Plasma exosomes from COVID-19 patients promote cytokine production in peripheral blood mononuclear cells (PBMCs). We obtained plasma specimens (200 – 250 µl each) collected from 50 hospitalized COVID-19 patients early upon admission and again from later in their hospitalization from the COVID-19 and Coronavirus Biorepository at the University Hospitals Cleveland Medical Center (UHCMC), Cleveland, OH (Supplementary table S1). We also received plasma specimens collected from age- and gender-matched hospitalized non-COVID-19 (non-COVID) donors from the UHCMC. These individuals were age 11 or greater. Among 10 patients who had available laboratory tests for IL-6, nine presented higher IL-6 serum levels (8.3 – 255.8 pg ml-1) during hospitalization than reference individuals (≤ 2 pg ml-1). We purified plasma exosomes using the differential ultracentrifugation protocol and quantified plasma exosome yields based on acetylcholinesterase activity17,18. Exosome-depleted plasma from the same patients/donors was obtained simultaneously and used as autologous controls. The yield of plasma exosomes was typically approximately 2× 109 exosomes from approximately 200 µl of each plasma sample. Given the similarity in size and density between SARS-CoV-2 virions and exosomes, exosomes purified from the plasma of COVID-19 patients may lead to coprecipitation of exosomes and viral particles. To eliminate the possibility of viral contamination in plasma exosomes, we heated the plasma collected from COVID-19 patients and non-COVID donors at 57 °C for 30 min to inactivate the virus14, a procedure that does not affect the stability of plasma exosomes before exosome preparation29,30.
To determine whether plasma exosomes from COVID-19 patients contained SARS-CoV-2 components, we used a one-step RT-PCR platform suitable for qualitative detection of SARS-CoV-2 nucleic acids in saliva and nasal swab samples without the need for RNA extraction (RayBiotech Inc.). We detected exosomal SARS-CoV-2 nucleic acids in over 90% of COVID-19 patients (45 positives vs. 5 negatives) following the manufacturer’s protocol. Among 50 plasma exosome samples from patients early in hospital admission, 41 (82%) were SARS-CoV-2 positive. However, only 10 out of 50 plasma exosome samples (one sample had no data) from the same patients later in their hospitalization were SARS-CoV-2 positive, suggesting diminishing COVID-19 plasma exosomes during hospitalization. To quantify SARS-CoV-2 nucleic acids in COVID-19 plasma exosomes, we extracted total RNA from COVID-19 plasma exosome samples of 8 patients for reverse transcription and subsequent quantitative PCR (RT-qPCR) using primer sets for the detection of SARS-CoV-2 S and N genes (ScienCell, Inc.). We detected SARS-CoV-2 RNA, either envelope spike protein 1 (S) RNA or nuclear (N) gene RNA or both, in 7 out of 8 COVID-19 plasma exosome samples. Specifically, S RNA was identified in 5 exosome samples, while N RNA was identified in 3 samples from patients early in the admission (referred to as to early stage COVID-19). In these same patients who were in the later phase of hospitalization (referred to as late-stage COVID-19), however, S and N RNA was detected in 3 and 2 plasma exosome samples, respectively (Table 1). Most importantly, we detected viral RNA in COVID-19 plasma exosomes in some patients up to 86 days after infection (Pt0001, 86 days; Pt0959, 8 days; Pt2077, 9 days). Our data indicate the presence of SARS-CoV-2-associated exosomes in the circulation of patients during the acute phase of COVID-19 and, importantly, the persistence of COVID-19 plasma exosomes approximately 3 months after recovery. To determine whether the detected viral nucleotide acids were derived from SARS-CoV-2-infected cells, we transfected SARS-CoV-2-ΔN/EGFP BAC or SARS-CoV-2-ΔN/EGFP-alpha BAC, in which the viral N gene was replaced by EGFP cDNA in the viral genome, into Vero E6 cells for exosome isolation. We found that exosomes isolated from culture supernatants of SARS-CoV-2-ΔN/EGFP VERO E6 cells contained the viral S gene using RT-qPCR (Table 2). Exosomes from culture media of A549 cells that overexpressed both S and N genes, as a positive control, were positive for both the S and N genes. Our results are consistent with a recent report about the presence of SARS-CoV-2 RNA in the exosomal cargo25 and suggest that the plasma from COVID-19 patients contains SARS-CoV-2-associated exosomes.
