated conditions, each represented by three biological replicates. The analysis identified 94 splicing events across the five canonical AS types (Figure 4A). Among these, skipped exons (SE) appeared most frequently, with 38 events, followed by retained introns (RI) (35 events), alternative 3′ splice sites (A3SS) (13 events), alternative 5′ splice sites (A5SS) (7 events), and mutually exclusive exons (MXE) (6 events) (Figures 4B and 4C; Table S5). These results suggest that exposure to Pf-EVs induces splicing alterations, with exon skipping modulation emerging as the dominant splicing event, potentially due to direct interference of the ACIN1/PNN complexes by ETRAMP binding.Figure 4. Splicing is altered in monocytes after treatment with Pf-EVs as revealed by RNA-seq analysis(A) Volcano plot of all AS events between Pf-EV-treated and untreated THP-1 cells. Shown are significant results based on Δ percent spliced in (PSI) (inclusion level difference), and significance (−log p) determined using rMATS-turbo.(B) Volcano plots of individual AS events (skipped exons [SE], retained introns [RI], alternative 3′ splice sites [A3SS], alternative 5′ splice sites [A5SS], and mutually exclusive exons [MXE]) between Pf-EV-treated and untreated THP-1 cells. Shown are significant results based on ΔPSI (inclusion level difference), and significance (−log p).(C) Raw counts of each AS event.(D) KEGG pathways enrichment analysis of AS events from rMATS between Pf-EV-treated and untreated THP-1 cells.(E) Detailed sashimi plots depicting the exon skipping events of ENOPH1 and KEAP1 transcripts. ΔPSI was calculated as the ratio between the PSI of Pf-EV-treated cells and that of untreated THP-1 cells.(F) Volcano plot of differentially expressed genes between Pf-EV-treated and untreated THP-1 cells. Shown are significant results based on log2 fold change (difference), and significance (−log p) determined using DEseq2 (version 1.38.3). Quantification was based on two independent biological experiments.(G) Network interactions of upregulated genes in THP-1 cells upon Pf-EV treatment as compared to untreated cells by STRING analysis.(H) qPCR analysis of the upregulated genes in Pf-EV-treated versus untreated THP-1 cells. HPRT served as a housekeeping control gene. Error bars indicate SEM between the three biological replicates. ΔCT values per gene were compared with a t test (CCL4, p = 0.009; CXCL10, p = 0.022; ICAM-1, p = 0.009; CCL3, p = 0.056; and IL-1β, p = 0.007); ^∗^p < 0.05 and ^∗∗^p < 0.01.

To gain preliminary insight into the potential functional consequences of these splicing changes, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis was performed on the set of genes associated with high-confidence events (false discovery rate [FDR] < 0.1 and |ΔPSI| > 0.1). The analysis revealed a significant overrepresentation of terms related to RNA metabolism, most notably genes encoding the proteins associated with the spliceosome (KEGG: hsa03041, adjusted p = 2.81 × 10^−4^), RNA transport (KEGG: hsa03013, adjusted p = 0.00568), and mRNA surveillance pathways (KEGG: hsa03015, adjusted p = 0.0216) (Figure 4D). Among the significant AS events, we identified several exon skipping alterations, for instance, affecting KEAP1 and ENOPH1 transcripts (Figure 4E). In KEAP1 (ENSG00000079999), Pf-EV-treated samples exhibited higher exon inclusion level compared to untreated samples (PSI_treated = 1.00; PSI_untreated = 0.59 on average, n = 3) (Figure 4E). Conversely, ENOPH1 (ENSG00000145293) exhibited enhanced exon skipping upon Pf-EV treatment, with a marked decrease in inclusion levels (PSI_treated = 0.49 on average; PSI_untreated = 1.00, n = 3), corresponding to a negative inclusion level difference (Figure 4E).

While AS analysis revealed notable transcriptomic remodeling in THP-1 cells following Pf-EV treatment, differential gene expression analysis (DESeq2) showed minimal global changes in gene expression levels. Nevertheless, a specific subset of immune response genes was significantly upregulated (Figure 4F). These genes are associated with defense response to other organisms, cytokine-mediated signaling pathways, and the TNF signaling pathway (Figure 4G). qPCR analysis confirmed the upregulation of key immune-related genes, including CCL4, CXCL10, ICAM-1, CCL3, and IL-1β, following exposure to Pf-EVs (Figure 4H). In summary, these results demonstrate that the presence of Pf mRNAs in the host nucleus results in widespread reprogramming of host splicing patterns.

Pro-inflammatory responses in host monocytes and macrophages induced by Pf-EVs

Given the critical role monocytes play during malaria infection and considering that Pf RNAs disrupt splicing (Figure 4A) and dysregulate the expression of many host proteins (Figures 3A–3C), we speculated that internalization of Pf-EVs might alter the decision-making processes of recipient immune cells. Analysis of SILAC data revealed that proteins upregulated after Pf-EV internalization were enriched in pathways associated with the innate immune response, type I IFN response, and defense response to RNA viruses, all categorized as pro-inflammatory responses. Thirty proteins were significantly upregulated in both intensities and ratios or were exclusively found in the EV-treated group, with 19 of these associated with immune system functions (Figures 5A and 5B; Table S4). For example, MX2 and OAS2 (defense response to virus) and ICAM-1, IFIT3, and STAT1 (type I IFN signaling pathway) were significantly upregulated (Figures 5A, 5B, and S8C). The enhanced expressions of ICAM-1, OAS2, and MX2 were validated in THP-1 cells post Pf-EV internalization using western blot analysis (Figure S8D). Elevation was significant for ICAM-1 and OAS2.Figure 5Pf-EVs induce a pro-inflammatory response in monocytes(A) Network interactions of proteins upregulated in THP-1 cells upon Pf-EV treatment as compared to untreated cells by STRING analysis.(B) GO annotation enrichments of human upregulated proteins in THP-1 cells upon Pf-EV treatment as compared to untreated cells. Count refers to the number of proteins that were annotated with a particular term.(C) Relative expression of CD80, TNF-α, IL-6, and IL-12 mRNAs in MDMs. Monocytes isolated from healthy donors were treated with M-CSF only; with M-CSF plus Pf-EVs; or with M-CSF, IL-4, and IL-10. Cells were harvested, and RNA was extracted and subjected to qPCR analysis. GAPDH served as a housekeeping control gene. Data are presented as the fold induction over untreated control (M-CSF – 6 h) and represent the mean and standard error of the mean of three biological replicates. ΔCt values were compared using an ANOVA model, accounting for treatment, time, and batch (CD80: Pf-EVs p = 0.0003 and IFN-γ [both] p < 0.0001; IL-6: Pf-EVs p = 0.0002 and IFN-γ [both] p < 0.0001; IL-12: Pf-EVs p = 0.0223 and IFN-γ [both] p < 0.0001; and TNF-α: Pf-EVs p = 0.0562, IFN-γ [6 h] p < 0.0001, and IFN-γ [24 h] p = 0.0004); ^∗^p < 0.05, ^∗∗^p < 0.01, ^∗∗∗^p < 0.001, and ^∗∗∗∗^p < 0.0001.(D) Model of Pf mRNA nuclear localization and binding by host cell proteins ACIN1 and PNN, leading to splicing alterations, endogenous protein downregulation, and upregulation of the pro-inflammatory response in host monocytes.Illustration created with BioRender.

