Rapidly spreading coronavirus disease 2019 (COVID-19) is currently affecting the world. Specifically, cytokine storms are a key feature in a substantial number of COVID-19 patients,1 and studies from our group and others suggest that the IL-6/IL-6R cascade plays a dominant role in symptom-correlated cytokine storms.2,3 Cell-free circulating RNAs (cfRNAs) in plasma carry information from pathologic sites, and they have been reported to play important roles in disease development;4 however, their involvement in COVID-19 has not yet been clarified. Here, we report the characteristics of plasma cfRNA profiles of COVID-19 patients, and we found that no SARS-CoV-2 RNA is present in the plasma of COVID-19 patients. Compared with healthy donors, significantly higher mRNA expression of IL-6R was observed; miR-451a, a known negative regulator of IL-6R translation, was downregulated, which may promote IL-6R expression at the protein level. In addition, three upregulated long noncoding RNAs (lncRNAs) carrying miR-451a binding sites might function as miRNA sponges to compete with IL-6R for miR-451a in COVID-19 patients. Taken together, we provide the cfRNA landscape of COVID-19 patient plasma and describe the possible mechanisms underlying elevated cytokine storms in COVID-19 patients. These findings will contribute to the identification of drug targets for this new disease.

We analyzed the cfRNAs from three healthy donors and five COVID-19 patients (Fig. 1a), and no SARS-CoV-2 RNA was detected (Fig. 1b, c). We detected 33,562 human genes, including 57% protein-coding genes, 4% miRNA genes, 15% lncRNA genes, and 24% other noncoding genes (Fig. 1d and Fig. S1a). Compared with healthy donors, we identified 2583 upregulated and 192 downregulated cfRNA genes in all COVID-19 patients (Fig. 1e). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses revealed that the functions of the upregulated genes are significantly enriched in antivirus-related pathways such as “type I interferon signaling pathway” and “innate immune response and response to virus” (Fig. 1f). In particular, many interferon-stimulated genes (ISGs) were upregulated in COVID-19 patients, including ISG15, IFI6, IFI16, and IFI27L (Fig. 1f), which is consistent with a previous study reporting that IFNα expression is increased in COVID-19 patients.5 Among these ISGs, ISG15 may be related to prolonged viral latency of SARS-CoV-2.6 A previous study found that the IL-6 concentration in COVID-19 patient plasma was higher than the normal range (0–7 pg/ml).5 We also detected high expression of IL-6R in COVID-19 patients compared with healthy donors (Fig. 1f). this high level of IL-6R transcription might be a result of increased type I interferon signaling.7 Moreover, regarding stages of disease progression, expression levels of interferon-stimulated genes, and IL-6R showed a downward trend from stage 1 to stage 4 (Fig. S1b and Supplementary Table S2).

Fig. 1
figure 1

The landscape characteristics of cell-free circulating RNAs (cfRNAs) in the plasma of COVID-19 patients; no SARS-CoV-2 RNA was detected in COVID-19 patient plasma. a Schematic diagram showing the sample preparation procedure in this study. cfRNAs were collected from 3 healthy donors and 5 COVID-19 patients. b The computational pipeline to detect human RNAs and SARS-CoV-2 RNAs. c Bar plot showing the mapping rate and number of mapped reads against the SARS-CoV-2 genome (n = 32). d Bar plot and pie plot showing the composition of detected genes in all samples. Data are shown as the mean ± SD (n = 32). e Heatmap showing fold changes (relative to healthy donors) of 2,583 upregulated cfRNA genes and 192 downregulated cfRNA genes in COVID-19 patients. Blue and red represent log2-transformed fold changes < 0 and > 0, respectively. f GO and KEGG enrichment analyses of 2,583 upregulated cfRNA genes in COVID-19 patients (left). Expression levels (TPMs) of IL-6R and representative interferon-stimulated genes (ISGs) are shown on the right. g Volcano plot showing up- and downregulated microRNA (miRNA) and long noncoding RNA (lncRNA) genes in COVID-19 patients relative to healthy donors (left). Relative expression of miR-451a, LOC105371414, LOC105374981 and LOC107987081 is shown on the right. h Base-pairing interaction between miR-451a and IL-6R (top) and the three upregulated lncRNAs (bottom). miR-451a target sites (seed sequences) are highlighted in red. i A proposed model for the regulatory network of miR-451a, IL-6R and lncRNAs in healthy donors and COVID-19 patients. Asterisks indicate statistically significant differences: **P < 0.01; ***P < 0.001

Full size image

In addition to increased RNA expression of IL-6R, we also found miR-451a, a reported translational repressor of IL-6R,8 to be one of the top five downregulated microRNA (miRNA) genes in COVID-19 patients (Fig. 1g, Fig. S1c, d, Supplementary Table S1 and Supplementary Table S2). Moreover, expression levels of miR-451a showed an upward trend from stage 1 to stage 4 (Fig. S1b and Supplementary Table S3), suggesting that decreased miR-451a may promote expression of IL-6R in COVID-19 patients at the protein level. LncRNAs can act as miRNA sponges to inhibit miRNA function.9 We identified in COVID-19 patients three upregulated lncRNAs, LOC105371414, LOC105374981, and LOC107987081, carrying miR-451a binding sites (Fig. 1g, h and Fig. S1d), which may compete with IL-6R for miR-451a binding. We also found that miR-374a, the target of which is CCL2,10 was downregulated in COVID-19 patients (Fig. 1g). Derepression of CCL2 may confer acute respiratory distress syndrome (ARDS) and cause cytokine storms in COVID-19 patients.11 We also detected 16 upregulated lncRNAs carrying miR-374a binding sites (Fig. 1g and Fig. S2b). In healthy donors, miR-451a/miR-374a can maintain the normal level of IL-6R/CCL2 by targeting IL-6R/CCL2 mRNAs. However, in COVID-19 patients, decreased expression of miR-451a/miR-374a and its binding to lncRNAs may promote expression of IL-6R/CCL2 at the protein level (Fig. 1i). These results suggest that decreased miR-451a/miR-374a and enhanced lncRNA levels may exacerbate IL-6-induced cytokine storms by promoting IL-6R/CCL2 translation in COVID-19 patients.

The cytokine storm in COVID-19 patients is characterized by increased IL-62. However, the mechanisms of cytokine storm-correlated symptoms from the perspective of cfRNAs remain unclear. Our study identified obvious differences in cfRNA molecules between COVID-19 patients and healthy donors. We found activation of type I interferon-responsive genes and low expression of miR-451a, which may lead to uncontrolled expression of IL-6R at both mRNA and protein levels, enhancing cytokine storms in COVID-19 patients. Furthermore, the three lncRNAs identified as upregulated in COVID-19 patients may compete with IL-6R for miR-451a to reverse overexpression of IL-6R at the protein level. Collectively, our work provides the cfRNA landscape of plasma in COVID-19 patients, offers insight into the potential mechanism to understand the elevated cytokine storms caused by IL-6 in COVID-19 patients and may shed light on drug development for this new disease.