Inhibition of acid or neutral sphingomyelinases differentially impacts RNA and protein cargo sorting to extracellular vesicles

Abstract

Extracellular vesicles (EVs) form through regulated biogenesis processes involving sphingomyelinases (SMases), enzymes that metabolize sphingomyelin to produce ceramide-a lipid influencing membrane rigidity and essential for EV generation. This study explores the impact of inhibiting neutral SMase (NSM) and acid SMase (ASM) on the sorting of EV protein and RNA cargoes in human MCF7 cells. Our results revealed that NSM inhibition reduces EV nanoparticles and diminishes RNA and protein cargoes, including endosomal, spliceosomal, and translation-related proteins. Conversely, ASM inhibition increased RNA-binding proteins within and enhanced the expression of ribonucleoprotein complex-associated RNA in released EVs, including several snRNAs and 7SL RNA. Intriguingly, ASM-inhibited EVs enhanced the migration and translational activity of recipient MCF10A cells. These findings suggest an important role for SMase-dependent vesiculation in governing RNA and protein trafficking to the extracellular space, unveiling potential implications for cellular communication and function.

Keywords: Biological sciences; Cell biology; Omics; Organizational aspects of cell biology.

Graphical abstract

Figure 1

Sphingolipid metabolism influences the release of EV RNA from cells (A) Cellular viability of MCF7 cells was measured after a 36-h incubation with sphingomyelinase (SMase) inhibitor drugs at different doses. Appropriate dosing concentrations were determined as 5 μM for FTY720 (ASM inhibitor) and 10 μM for GW4869 (NSM inhibitor), with DMSO (vehicle) used as the control condition in a volumetric equivalent volume of 0.2% relative to the total media. All subsequent experiments utilize these doses. (B) Activity of neutral sphingomyelinase (NSM) as quantified by measuring of relative fluorescence units (RFU) in cellular lysates derived from control or SMase inhibitor-treated cells. (C) Activity of acid sphingomyelinase (ASM) as quantified by measuring of relative fluorescence units (RFU) in cellular lysates derived from control or SMase inhibitor-treated cells. (D) Western blot lanes of cellular and EV proteins isolated from control cells or cells treated with SMase inhibitor drugs. The samples were probed for EV markers TSG101 and CD81, as well as the negative control marker calnexin (CANX). (E) Nanoparticle Tracking Analysis (NTA) performed on size exclusion chromatography (SEC)-purified EVs derived from the conditioned media of MCF7 cells treated with vehicle or EV-inhibitor drugs. (F) Transmission electron micrographs (TEM) depicting iodixanol gradient-purified EVs derived from either control cells or cells treated with SMase inhibitor drugs. The EVs were negatively stained with 2% uranyl acetate and imaged using a FEI Tecnai T12 120kV transmission electron microscope. (G) Nanoparticle diameter distributions, measured in nanometers (nm), as quantified using TEMs of EVs with the TEM ExosomeAnalyser tool. (H) Distribution of total particle area, measured in square nanometers (nm²), as quantified using TEMs of EVs with the TEM ExosomeAnalyser tool. (I) Quantifications of total detected nanoparticles, isolated via size exclusion chromatography (SEC). Measurements were recorded using nanoflow cytometry on the CytoFLEX platform. (J) Quantifications of RNA-positive nanoparticles, isolated via SEC. Measurements were recorded using nanoflow cytometry on the CytoFLEX platform. RNA labeling was done using SYTO RNASelect, a cell-permeable stain for nucleic acids that is selective for RNA. For all graphs, unless otherwise stated, the error bars represent the standard error of the mean (SEM) and the statistics were calculated by ordinary one-way ANOVA with Tukey post-hoc test (where ns = p > 0.05, ∗ = p ≤ 0.05, ∗∗ = p ≤ 0.01, ∗∗∗ = p ≤ 0.001, and ∗∗∗∗ = p ≤ 0.0001) on GraphPad Prism. See also Figure S1.