To confirm the relevance of the effects detected in THP-1 cells in primary immune cells, we examined the response of primary monocyte-derived macrophages (MDMs) to Pf-EVs. Monocytes isolated from peripheral blood mononuclear cells from three healthy donors were differentiated into macrophages in culture. Subsequently, MDMs were treated with Pf-EVs, and RNA was extracted and subjected to qPCR analysis to assess whether the MDM response was skewed toward a pro-inflammatory or an anti-inflammatory phenotype. A strong pro-inflammatory response was observed in treated MDMs, with significant upregulation of mRNAs encoding *IL-*6, IL-12, and CD80 (Figure 5C). For TNF-α mRNA, an upregulation trend was observed, with a 3-fold increase (Figure 5C, p = 0.056). As controls, stimulation with IFN-γ and lipopolysaccharide (LPS) was employed. A pro-inflammatory response was observed in MDMs stimulated with IFN-γ and LPS (Figure 5C). To conclude, Pf-EV internalization stimulated the expression of various pro-inflammatory cytokines and chemokines indicative of a pro-inflammatory response in both monocytes and primary MDMs. This probably benefits the malaria parasite, as was shown for other pathogens.56^,^57^,^58

Our findings reveal a previously unknown virulence strategy employed by the malaria parasite mediated by extracellular RNA (Figure 5D).

Discussion

The intricate dynamic between pathogens and their hosts presents a formidable obstacle to studying and treating infectious diseases. Eukaryotic pathogens have remarkable adaptability and employ diverse mechanisms, such as the secretion of vesicles, to facilitate advantageous cell-to-cell communication.15^,^59^,^60

Secreted EVs carry not only proteins61^,^62 but also small RNAs.63 However, only a few studies have addressed the question of whether EVs also carry longer RNA species64^,^65^,^66 or full-transcript mRNAs.63^,^67^,^68 The RNAs carried by pathogenic EVs, including secreted by bacterial cells69^,^70 and cells infected by parasites,13^,^14^,^16^,^18^,^56 have been demonstrated to influence host cells in various ways, including enabling immune system evasion. Interestingly, pathogen mRNAs carried by EVs are translated in host cells during tuberculosis infection69 and in fungal infection of Arabidopsis plants.71 In the case of tuberculosis, the translated mRNA manipulates host monocytes,69 whereas in the case of Arabidopsis, the mRNA transported by the EV limits fungal infection.71

By capturing the molecular signatures exchanged between host and parasite, EVs from clinical malaria patients offer unparalleled opportunities for biomarker discovery72^,^73^,^74 and the rational design of EV-based vaccines.75 In Plasmodium vivax infections, plasma-derived EVs interact with human spleen fibroblasts and induce ICAM-1 via NF-κB translocation, suggesting a mechanism for splenic sequestration and contributing to disease severity.72

Metabolites, ions, and molecules smaller than ∼40 kDa can freely traverse the nuclear envelope, but macromolecules like proteins, mRNAs, tRNAs, ribosome subunits, and viruses must be actively transported through the nuclear pore complex.76^,^77^,^78 The RNAs of certain RNA viruses, including HIV-1,79 influenza A virus,80 and hepatitis delta virus,81 are imported into host cell nuclei. In these examples, the pathogens use their own machinery to transport RNA into the nucleus of the host cell.79^,^81

RNA-binding proteins are critical for protecting mRNAs from degradation and orchestrating their functions.82^,^83 In humans, only a handful of pathogens, predominantly viral, have been shown to manipulate the host splicing process to their advantage.24^,^84^,^85 Among these, the West Nile virus was suggested to interfere with the EJC and subsequently the nonsense-mediated RNA decay process to protect viral RNA from decay.85

Splicing dysregulation can influence gene expression at multiple, interconnected levels. Altered isoforms, even when stable, may encode truncated or dysfunctional proteins that compromise cellular function. Other splice variants may evade degradation but be inefficiently exported or translated, thereby reducing protein abundance without major changes in steady-state mRNA levels.22^,^50 Our findings indicate that parasite-derived RNAs perturb the splicing regulators ACIN1 and PNN, shifting isoform production toward transcripts that are inherently unstable or translationally compromised. This model provides a mechanistic explanation for the limited transcriptomic changes we observed in parallel with more pronounced alterations in protein abundance.