COVID-19 plasma exosomes stimulate immune responses in PBMCs. The production of cytokines, particularly IL-6, IL-8, and TNF-α, is associated with the progression and severity of COVID-1931,32. To determine the response of immune cells to plasma exosomes, we first treated PBMCs with COVID-19 plasma exosomes and liposaccharides (LPS) or left untreated for 16 h, followed by flow cytometry to determine intracellular IL-6, IL-8, and TNF-α in PBMCs gated on CD3. We found that COVID-19 plasma exosomes and LPS stimulated significant production of IL-6, IL-8, and TNF-α compared with untreated CD3+ lymphocytes, indicating that PBMCs were able to respond to COVID-19 plasma exosomes (Supplementary Fig. 1).
To determine whether SARS-CoV-2-associated exosomes played a role in the regulation of immune responses, we treated PBMCs isolated from healthy donors with plasma exosomes (4 × 109 ml-1, equivalent to exosomes out of 200 – 300 µl of plasma) for 16 h in RPMI media and then collected culture supernatants for semiquantitative antibody arrays to measure cytokines. Plasma exosomes derived from early-stage COVID-19 patients (E) stimulated the production of IL-6, IL-8, TNF-α, IL-11, IL-17, MIP-1β, TGFβ3, and BMP4 compared with those from the same patients in the late stage (L) (Fig. 1a, b). This suggests that SARS-CoV-2-associated plasma exosomes played a role in the elevation of cytokines observed in COVID-19 patients 31. The response of PBMCs to early- and late-stage COVID-19 plasma exosomes was clearly separate in the principal components analysis (PCA) space (Fig. 1c). We quantified the fold change in proteins released from PBMCs in response to COVID-19 plasma exosomes based on their statistical significance using a simple linear model. Our analyses revealed a dramatic increase in TNF-α, IL-8, IL-6, VEGFD, VEGFR3, TGFβ3, IL-5, GM-CSF, GDF15, and IFNγ, with BDNF, HB-EGF, PDGF-AA, and IGFB-2 proteins decreasing compared with all cytokines relative to their fold changes in volcano plots (Fig. 1d).
We then treated PBMCs with COVID-19 plasma exosomes purified from patients’ plasma specimens collected early in admission and again later in hospitalization and plasma exosomes from matched non-COVID donors. We found that plasma exosomes from COVID-19 patients upon hospital admission significantly increased the expression of IL-6, IL-8, and TNF-α in PBMCs gated on CD3+ lymphocytes, CD4+ T cells, CD8+ T cells, and CD14+ monocytes compared with plasma exosomes from the same patients later in their hospitalization or those from non-COVID controls (Fig. 2a, b). Moreover, COVID-19 plasma exosomes were unable to induce the expression of IL-6 in CD8+ T cells relative to treatment with non-COVID-19 plasma exosomes. Importantly, exosome-depleted COVID-19 plasma failed to stimulate cytokine production in PBMCs, suggesting the potential of SARS-CoV-2-associated exosomes to elevate cytokine levels in COVID-19 patients. To determine if SARS-CoV-2-associated exosomes initiated the response of PBMCs, we treated PBMCs with exosomes isolated from culture supernatants of SARS-CoV-2-ΔN/EGFP VERO E6 cells, followed by flow cytometry for cytokine expression. We found that exosomes from SARS-CoV-2-ΔN/EGFP VERO E6 cells significantly induced the expression of IL-6 and TNF-α in CD14+ monocytes and IFNγ in CD8+ cells (Supplementary Fig. 2a, b). Although these exosomes only marginally induced levels of IL-6, TNF-α, and IFNγ in CD4+ T cells, IL-6 and TNF-α in CD8+ T cells, and IL-8 in CD14+ T cells, the trend of exosome induction of cytokines was evident.