Figure 2

NSM or ASM inhibition differentially impacts the EV proteome (A) Total ion current (TIC) chromatograph area of proteins isolated from either control or SMase inhibitor-treated cells, as estimated by initial injection volume and total sample volume. (B) TIC chromatograph area of proteins isolated from EVs derived from either control or SMase inhibitor-treated cells, as estimated by initial injection volume and total sample volume. (C) Multidimensional scaling (MDS) of liquid chromatography-tandem mass spectrometry (LC-MS/MS) proteomics data derived from cellular and EV purified proteins under control or SMase inhibitor-treated conditions. (D) Pearson correlations of cellular and EV proteomes under control or SMase inhibitor-treated conditions. (E) Clustered heatmap analysis of peptide intensities (relative Z score) of EV proteomes from control or SMase inhibitor-treated conditions. Blue lines denote lower association, while red lines denote higher association of each protein relative to the other two groups. Only statistically significant proteins were included (adjusted p value ≤0.05 and Log₂ fold change ≤ −1 or ≥1). (F) Cytoscape ClueGO analysis displaying a gene ontology (GO) biological process network of associated protein clusters, as identified by ORA, of statistically significant downregulated or upregulated EV proteins (adjusted p value ≤0.05 and Log₂ fold change ≤ −1 or ≥1) from control or SMase inhibitor-treated conditions. Green nodes represent downregulated FTY720 EVs, purple nodes represent downregulated GW4869 EVs, and red nodes represent upregulated FTY720 EVs. (G) Boxplots of peptide intensities of EV protein clusters (subfigure E), categorized by overrepresentation analyses (ORA) with g:Profiler, displaying the total Log₂ peptide intensities of all proteins in the groups and their changes under control or SMase inhibitor-treated conditions. Only statistically significant proteins were included (adjusted p value ≤0.05 and Log₂ fold change ≤ −1 or ≥1). Examples from clusters 1–3 displayed. Statistical analysis details for subfigures C–G are provided in the methodology. For subfigures A and B, the error bars represent the standard error of the mean (SEM) and the statistics were calculated by ordinary one-way ANOVA with Tukey post-hoc test (where ns = p > 0.05, ∗ = p ≤ 0.05, ∗∗ = p ≤ 0.01, ∗∗∗ = p ≤ 0.001, and ∗∗∗∗ = p ≤ 0.0001) on GraphPad Prism. See also Figure S2.