The extent of splicing dysregulation observed upon Pf-EV internalization was further supported by KEGG pathways enrichment analysis, with the top enriched terms for affected genes pointing to direct disruptions in core splicing mechanisms and RNA surveillance pathways. These alterations broadly impact RNA metabolism, reinforcing a model in which malaria-derived ETRAMP transcripts rewire the host cell’s RNA processing landscape. Meta-computational analysis combining host gene expression and protein translation datasets upon Pf-EV introduction identified genes that were significantly impacted at both levels in short- and long-term responses Figure S9). Integration of RNA-seq and SILAC datasets (3 h and 16 h post-EV treatment, respectively) using the R software identified the most significantly upregulated and downregulated transcripts and proteins. Among the genes most significantly dysregulated at both the RNA and protein levels, ZWINT (UniProt: O95229) and NDUFS7 (UniProt: O75251) were most prominently affected by EV uptake. For ZWINT, the data suggest that increased transcription at 3 h post-EV treatment may lead to a corresponding increase in translation at 16 h post-treatment. Interestingly, for NDUFS7, we observed elevated transcription at 3 h post-treatment but reduced translation at 16 h post-EV treatment. This discordant regulation of NDUFS7, with increased mRNA levels but reduced protein levels, may suggest post-transcriptional/translation control mechanisms following EV uptake for this target.

EVs have emerged as key mediators of host-pathogen interactions,86^,^87 carrying a diverse array of molecular cargo that can modulate the immune system. Their contents include not only RNA10^,^88 but also DNA13^,^89 and proteins.11^,^90^,^91 Thus, we cannot rule out the possibility that additional cargo components, independent of ETRAMP, contribute to the splicing changes observed in host cells. Furthermore, while our RAP assay demonstrates strong binding of ETRAMP transcripts to ACIN1 and PNN, proteins associated with the EJC, it remains possible that these specific ETRAMPs are not the sole drivers of the observed splicing dysregulation and immune response but rather act in concert with other EV-derived parasitic components.

Parasitic EVs have also been shown to promote cytokine release and antigen presentation, thereby enhancing both innate and adaptive immunity, while also contributing to immunopathology.60^,^92 Moreover, in some cases, these EVs appear to preferentially target immune cells more effectively than non-pathogenic control EVs. For example, in Plasmodium vivax malaria, EVs derived from patient plasma interact more extensively within the spleen with human T cells, monocytes, and B cells, as compared to EVs from healthy donors.93 Together, these studies emphasize the multifaceted role of EVs as powerful modulators of immune responses at the host-pathogen interface.

Limitations of the study

In this study, we investigate the distribution of EV RNA cargo, focusing on three transcripts secreted by Pf-infected RBCs and delivered to host monocytes. Our analysis is limited to these three transcripts rather than the full EV RNA cargo; we cannot infer the fate/function of other EV RNA species or whether additional RNAs also enter the nucleus. While sashimi plots illustrate splice junction usage and overall exon coverage, reflecting aggregated read support rather than replicate-level statistical estimates such as those derived from rMATS, they can yield visual impressions that are not fully concordant with the quantified PSI estimates.

Resource availability

Lead contact

Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Neta Regev-Rudzki (neta.regev-rudzki@weizmann.ac.il).

Materials availability

This study did not generate new, unique reagents.

Data and code availability

• •The datasets generated during this study are available from the corresponding author upon reasonable request. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PRIDE: PXD054139.
• •The RNA-seq data have been deposited in the BioProject database with the dataset identifier BioProject: PRJNA1258730. Accession number is GEO: [GSE247350](GSE247350). https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE247350.
• •This paper does not report original code.

Acknowledgments

We thank the Malaria Research Reference Reagent Resource Center (MR4) for their generous supply of parasite strains. We thank the Crown Genomics Institute and the Mantoux Bioinformatics Institute of the Nancy and Stephen Grand Israel National Center for Personalized Medicine at Weizmann Institute of Science for their work on RNA-seq and analysis. Illustrations were created with BioRender.com. N.R.-R. is funded by the European Union (ERC, MalChemAtlas, 101086598). Views and opinions expressed are, however, those of the authors only and do not necessarily reflect those of the European Union or the European Research Council. Neither the European Union nor the granting authority can be held responsible. This research (grant no. P141807 ) was carried out with the aid of a grant from the Canadian Institutes of Health Research (CIHR), the International Development Research Centre (IDRC), the Israel Science Foundation (ISF), and the Azrieli Foundation. Z.M. is supported by the ISF (1033/21). N.R.-R. is deeply grateful for the support of the Henry Chanoch Krenter Institute for Biomedical Imaging and Genomics, the Dr. Barry Sherman Institute for Medicinal Chemistry, the Karen Siem Fellowship for Women in Science, the Henri Gutwirth Award, the Brenden-Mann Women’s Innovation Impact fund, and the Pasteur-Weizmann joint research program.