SARS-CoV-2-associated plasma exosomes differentially interact with immune cells. To determine whether COVID-19 plasma exosomes directly regulated the immune responses of a specific type of immune cell within PBMCs, we separated CD4+ T cells, CD8+ T cells, and CD14+ monocytes from PBMCs using MACS MicroBeads (Miltenyi Biotec Inc. ), followed by treatment with plasma exosomes and flow cytometry gated on live cells. We found that in CD4+ T cells, COVID-19 plasma exosomes from early-stage patients stimulated the production of IL-6, IL-8, and TNF-α compared with plasma exosomes from these same patients in their later hospitalization and those from non-COVID donors, while the expression of IFNγ was not significantly affected (Fig. 3a, Supplemental Fig. 3a and 3b). In CD8+ T cells, plasma exosomes from early and later hospitalization of the same patients increased IL-6, TNF-α, and IFNγ production compared with those from non-COVID donors. Similarly, COVID-19 plasma exosomes from patients in both the early and late stages of the disease stimulated the expression of IL-6, IL-8, and TNF-α in CD14+ monocytes (Fig. 3a). Th-17 T cells are a subset of CD4+ T helper cells characterized by the production of IL-17 and may have evolved for host protection against microbes33. We found that plasma exosomes from early-stage patients stimulated IL-17 and IL-6 expression in CD4+ Th17 cells. However, TNF-α production remained unchanged (Fig. 3b, Supplementary Fig. 3c). Regulatory T cells (Tregs) are able to produce soluble factors, such as IL-10, to suppress the activation, proliferation, and cytokine production of CD4+ T cells and CD8+ T cells34. COVID-19 plasma exosomes from patients later in their hospitalization induced the production of IL-6, TGFβ, and IL-10 compared with plasma exosomes from early-stage patients or non-covid donors (Fig. 3b, Supplementary Fig. 3c). In addition, COVID-19 plasma exosomes failed to affect cytokine production in CD4+/CD45RO+ central memory T cells that were selected using a CD4+ central memory T cell isolation kit (Miltenyi) (Fig. 3b, Supplementary Fig. 3c). Our findings suggested that certain subpopulations of T cells, including Th-17 cells, Treg cells, and CD4+ central memory T cells, were less responsive to COVID-19 plasma exosomes than CD4+ T cells, CD8+ T cells, and CD14+ monocytes.
SARS-CoV-2 viral dsRNA contributes to immune responses to virus-associated exosomes. A wide spectrum of viruses infect permissive cells and produce viral nuclei acids that, when released from infected cells, subsequently induce innate immune responses via pattern recognition receptors (PRRs), including the endosomal receptor TLR335-37. We have reported that exosomes purified from the plasma of people with HIV and those isolated from culture supernatants of HIV-infected T cells contain viral double-stranded RNA (dsRNA), which induces the expression of proto-oncogenes and IFN-stimulated genes (ISGs) in cancer cells via TLR317. SARS-CoV-2 infection and replication yield dsRNA intermediates, which are potentially involved in eliciting innate immune responses of respiratory tract-derived cells and cardiomyocytes35. To determine whether COVID-19 plasma exosomes contained dsRNA, we quantified dsRNA by using the viral dsRNA detection system (PerkinElmer, Waltham, MA) based on the fluorescence resonance energy transfer (FRET) platform with a monoclonal dsRNA antibody38. We identified viral dsRNA in plasma exosomes from each COVID-19 patient we tested, which was not present in plasma exosomes from non-COVID donors or in exosomes from Jurkat T cells (Fig. 4a). Our positive control showed dsRNA in exosomes isolated from culture supernatants of HIV-infected J1.1 T cells, as expected17. The presence of viral dsRNA in COVID-19 plasma exosomes was further validated by the detection of dsRNA in exosomes isolated from culture supernatants of VERO E6 cells transfected with either SARS-CoV-2-ΔN/EGFP or the SARS-CoV-2-ΔN/EGFP UK variant. However, exosome-depleted supernatants did not contain dsRNA (Fig. 4b). To determine if COVID-19 plasma exosomes would transfer the viral dsRNA into recipient cells, we incubated CD3+ lymphocytes with COVID-19 plasma exosomes, followed by dsRNA quantification in cells. We identified dsRNA in CD3+ lymphocytes incubated with COVID-19 plasma exosomes and HIV-infected J1.1 T-cell exosomes but not those treated with plasma exosomes from non-COVID donors (Fig. 4c). These results indicate that COVID-19 plasma exosomes are able to transfer viral dsRNA cargoes to immune cells effectively, as we reported before17. To determine whether COVID-19 plasma exosomes contained the viral S and/or N protein, we extracted total exosomal proteins (100 µg total protein for each loading) for immunoblotting using antibodies against the S and N proteins. We did not identify viral proteins in COVID-19 plasma exosomes that were positive for the exosome marker CD9 using immunoblotting, although exosomes from culture supernatants of A549 cells overexpressing S and N proteins presented both S and N proteins (Fig. 4d). Consistent with this observation, we did not detect the S protein in exosomes from culture supernatants of SARS-CoV-2-ΔN/EGFP VERO E6 cells, while the exosome marker CD63 protein was detected (Fig. 4e). While the detection of SARS-CoV-2 protein(s) was reported in COVID-19 plasma exosomes immobilized in a chip platform26, our findings suggest that viral RNA may serve as a major cargo of COVID-19 plasma exosomes and, once delivered into recipient cells, induce cellular responses.