Figure 3

Inhibition of ASM alters the transcript diversity of EVs (A) Spectrophotometer quantification of nucleic acid concentration, in nanograms per microliter (ng/uL), of total RNA isolated from either control or SMase inhibitor-treated cells. (B) Spectrophotometer quantification of nucleic acid concentration, in picograms per microliter (pg/uL), of total RNA isolated from either control or SMase inhibitor-treated cells. (C) Principal-component analysis (PCA) of RNA sequencing data derived from cellular and EV purified RNAs from control or SMase inhibitor-treated specimens. (D) Percentage distribution of RNA biotypes in EV transcriptomes from either control or SMase inhibitor-treated conditions. “Others” describe a varied group of minor transcripts, including but not limited to miscRNA, snRNA, miRNA, etc. (see Figure 4E). (E) Percentage distribution of RNA biotypes in EV transcriptomes from either control or SMase inhibitor-treated conditions. This subfigure represents the “Others” group in the Figure 4D. (F–H) Clustered heatmap analyses of transcript biotypes (subfigures F = protein coding, G = lncRNA, H = miscRNA) expression (relative Z score) in transcriptomes from control or SMase inhibitor-treated conditions. Blue lines denote lower expression, while red lines denote higher expression of all transcripts detected relative to the other two conditions. Genes displayed represent the 10 most differentially expressed transcripts from the highlighted clusters. Only statistically significant transcripts were included (adjusted p value ≤0.05 and Log₂ fold change ≤ −1 or ≥1). (I) Dot plot illustrating the asymmetrical distribution of transcripts detected in acid sphingomyelinase (ASM)-inhibited EVs (FTY720) compared to ASM-inhibited cellular transcriptomes, normalized to control cellular transcriptomes. The upper dotted red box, labeled J, represents statistically significant upregulated transcripts (adjusted p value ≤0.01 and log₂ fold change ≥2) in ASM-inhibited EVs that remain unchanged in cells in response to the drug condition. The lower dotted red box, labeled K, represents statistically significant upregulated transcripts (adjusted p value ≤0.01 and Log₂ fold change ≥2) in ASM-inhibited EVs, which are simultaneously downregulated in ASM-inhibited cells (adjusted p value ≤0.01 and Log₂ fold change ≥2) in response to the drug condition. (J) FuncAssociate v3.0 gene ontology (GO) enrichment analysis of statistically significant upregulated transcripts (adjusted p value ≤0.01 and Log₂ fold change ≥2) in ASM-inhibited EVs that remain unchanged in cells in response to the drug condition (see subfigure I). The calculated logarithm (base 10) of the odds (LOD) ratio and adjusted p value are displayed. (K) FuncAssociate v3.0 gene ontology (GO) enrichment analysis of statistically significant upregulated transcripts (adjusted p value ≤0.01 and Log₂ fold change ≥2) in ASM-inhibited EVs, which are simultaneously downregulated in ASM-inhibited cells (adjusted p value ≤0.01 and log₂ fold change ≥2) in response to the drug condition (see subfigure I). The calculated logarithm (base 10) of the odds (LOD) ratio and adjusted p value are displayed. For subfigures A and B, the error bars represent the standard error of the mean (SEM) and the statistics were calculated by ordinary one-way ANOVA with Tukey post-hoc test (where ns = p > 0.05, ∗ = p ≤ 0.05, ∗∗ = p ≤ 0.01, ∗∗∗ = p ≤ 0.001, and ∗∗∗∗ = p ≤ 0.0001) on GraphPad Prism. See also Figure S3.

Figure 4

RBPs as major differentially regulated EV cargoes in response to NSM or ASM inhibition (A) MA plot of RNA-binding proteins (RBPs), detected by liquid chromatography-tandem mass spectrometry (LC-MS/MS), illustrating the Log₂ fold change and Log₂ mean intensity of RBPs in EVs in response to neutral sphingomyelinase (NSM) inhibitor treatments (GW4869), relative to control EVs. (B) MA plot of RNA-binding proteins (RBPs), detected by LC-MS/MS, illustrating the Log₂ fold change and Log₂ mean intensity of RBPs in EVs in response to acid sphingomyelinase (ASM) inhibitor treatments (FTY720), relative to control EVs. (C) MA plot of RNA-binding proteins (RBPs), detected by LC-MS/MS, illustrating the Log₂ fold change and Log₂ mean intensity of RBPs in EVs in response to sphingomyelinase inhibitor treatments (GW4869/FTY720), relative to each other. (D) Bar graphs illustrating the relative expression of 7SL RNAs (from transcripts per million—TPMs), as detected by RNA sequencing (RNA-seq). The graphs display changes in transcript expression in EVs under both control and SMase inhibitor treatment conditions. (E) Clustered heatmap analysis of peptide intensities (relative Z score) depicting the changes in the expression of proteins associated with the signal recognition particle (SRP) within EVs under both control and SMase inhibitor-treated conditions. The color scale represents the strength of association, where blue signifies lower association and red indicates higher association for each protein compared to the other two groups. Only statistically significant proteins were included (adjusted p value of ≤0.05 and a Log₂ fold change of ≤ −1 or ≥1). (F) Bar graphs illustrating the relative expression of U snRNAs (from TPMs), as detected by RNA-seq. The graphs display changes in transcript expression in EVs under both control and SMase inhibitor treatment conditions. (G) Clustered heatmap analysis of peptide intensities (relative Z score) depicting the changes in the expression of proteins associated with small nuclear ribonucleoproteins (snRNPs) of the spliceosome within EVs under both control and SMase inhibitor-treated conditions. The color scale represents the strength of association, where blue signifies lower association and red indicates higher association for each protein compared to the other two groups. Only statistically significant proteins were included (adjusted p value of ≤0.05 and a Log₂ fold change of ≤ −1 or ≥1). For subfigures D and F, the error bars represent the standard error of the mean (SEM) and the statistics were calculated by ordinary one-way ANOVA with Tukey post-hoc test (where ns = p > 0.05, ∗ = p ≤ 0.05, ∗∗ = p ≤ 0.01, ∗∗∗ = p ≤ 0.001, and ∗∗∗∗ = p ≤ 0.0001) on GraphPad Prism. See also Figure S4.