Author contributions

Conceptualization, P.A.K., E. Kiper, T.Z., E. Kozela, R.N., D.A., A.C.C., Y.L., R.R., E.W., M.C., I.R.-G., E.P.-K., Z.P., O.S., I.U., C.L., Z.M., and N.R.-R; methodology, P.A.K., E. Kiper, T.Z., S.Y., E. Kozela, N.Z., R.N., D.A., H.O., A.C.C., I.R.-G., and E.P.-K.; investigation, P.A.K., E. Kiper, T.Z., S.Y., E. Kozela, N.Z., R.N., D.A., H.O., and A.C.C.; writing – original draft, P.A.K., E. Kiper, Z.P., T.Z., R.R., E.W., Z.M., and N.R.-R.; writing – review & editing, P.A.K., E. Kiper, Z.M., and N.R.-R.; funding acquisition, N.R.-R.; and supervision, Z.M. and N.R.-R.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesRabbit monoclonal anti-Lamin B2Cell Signaling TechnologyCat# D8P3URabbit polyclonal anti-CalnexinAbcamCat# ab22595Rabbit polyclonal anti-Rab5AbcamCat# ab218624Rabbit polyclonal anti-Rab7AbcamCat# ab137029Mouse monoclonal anti-GAPDH (6C5)AbcamCat# ab8245Mouse monoclonal anti-ICAM-1Santa Cruz BiotechnologyCat# sc-8439Mouse monoclonal anti-Mx2Santa Cruz BiotechnologyCat# sc-271527Mouse monoclonal anti-OAS2Santa Cruz BiotechnologyCat# sc-271117Goat anti-rabbit IgG-HRPAbcamCat# ab6721Goat anti-mouse IgG-HRPAbcamCat# ab6789Biological samplesHuman A^+^ erythrocytesLocal blood bankN/AChemicals, peptides, and recombinant proteinsTrizol Tri reagentBio-Lab Ltd.Cat# 9010233100CycloheximideSigma-AldrichCat# 239763-MSodium deoxycholate (DOC)Sigma-AldrichCat# D6750DTTBio BasicCat# DB0058Tris-HClSigma-AldrichCat# T5941KClEMD Merck MilliporeCat# 7447-40-7TritonSigma-AldrichCat# X100Acetic acidSAFCCat# ARK2183LiClSigma-AldrichCat# L7026DDMSigma-AldrichCat# D4641PBS-Tween20Fisher BioreagentsCat# BP337-500TURBO DNaseThermo Fisher ScientificCat# AM2238RNase inhibitorEurxCat# E4210Protease inhibitor cocktailAPExBIOCat# K1007SDSBio-Lab Ltd.Cat# 19812323Poly-D-lysineGibcoCat# A38904-01PFABar NaorCat# BN15710Saline Sodium Citrate buffer (SSC)Thermo Fisher ScientificCat# AM9763Thiazole orangeSigma-AldrichCat# 390062Hoechst 33342Sigma-AldrichCat# B2261CFSESigma-AldrichCat# 2188Rhodamine B octadecyl ester perchlorate (R18)Sigma-AldrichCat# 83685Sera-mag magnetic beadsCytivaCat# 29343052Streptavidin-coated magnetic beadsNew England BiolabsCat# S1420SSytoRNA SelectInvitrogenCat# S32703ProLong Gold Antifade MountantInvitrogenCat# P36930Critical commercial assaysMycoAlert Plus kitLonzaCat# LT07-318High-Capacity cDNA Reverse Transcription KitThermo Fisher ScientificCat# 4368814Direct-zol RNA Miniprep KitZymo ResearchCat# R2051Western ECL SubstrateBiological IndustriesCat# 20-500-120RNeasy Mini KitQiagenCat# 74104Qubit RNA BR assay kitThermo Fisher ScientificCat# Q10210TruSeq Stranded mRNA Library Prep KitIlluminaN/AHiScribe T7 In Vitro Transcription KitNew England BiolabsCat# E2040SHCR v3.0 smFISH assay (B1/B2/B3 amplifier system)Molecular Instrumentshttps://www.molecularinstruments.comDeposited dataRNA-seq dataThis paperPRJNA1258730Proteomics (SILAC-MS) dataThis paperPXD054139Proteomics (RNA antisense purification-MS) dataThis paperPXD054139Experimental models: Cell linesPlasmodium falciparum NF54 strainMR4MRA-1000Human monocytic cell line THP-1ATCCTIB-202OligonucleotidesqPCR primers for Pf and human genesThis paperSee Tables S1 and S2RNA antisense probes for ETRAMP11.2This paperSee Table S2Software and algorithmsImageJ/FIJISchneider et al.https://imagej.net/ij/Imaris Cell ModuleBitplane (Oxford Instruments)https://imaris.oxinst.comSpectroFlo Software v3.3.0Cytek Bioscienceshttps://cytekbio.comQuantStudio Software v1.1Applied BiosystemsN/AMaxQuant v1.5.2.8Cox & Mannhttps://www.maxquant.orgIDEAS v6.2 softwareAmnis Corp.N/AR v4.3.3R Foundationhttps://www.r-project.orgPerseus softwareMaxQuanthttps://maxquant.net/perseus/STAR alignerDobin et al.https://github.com/alexdobin/STARDESeq2Love et al.https://bioconductor.orgDEXSeqAnders et al.https://bioconductor.orgOtherNanosight NS300Malvern InstrumentsN/ACytek Aurora spectral flow cytometerCytek BiosciencesN/AImageStreamX mark II imaging flow-cytometerAmnis Corp.N/ALeica DMi8 confocal microscopeLeica MicrosystemsN/AQ Exactive HFX mass spectrometerThermo Fisher ScientificN/AJPK Nanowizard III AFMJPK InstrumentsN/A

Experimental model and study participant details

Plasmodium falciparum NF54 strain

The NF54 Pf parasite line was generously provided by the Malaria Research Reference Reagent Resource Center, MR4, obtained through BEI Resources, NIAID, NIH (Pf, strain NF54 (Patient Line E), MRA-1000, contributed by Megan G. Dowler). Parasites were cultured in pooled donor RBCs provided by the Israeli Magen David Adom blood bank at 4% hematocrit and incubated at 37°C in gas mixture of 1% O_2_ and 5% CO_2_ in N_2_.94 Parasites were maintained in RPMI-1640 medium pH 7.4 (Diagnovum, D840-P10), 25 mg/mL HEPES (Sigma, H3375), 50 μg/mL hypoxanthine (Mercury, 4010CBC), 2 mg/mL sodium bicarbonate (J.T. Baker, 144-55-8), 20 μg/mL gentamycin (Sigma, G9654), and 0.5% Albumax II (Gibco, 11021045).94 Growth was monitored using methanol-fixed Giemsa-stained blood smears.13 Pf cultures were tested for mycoplasma twice a month using a MycoAlert Plus kit.

THP-1 human monocytic cell line

The human monocytic leukemia cell line THP-1 was used as a model of human monocytes. THP-1 cells were cultured as previously described.13 Briefly, THP-1 cells were grown in complete RPMI-1640 (Sartorius, 01-100-1A) supplemented with 10% fetal bovine serum (Sigma, F7524), 1% non-essential amino acids (Sartorius, 01-340-1B), and 0.1% penicillin/streptomycin solution (Diagnovum, D910-100ML) in a humidified incubator at 37°C in 5% CO_2_. THP-1 cultures were tested for mycoplasma once a month using a MycoAlert Plus kit.

RBCs

This study was approved by the Institutional Review Board (IRB) of the Weizmann Institute. Human peripheral blood cells were obtained from healthy donors under this approval. Donor samples were de-identified prior to use. All samples were collected in accordance with institutional guidelines and relevant regulations.