SARS-CoV-2 viral dsRNA contributes to immune responses to COVID-19 plasma exosomes. TLR3 detects dsRNA derived from the viral genome released from damaged host cells, viral particles, and/or extracellular vesicles from infected cells17,39,40. Various cell types, including immune cells, epithelial cells, and endothelial cells, express TLR3 and respond to dsRNA to trigger TLR3 signaling41,42. To determine whether COVID-19 plasma exosomes induce the immune response by assembling the action of viral dsRNA, we treated MicroBeads selected CD4+ T cells, CD8+ T cells, and CD14+ monocytes with plasma exosomes from COVID-19 patients and non-COVID donors as well as polyinosinic-polycytidylic acid (poly(I:C)), a synthetic analog of dsRNA and a potent activator of TLR343,44, and poly(I:C) treated with an RNase. We found that COVID-19 plasma exosomes significantly induced the expression of cytokines and chemokines in CD4+ and CD8+ T cells as well as in CD14+ monocytes compared with treatment with plasma exosomes from non-COVID-19 patients and healthy donors (Fig. 5a). While poly(I:C) moderately induced the expression of IL-6 in CD4+ T cells and IL-6 and TNF-α in CD14+ monocytes compared with plasma exosomes from healthy donors, it had no effect on the expression of cytokines in CD8+ T cells (Fig. 5a). Our results suggested that exogenous dsRNA mimics were unable to recapitulate the function of exosomal viral dsRNA. We speculated that plasma exosomes would protect the viral RNA cargo from RNase degradation and effectively deliver it to recipient cells for signaling. Indeed, COVID-19 plasma exosomes treated with RNase were still able to induce expression of cytokines in lymphocytes and monocytes (Fig. 5b), indicating that the viral dsRNA within COVID-19 plasma exosomes played a critical role in the regulation of the immune responses of PBMCs and was able to transmit pathogenic factors to nonpermissive cells.
To determine if TLR3 mediated the immune response to COVID-19 plasma exosomes, we treated PBMCs with the dsRNA/TLR3 small molecular inhibitor, a competitive inhibitor of dsRNA binding to TLR3 with high affinity and specificity17,45, followed by stimulation with COVID-19 or non-COVID-19 control plasma exosomes for flow cytometry. Our results showed that the TLR3 inhibitor blocked the production of IL-6, TNF-α, and IFNγ in CD4+ T cells and CD8+ T cells; however, the inhibitor failed to affect the expression of cytokines in CD14+ monocytes (Fig. 6a). Thus, while TLR3 played a major role in cytokine production in T cells, monocyte pattern recognition receptors (PRRs) other than TLR3 might function in response to SARS-CoV-2-associated plasma exosomes. In addition, plasma exosomal proteins of COVID-19 patients may play a role in the immune response to SARS-CoV-225. Given that some cytokines and chemokines induced by COVID-19 plasma exosomes, such as IL-6 and IL-8, are not typical signature cytokines or chemokines of T cells, we investigated how PRRs of T lymphocytes and monocytes responded to COVID-19 plasma exosomes. MicroBeads selected CD4+ T cells, CD8+ T cells, and CD14+ monocytes from PBMCs were treated with plasma exosomes from early-stage COVID-19 patients and non-COVID donors for 16 h at 37 °C, followed by flow cytometry for the expression of TLR3, TLR7, TLR8, and TLR9 gating on live cells. We found that COVID-19 plasma exosomes significantly induced the expression of TLR3 and TLR9 in all subsets of immune cells tested, while COVID-19 plasma exosomes were unable to induce TLR7 expression in CD8+ T cells. The expression of TLR8 in CD8+ T cells was not significantly different between COVID-19 and non-COVID-19 plasma exosome treatments. Poly(I:C), however, failed to stimulate the expression of any of these TLRs in these PBMCs (Fig. 6b). Our findings suggest that COVID-19 exosomes derived from SARS-CoV-2-infected cells are able to make innate and adaptive immune cells susceptible to exosomal viral cargoes and induce robust proinflammatory cytokine/chemokine responses, which may contribute to the severity and delayed recovery of COVID-19.