Figure 5

ASM-inhibited EVs enhance protein translation in recipient cells (A) Quantification of total EV proteins within the equal volumes of EVs utilized for co-culture experiments. The EVs used are present within equivalent isolation volumes purified from similar cell numbers that have been incubated under both control and SMase inhibitor conditions. (B) Representative light microscope images of wounded MCF10A cells incubated with MCF7 EVs, obtained under control or SMase inhibitor conditions, as captured using the Incucyte Live-Cell Analysis System. The images were taken at the initiation (0 h 00 m) and conclusion (48 h 00 m) of the co-culture experiment. The Incucyte software applied a wound mask. The scale bar represents 700 μm. (C) Quantification of the percentage relative wound density (%RWD) in wounded MCF10A cells over a 48-h period. The cells were incubated with MCF7 EVs obtained under control or SMase inhibitor conditions. Images were captured by an automated camera using the Incucyte live-cell analysis system, and %RWD was calculated using its Scratch Wound Analysis Software Module. %RWD measures the spatial cell density in the wound area relative to the spatial cell density outside of the wound area at each time point. (D) Schematic of the cell proliferation assay through the incorporation of 5-ethynyl-2′-deoxyuridine (EdU) in newly synthesized DNA, followed by the chemoselective ligation of Alexa Fluor 488 picolyl azide by a click chemistry reaction. (E) Percentage of MCF10A cells in the total population labeled by 5-ethynyl-2′-deoxyuridine (EdU) after a 24-h incubation with EdU and MCF7 EVs obtained under control or SMase inhibitor conditions. EdU incorporates into newly synthesized DNA, and quantification is based on high-content microscopy imaging. (F) Schematic of the protein translation assay through the puromycylation of nascent proteins at the ribosome with O-propargyl puromycin (OPP), followed by the chemoselective ligation of Alexa Fluor 488 picolyl azide by a click chemistry reaction. (G) Representative high-content screening microscopy images displaying nuclear stain (NuclearMask) and OPP-labeled protein foci (Alexa Fluor (AF)-488). Red arrows point to example OPP-labeled foci. (H) Distribution of quantified mean intensities (arbitrary units – a.u.) of O-propargyl puromycin (OPP)-labeled foci in MCF10A cells following a 24-h co-culture with MCF7 EVs obtained under control or SMase inhibitor conditions. OPP integrates into the nascent peptides of newly translated proteins, and quantification is performed based on images captured with high-content screening microscopy. For all graphs, unless otherwise stated, the error bars represent the standard error of the mean (SEM) and the statistics were calculated by ordinary one-way ANOVA with Tukey post-hoc test (where ns = p > 0.05, ∗ = p ≤ 0.05, ∗∗ = p ≤ 0.01, ∗∗∗ = p ≤ 0.001, and ∗∗∗∗ = p ≤ 0.0001) on GraphPad Prism. See also Figures S5 and S6.

Figure 6

SMase inhibition alters the RBP and RNA content of EVs Graphical representation of the proposed model illustrating the distinct effects of ASM and NSM inhibition on extracellular vesicle (EV) populations and cargo composition. ASM inhibition promotes an enrichment of proteins associated with endosomal pathways, suggesting a shift toward endosome-derived EVs (exosomes) over plasma membrane-derived microvesicles. These exosomes exhibit increased levels of RNA-binding proteins (RBPs) and RNA cargoes. In contrast, NSM inhibition decreases the presence of RBPs and RNA cargoes, highlighting the contrasting roles of ASM and NSM in regulating EV cargo profiles and their cellular origins.