Method details

Isolation of EVs

High parasitemia (≥5%) Pf culture or uninfected RBC (uRBC) growth media was collected 24 h post synchronization with Sorbitol. EV extraction was performed as reported previously13 and in accordance with MISEV guidelines.95 Briefly, cellular debris were removed by differential centrifugation at 413g for 5 min, 1,900 g for 10 min, and 15,180 g for 1 h. Supernatants were then filtered through a 0.45-μm filter and then concentrated using Vivacell units with molecular weight cut-off of 100 kDa (Sartorius, VC1042) according to the instructions of the manufacturer. Next, the EVs were pelleted by ultracentrifugation at 150,000 g for 16 h at 4°C. Finally, the pellet was carefully resuspended in sterile DPBS without Ca^+^ and Mg^2+^ (Sartorius, 02-023-1A).

Nanoparticle tracking analysis

Vesicle size distribution and concentration were determined using Nanosight NS300 (Malvern Panalytical). Sample size distributions in a liquid suspension by the analysis of Brownian motion via light scattering using a 405-nm laser and a coupled CCD camera.96 EV concentration was in the range of 1 × 10^12^ particles/ml.

Spectral flow cytometer analysis

EVs isolated from uninfected and Pf-infected RBC cultures were first labeled for RNA with SytoRNA Select (5 mM stock in DMSO, Thermo Fisher Scientific) by incubation with the dye at the at final concentration 1:1000 for 30 min at 37°C in the dark. Immediately afterward, lipophilic membrane Rhodamine B octadecyl ester perchlorate (R18) dye was added (1:200; 2 mg/mL stock in DMSO, Sigma Aldrich) for 20 min and incubated again for 30 min at 37°C in the dark. Unstained samples and samples stained only with SytoRNA Select and only with R18 were also prepared. The dyes were mixed with the EV samples by pipetting to improve their distribution and to minimize dye aggregate formation or precipitation.

The EV samples were loaded on 20% sucrose cushion (prepared in DPBS without Ca^2+^ and Mg^2+^) and ultracentrifuged as described previously.97 The obtained pellets were resuspended in 150 μL of DPBS without Ca^2+^ and Mg^2+^, and the nanoparticle concentration was measured using an NS300 Nanosight. Samples were then analyzed by spectral flow cytometry using a Cytek Aurora equipped with five lasers (16UV, 16V, 14B, 10YG, 8R; Cytek Biosciences). Aliquots of 40 μL were analyzed at a flow rate of 15 μL/min. The gains of all the fluorescence detectors were enhanced equally by 225% compared to default Cytek assay settings to increase the sensitivity to weak signals while maintaining detection of the characteristic spectral signature of fluorescent dyes. The trigger threshold was set to 500 on the B3 detector. Spectral unmixing of the raw data has been performed using SpectroFlo Software v3.3.0.

RNA isolation from cells and EVs

For RNA isolation from THP-1 cells, treated cells were thoroughly washed three times with PBS prior to RNA extraction, then the cells were resuspended in Trizol Tri reagent,31 and RNA was extracted according to Invitrogen’s Trizol protocol as previously described.14 For RNA isolation from EVs, the purified EV pellet was resuspended in 75 μL of DPBS without Ca^2+^ and Mg^2+^, Trizol Tri reagent was added, and RNA was extracted as described.14 The RNA concentration and purity were evaluated spectrophotometrically using a NanoDrop spectrophotometer (Thermo Fisher Scientific).

Pf mRNA stability assay

THP-1 cells were treated with Pf-EVs, and cell samples were collected at 1, 4, 16, and 24 h after EV addition. Cells were then washed three times with DPBS without Ca2+ and Mg2+ prior to RNA isolation. cDNA was then synthesized using a High-Capacity cDNA synthesis kit to the manufacturer’s protocol.98 ETRAMP11.2, ETRAMP2, and ETRAMP14.1 mRNA levels were assessed by quantitative real-time PCR (qPCR) using a 6 FlexReal-Time PCR System (Applied Biosystems) and QuantStudio real-time PCR software v1.1. Data are presented as the fold induction relative to the sample treated with Pf-EVs for 1 h and are mean relative quantities and associated errors of three biological replicates. qPCR primers used for this study are listed in Table S1.

Polysome pull-down assay

THP-1 cells were grown in a 15-cm Petri dish and were either treated with Pf-EVs or not as previously described.13^,^14 After 4 h, ribosome profiling was conducted as was previously described.14^,^99 In brief, the cells were incubated with 100 μg/mL cycloheximide for 5 min. Next, the cells were washed with cold buffer containing 20 mM Tris-HCl, pH 8, 140 mM KCl, 5 mM MgCl_2_, and 100 μg/mL cycloheximide. The cells were lysed with the same buffer that also contained 0.5% Triton, 0.5% sodium deoxycholate, 1.5 mM DTT, 150 units RNase inhibitor, and protease inhibitor cocktail. Cell lysates were vortexed and centrifuged at 12,000 g at 4°C for 5 min. The cleared lysates were loaded onto a 10%–50% sucrose (J.T. Baker, 4072.1000) density gradient and centrifuged at 247,600 g in an SW41 rotor for 105 min at 4°C. Gradients were fractionated, and the optical density at 254 nm was continuously recorded using an ISCO UA-6 absorbance detector. The collected samples were pooled to create three fractions: polysome-free (no ribosomes), light (2–5 ribosomes), and heavy (more than five ribosomes).33 RNA was isolated by extraction using Trizol Tri reagent (Invitrogen, cat. 15596026) following the manufacturer’s instructions as was previously described.33 RNA purification was performed using a Direct-zol RNA kit. RNA concentration and purity were evaluated spectrophotometrically using a NanoDrop spectrophotometer (Thermo Fisher Scientific).

