This study profiled the protein cargo of plasma-derived small extracellular vesicles (SEVs) from individuals with post-acute sequelae of SARS-CoV-2 infection (PASC, long COVID) and compared them with SEVs from PASC-negative controls. SEVs were isolated by size-exclusion chromatography and analyzed on the Olink Explore HT proteomics platform. Proteomic analyses identified hundreds of differentially expressed SEV proteins enriched in pathways related to coagulation, inflammation, apoptosis, fibrosis, extracellular matrix remodeling, mitochondrial dynamics, and immune activation. Several proteins previously implicated in acute COVID-19—including FN1, HCF-H, HGF, and IL-17RA—were persistently increased in PASC SEVs; HGF and IL-17RA showed greater differences in SEVs than in matched plasma. The authors propose that SEV proteomic alterations reflect inflammatory, thrombotic, and neurobiological dysregulation in long COVID and that SEVs are a promising source of biomarkers and mechanistic insight.
Post-acute sequelae of SARS-CoV-2 infection (PASC, long COVID) comprises a heterogeneous, multisystem condition with symptoms that can persist for months. The study frames PASC as frequently affecting pulmonary, neuropsychiatric, cardiovascular, endocrine, renal, gastrointestinal, hepatobiliary, and dermatologic systems. Proposed mechanisms include immune dysregulation, viral persistence or viral component retention, autoimmunity, and endotheliopathy. Given that extracellular vesicles (EVs) carry proteins, lipids, and RNAs and can reflect cellular and systemic pathobiology, plasma-derived SEV cargo may serve as a stable liquid-biopsy source to identify biomarkers and gain mechanistic insight into PASC.
Participants classified as PASC-positive (N=20 total enrolled; N=13 included in Olink proteomics) reported one or more symptoms persisting beyond 6–9 months after infection, except for three participants with shorter intervals noted in the methods. PASC-negative individuals (N=11 total enrolled; N=10 included in proteomics) were drawn from the ACTIV-2 trial cohort and reported no persistent symptoms 6–15 months after confirmed SARS-CoV-2 infection. Symptoms were captured with REDCap surveys and clinical documentation using the WHO Global COVID-19 clinical case report form for post-COVID conditions. The most common persistent symptoms in the PASC proteomics cohort included fatigue, shortness of breath, brain fog, sleep disturbances, musculoskeletal or chest pain, and rash.
EDTA plasma was processed to platelet-free plasma and SEVs were isolated by size-exclusion chromatography (qEV original 35 nm columns). Pooled fractions 7–10 were concentrated to 500 µl. Nanoparticle tracking analysis (NTA) was performed using NanoSight; five 60-second videos per sample were recorded. Western blot of SEV lysates assessed canonical EV markers (CD9, CD81, CD63, Integrin β1, Alix, TSG101). Protein abundance for proteomics was reported as normalized protein expression (NPX) on a log2 scale.
Equal volumes of SEV lysate from PASC-positive (n=13) and PASC-negative (n=10) individuals were randomized and analyzed on the Olink Explore HT panels at an external genomics center. Internal assay controls and sample quality metrics were used; samples failing technical criteria were excluded. For validation, selected targets were assayed by ELISA using commercial kits as described.
Differential protein expression used limma linear models. Symptom-associated comparisons employed Wilcoxon Rank Sum tests. Visualization included heatmaps and volcano plots. Pathway enrichment employed Enrichr for GO/KEGG, Ingenuity Pathway Analysis (IPA), STRING/Reactome mapping, and Cytoscape visualization. A p-value <0.05 was considered statistically significant; analyses were performed using R and GraphPad Prism for SEV concentration comparisons.
Among the proteomics cohort, age, sex, vaccination status, and smoking status did not differ significantly between PASC-positive and PASC-negative groups. BMI and interval from positive test to enrollment showed no statistically significant differences. Pulmonary disease, depressive disorder, and GERD were recorded more frequently in PASC-positive participants; coronary artery disease was observed only in the PASC-negative group. Detailed counts and medians are reported in Table 1 of the source article.
NTA showed that particle size distributions were comparable between groups, with most SEVs in the 50–200 nm range. Total SEV concentration normalized per ml of plasma or per microgram of EV protein did not differ significantly between PASC-positive and PASC-negative participants. Western blot confirmed presence of canonical EV proteins (tetraspanins CD9, CD63, CD81; Integrin β1; Alix; TSG101) in SEV preparations.
PEA proteomics identified substantial alteration of SEV-associated proteins in PASC. Heatmap clustering revealed six expression clusters, with one cluster predominantly downregulated and five clusters predominantly upregulated in PASC. The authors report 269 significantly dysregulated proteins overall, including proteins with >2-fold change (84 upregulated and 21 downregulated by >2-fold as summarized in the abstract). Volcano plots and hierarchical clustering illustrated these differences.
Differential SEV proteins were enriched in pathways related to coagulation and inflammation, as well as apoptosis, fibrosis, extracellular matrix remodeling, mitochondrial dynamics, and immune activation. Canonical pathway analysis using IPA and Reactome mapping supported involvement of thrombotic and immunologic signaling networks in PASC-associated SEV cargo.
Several proteins with increased abundance in PASC SEVs—specifically FN1, HCF-H, HGF, and IL-17RA—had been noted previously in acute COVID-19. The authors highlight that HGF and IL-17RA exhibited greater differences in SEVs than in matched plasma and were significantly altered in SEVs but not plasma, suggesting SEVs may amplify disease-relevant signals compared with bulk plasma.
The data indicate persistent SEV-associated signatures of inflammation and coagulation in individuals with long COVID, supporting the concept that SEVs could serve as a biomarker source and potential mediators of ongoing pathobiology. The enrichment of extracellular matrix, apoptosis, and mitochondrial pathways provides mechanistic hypotheses linking SEV cargo to tissue remodeling, immune activation, and thrombotic risk in PASC.
The study reports sample sizes for proteomic analysis (n=13 PASC, n=10 controls) and notes sample exclusions for technical QC but does not provide exhaustive subgroup analyses for all reported markers in the source text. The authors used multiple bioinformatics tools for pathway mapping. Data availability and additional details are referenced in the original article's data availability statement.