Nuclear fractionation

The nuclear fractionation assay was adapted from the Cold Spring Harbor Laboratory protocol.38 Briefly, THP-1 cells treated or not with Pf-EVs were washed three times with DPBS without Ca^2+^ and Mg^2+^ prior to the addition of the cell disruption buffer (10 mM KCl, 1.5 mM MgCl_2_, and 20 mM Tris-HCl, pH 7.5). After a 15-min incubation with cell disruption buffer, cells were transferred to the Dounce homogenizer (Kimble, 885302) and homogenized until cells were fully fractionated. Fractionation of the cells was verified by light microscopy. Post fractionation, Triton 100X was added to a final concentration of 0.1%, and nuclei were pelleted by a centrifugation at 1500g for 5 min. The supernatant, containing cytoplasm, was transferred to a fresh tube. The nuclei were washed with DPBS without Ca^2+^ and Mg^2+^ at least three times prior to further processing. RNA was isolated from 75% of both cytoplasmic and nuclear fraction, and proteins were isolated from the other 25%. Proteins were extracted using RIPA buffer (150 mM NaCl (Bio Lab Ltd, 19030594), 1% Triton 100X, 0.1% SDS, 50 Mm Tris-HCl, 0.5% DOC and 20 μg of the obtained protein was subjected to gel electrophoresis using 10% SDS-PAGE and transferred to nitrocellulose membranes (Pall Corporation, 66485). The nitrocellulose membrane was blocked for 1 h with 5% skim milk (Tnuva) in 0.05% PBS-Tween20 (Fisher Bioreagents, BP337-500). The primary antibodies used: anti-Lamin B2, anti-calnexin, anti-Rab5, anti-Rab7, and anti-GAPDH (6C5).

All primary antibodies were used at a dilution of 1:1000 and incubated overnight at 4°C. The secondary antibodies used were goat anti-rabbit diluted 1:10,000 and goat anti-mouse diluted 1:20,000, both conjugated with HRP and incubated for 1 h at room temperature. The membrane was developed using Western ECL Substrate. Quantification of the blots was done using ImageJ software.

Hybridization chain reaction (HCR) single-molecule fluorescence in situ hybridization (smFISH)

The assay was performed according to the Molecular Instruments protocol.100 Briefly, THP-1 cells (0.5 × 10^6^ cells per well) were plated in a 24-well plate that contained coverslips (Bar-Naor, BNCB015RA1.5SN) coated with poly-D-lysine. Pf-EVs were added at a ratio of 50,000 EVs per cell; untreated THP-1 cells served as control. After 1 h, the cells were then washed twice with DPBS without Ca^2+^ and Mg^2+^ and fixed with 4% PFA for 10 min on ice. Post fixation, cells were washed twice with DPBS without Ca^2+^ and Mg^2+^. Cells were then permeabilized with 70% ethanol for 2 h at 4°C. After permeabilization, cells were washed twice with 2x Saline Sodium Citrate buffer (SSC) prior to pre-hybridization with probe hybridization buffer (Molecular Instruments, BPH02624) for 30 min at 37°C. Probes for ETRAMP11.2 (B1) and ETRAMP2 (B2) (custom made by Molecular Instruments) in hybridization buffer were added to the EV-treated cells and a probe for β-actin (B3) was added to untreated cells. After incubation overnight at 37°C, cells were washed four times with wash buffer (Molecular Instruments, BPW02824) for 5 min at 37°C and twice with 5x SSC, 0.1% Tween 20 for 5 min at room temperature. Pre-amplification was performed with amplification buffer (Molecular Instruments, BAM03124) for 30 min at room temperature. Amplifiers (B1-AF647, B2-AF594, Molecular instruments) were boiled at 95°C for 90 s, cooled for 30 min at room temperature, diluted in amplification buffer, and added to the cells. Amplifiers B1-647, B2-594, and B3-488 were added, and samples were incubated overnight at room temperature. Cells were washed five times with 5x SSC, 0.1% Tween 20 for 5 min at room temperature. In the last two washes, Hoechst 33342 and CFSE were added at a dilution of 1:000. Finally, Prolong Gold antifade reagent was added, and samples were placed on 13-mm round coverslips, which were mounted onto a glass slide (Menzel Glaser Gmbh, AA00008032E), sealed, and kept at 4°C prior to imaging with confocal microscopy.

Confocal microscopy

Samples were imaged using a Leica DMi8 confocal laser-scanning microscope equipped with pulsed white-light and 405-nm lasers, an HC PL APO 63x/1.4 oil-immersion objective, and HyD SP GaAsP detectors. Optical sections of 0.22 μm thick were collected for each sample using the following fluorophores: Hoechst 33342 (ex. 415 nm, em. 425–455 nm), CFSE and B3-488 (ex. 488 nm, em. 500–550 nm), B2-594 (ex. 594 nm, em. 605–640 nm), and B1-647 (ex. 647 nm, em. 660–750 nm). The pinhole was 1 AU. Confocal microscopy images were analyzed using Imaris Cell module.101^,^102 The signal ratio between the nuclei and the cytoplasm was calculated following segmentation of the nuclei, cytoplasm, and fluorescent spots. Displayed are brightfield images and overlayed max projections of Hoechst, CFSE, and fluorescent probe channels from FIJI software (ImageJ).103

RNA antisense purification

A set of 80-nucleotide oligonucleotides tiled across the entire ETRAMP11.2 mRNA sequence were designed in a manner similar to that previously published.47 All PCR primers used for this assay are listed in Table S2. The synthesized oligonucleotides were amplified through PCR, which incorporated the T7 promoter sequence.47 The amplified oligonucleotides were purified using Sera-mag magnetic beads. Single-stranded RNA probes were generated by a 16-h in vitro transcription reaction catalyzed by T7 RNA polymerase from 250 ng template DNA carrying the T7 promotor sequence using the HiScribe T7 Kit. RNA was purified using a RNeasy kit. Single-stranded DNA probes were generated by PCR in the presence of biotinylated primer (Sigma; see Table S2 for primer sequence)47 in a reaction with 1 μg of RNA template using the High-Capacity cDNA synthesis kit according to the manufacturer’s protocol.98 Template RNA was degraded by incubation in 100 mM NaOH (Bio Lab, 1908029) at 75°C for 10 min. The reaction was stopped by the addition of acetic acid to a final concentration of 100 mM. Biotinylated probes were purified using the RNeasy kit with minor modifications to the manufacturer’s protocol (samples were mixed with 3.5 volumes of RLT buffer then, 1.5 volumes of ethanol were added to the DNA/RLT mix).

After 1 h incubation of THP-1 cells treated or not with Pf-EVs, the cells were irradiated at 254 nm with 0.8 J/cm^2^ for 5 min. Next, cells were scraped and washed twice with DPBS without Ca^2+^ and Mg^2+^, and the cell pellets were flash-frozen in liquid N_2_. Cell pellets were lysed with cold lysis buffer (10 mM Tris–HCl, pH 7.5, 500 mM LiCl, 0.5% DDM, 0.2% SDS, 0.1% sodium deoxycholate in combination with sonication ((30 s on, 30 s off) x 5 cycles). Cell lysates were treated with 20 U of TURBO DNase at 37°C for 10 min. Samples were placed on ice, and the DNase reaction was stopped by addition of 10 mM EDTA (J.T. Baker, 8993-01), 5 mM EGTA (Bio Basic, 67-42-5), and 2.5 mM TCEP (Sigma, 646547). Cell lysates were then mixed with 1.5x hybridization buffer (15 mM Tris–HCl, pH 7.5, 7.5 mM EDTA, 750 mM LiCl, 0.75% DDM, 0.3% SDS, 0.15% sodium deoxycholate, 6 M urea, 3.75 mM TCEP) and incubated on ice for 10 min followed by a 10-min centrifugation at 16,000 g at 4°C. The lysates were flash-frozen in liquid N_2_. Streptavidin-coated magnetic beads were added, and samples were incubated for 30 min at 37°C with intermittent mixing. Streptavidin magnetic beads were washed prior to use three times with wash buffer (500 mM NaCl (Bio Lab Ltd, 19030594), 20 mM Tris–HCl pH 7.5, 1 mM EDTA) and twice with hybridization buffer (10 mM Tris–HCl pH 7.5, 5 mM EDTA, 500 mM LiCl, 0.5% DDM, 0.2% SDS, 0.1% sodium deoxycholate, 4 M urea, 2.5 mM TCEP).

Biotin-labeled DNA probes were denatured at 85°C for 3 min and placed on ice. Aliquots of 5 μg probe were added to the heated lysate and incubated at 67°C for 2 h with intermittent mixing. Then streptavidin magnetic beads (washed as described above) were incubated with the probe-RNA complex at 67°C for 30 min with intermittent mixing. Finally, the streptavidin beads were magnetically separated, washed four times with hybridization buffer followed by three washes with DPBS without Ca^2+^ and Mg^2+^ before on-bead digestion and mass spectrometry.

Stable isotope labeling with amino acids in cell culture (SILAC) proteolysis and mass spectrometry

THP-1 cells treated or not with Pf-EVs were resuspended in arginine- and lysine-free RPMI (Biological Industries, 06-1100-35-1A) supplemented with 5% dialyzed fetal bovine serum (Biological Industries, 04-011-1B) and 100 U/ml penicillin/streptomycin (Biological Industries, 03-031-1B). To this culture, L-arginine and L-lysine carrying “heavy” ^13^C and ^15^N isotopes were added (Cambridge Isotope Laboratories, 10-CNLM-539-H-1 and 8-CNLM-291-H-1, respectively). Cells were cultured in a 96-well plate for 16 h to allow incorporation of the labeled amino acids into newly synthetized proteins.

The proteins in the samples were extracted in 9 M urea (EMD-Merck Millipore, 1084871000), 10 mM DTT (Bio Basic, DB0058), and 400 mM of ammonium bicarbonate (Sigma, A6141) by two cycles of 10 s sonication. The proteins were reduced with 3 mM DTT (60°C for 30 min), modified with 10 mM iodoacetamide (Sigma, I6125) in 100 mM ammonium bicarbonate (room temperature for 30 min in the dark), and 20 μg aliquots were digested in 2 M urea, 25 mM ammonium bicarbonate with modified trypsin (Promega, VA9000), overnight at 37°C in a 1:50 enzyme-to-substrate ratio. The tryptic peptides were desalted using Top Tip C18 tips (Glygen, 5010–21002), dried, and resuspended in 0.1% formic acid (Sigma, 5.33002).

Aliquots of 2 μg of each sample were resolved by reverse-phase chromatography on 0.075 × 180-mm fused silica capillaries (J&W) packed with Reprosil reversed phase material (Dr. Maisch GmbH). The peptides were eluted with linear 180-min gradient of 5%–28% acetonitrile (Sigma, 34851) with 0.1% formic acid (Sigma, 5.33002) in water, followed by a 15-min gradient of 28%–95% acetonitrile with 0.1% formic acid in water, and 25 min at 95% acetonitrile with 0.1% formic acid in water at flow rates of 0.15 μL/min. Mass spectrometry was performed on a Q Exactive HFX mass spectrometer (Thermo Fischer Scientific) in a positive mode (m/z 300–1800, resolution 120,000 for MS1 and 15,000 for MS2) using a full scan followed by high collision induces dissociation (at 27 normalized collision energy) of the 30 most dominant ions (>1 charges) selected from the first scan. The AGC settings were 3 × 10^6^ for the full scan and 1 × 10^5^ for the MS/MS scans. The intensity threshold for triggering MS/MS analysis was 1 × 10^4^. A dynamic exclusion list was enabled with exclusion duration of 20 s.

The mass spectrometry data were analyzed using MaxQuant software 1.5.2.8. for peak picking identification and quantitation using the Andromeda search engine with data searched against the human proteome from the Uniprot database and the Pf database with mass tolerance of 6 ppm (after recalibration) for the precursor masses and for the fragment ions. Oxidation on methionine and protein N-terminus acetylation were accepted as variable modifications, and carbamidomethyl on cysteine was accepted as a static modification. Minimal peptide length was set to six amino acids, and a maximum of three missed cleavages was allowed. Peptide- and protein-level false discovery rates were filtered to 1% using the target-decoy strategy. Protein tables were filtered to eliminate the identifications from the reverse database. The Heavy/Medium ratios for all peptides belonging to a particular protein species were pooled by the software, providing an average ratio for each protein.

mRNA expression analysis

THP-1 cells were treated with Pf-EVs at a ratio of 50,000 EVs per cell, and cell samples were collected at 1-, 2-, 3-, 4- and 16-h post EV treatment, from which the RNA was purified, and cDNA synthesized as described above. Untreated samples were also analyzed. Amounts of FN3KRP, ZWINT, TPX2, CNOT3, TEX264, GFI-1, RUNX1, RHOT2, NHLRC2, SCLY, ATAD2, and PCYT2 mRNAs were determined by quantitative real-time PCR using the SYBR-fast green detection system (flex 6 Real-Time PCR System, Applied Biosystems, QuantStudio real-time PCR software v1.1). Expression levels were normalized to GAPDH. qPCR primers used are listed in Table S2. ΔCT values were compared with a linear model testing the effect of treatment, time, treatment and time interaction, and batch. The effect of treatment per time point was tested with estimated marginal means using the R package ‘emmeans’. The analyses were conducted in R v. 4.3.3.

RNA sequencing

THP-1 cells were treated with Pf-EVs at a ratio of 50,000 EVs per cell or were not treated, and cell samples were collected after 3 h. RNA was purified using RNeasy mini kit. RNA concentrations were determined using the Qubit RNA BR assay kit. Sequencing libraries were prepared using the TruSeq Stranded mRNA kit. Paired-end reads were sequenced on 1 lane of an Illumina NovaSeq. The output was ∼114 million reads per sample.

Poly-A/T stretches and Illumina adapters were trimmed from the reads using cutadapt,104 and reads shorter than 30 bases were discarded. Reads were mapped to the Homo sapiens reference genome GRCh38 using STAR105 with gene annotations downloaded from Ensembl. EndToEnd option was used, and outFilterMismatchNoverLmax was set to 0.04. Reads with the same unique molecular identifiers were removed using the PICARD MarkDuplicate tool using the BARCODE_TAG parameter. Expression levels for each gene were quantified using htseq-count.106 Differentially expressed genes were identified using DESeq2107 with the betaPrior, cooksCutoff, and independent Filtering parameters set to false. Raw p values were adjusted for multiple testing using the procedure of Benjamini and Hochberg. Differential exon usage analysis was performed using DEXSeq,108^,^109 which models read counts at the exon level while accounting for biological variability. Exon-level counts were also obtained using DEXSeq. Normalization and statistical testing were carried out using the default DEXSeq GLM approach to determine significant differential exon usage. The pipeline was run using snakemake.110

Atomic force microscopy (AFM)

Pf-EV samples were diluted 1:1000. Freshly cleaved mica was first incubated with 50 μL of 10 mM MgCl_2_ (Sigma, M8266) for 2 min, then washed with 1 mL PBS. A 20-μL drop of diluted EV sample was incubated on the Mg^2+^ modified mica for 5 min and then washed three times with 200 μL of DPBS without Ca^2+^ and Mg^2+^. AFM imaging was performed using a JPK atomic force microscope (JPK Nanowizard III) as described previously.97 Scans were made in QI mode using a qp-BioAC-Cl-CB3 (Nanosensors) with a nominal spring constant of 0.06 N/m. Detector sensitivity and spring constant were determined using JPK software. Image analysis was performed using Gwyddion111 or JPK-SPM data processing software and images were assembled in Adobe Illustrator 2019.

Thiazole orange-labeled EV uptake into THP-1 cells

Pf-EVs were stained using thiazole orange (1 mg/mL stock in methanol) at a dilution of 1:1000 as previously described.28^,^31 THP-1 cells were incubated with the labeled Pf-EVs for 5 minutes at 37°C. Untreated cells and cells treated with thiazole orange in PBS were used as controls. Pf-EV internalization was detected by imaging flow cytometry as previously described28^,^31 using an ImageStreamX mark II imaging flow-cytometer. At least 5 × 10^4^ cells were collected from each sample, and data were analyzed using the IDEAS v6.2 software (Amnis Corp.). THP-1 cells were gated for single cells, using the area and aspect ratio features, and for focused cells, using the Gradient RMS feature, as previously described.28^,^31^,^112 Cropped cells were further eliminated by plotting the cell area of the bright field image against the Centroid X feature (the number of pixels in the horizontal axis from the left corner of the image to the center of the cell mask). Pf-EV internalization was evaluated based on the intensity (the sum of the background-subtracted pixel values within the masked area of the image) and the max pixel (the largest value of the background-subtracted pixel). Gating was set according to the untreated sample.

Western blot assay

THP-1 cells (1.5 × 10^6^ cells per well) were plated in a 6 well plate and Pf-EVs, uRBC-EVs, or media were added to the cells. Cells were treated with EVs at a ratio of 50,000 EVs per cell. The samples were incubated for 20 h. THP-1 cells were then lysed using RIPA buffer (150 mM NaCl (Bio Lab Ltd, 19030594), 1% Triton 100X, 0.1% SDS, 50 mM Tris-HCl, 0.5% DOC and 20 μg of the obtained protein was subjected to gel electrophoresis using 10% SDS-PAGE and transferred to nitrocellulose membranes (Pall Corporation, 66485). The nitrocellulose membrane was blocked for 1 h with 5% skim milk (Tnuva) in 0.05% PBS-Tween20. The primary antibodies used for detection were obtained commercially from Santa-Cruz, Cell signaling and Abcam. The following primary antibodies were used in this work: anti-ICAM-1, anti-Mx2 and anti-OAS2.

All primary antibodies were used at a dilution of 1:1000 and incubated overnight at 4°C. The secondary antibodies used were goat anti-rabbit diluted 1:10,000 and goat anti-mouse diluted 1:20,000, both conjugated with HRP and incubated for 1 h at room temperature. The membrane was developed using Western ECL Substrate. Quantification of the blots was done using ImageJ software.

Quantification and statistical analysis

All statistical analyses performed in the present study were carried out on the mean of at least three independent biological replicates using “R” version .5.1, as indicated in each figure legend. The mean of each set of biological repeats is represented in all cases. The tests were t-tests or ANOVA, as mentioned in each figure legend. Multiple comparisons were performed where needed, as indicated in the respective figure legends, and p values were calculated to show statistical significance. Asterisks in the figures indicate statistical significance: p < 0.05^∗^, p < 0.01^∗∗^, p < 0.001^∗∗∗^ and p < 0.0001^∗∗∗∗^.