Differential Expression of Viral Transcripts From Single-Cell RNA Sequencing of Moderate and Severe COVID-19 Patients and Its Implications for Case Severity

Abstract

With steady increase of new COVID-19 cases around the world, especially in the United States, health care resources in areas with the disease outbreak are quickly exhausted by overwhelming numbers of COVID-19 patients. Therefore, strategies that can effectively and quickly predict the disease progression and stratify patients for appropriate health care arrangements are urgently needed. We explored the features and evolutionary difference of viral gene expression in the SARS-CoV-2 infected cells from the bronchoalveolar lavage fluids of patients with moderate and severe COVID-19 using both single cell and bulk tissue transcriptome data. We found SARS-CoV-2 sequences were detectable in 8 types of immune related cells, including macrophages, T cells, and NK cells. We first reported that the SARS-CoV-2 ORF10 gene was differentially expressed in the severe vs. moderate samples. Specifically, ORF10 was abundantly expressed in infected cells of severe cases, while it was barely detectable in the infected cells of moderate cases. Consequently, the expression ratio of ORF10 to nucleocapsid (N) was significantly higher in severe than moderate cases (p = 0.0062). Moreover, we found transcription regulatory sequences (TRSs) of the viral leader sequence-independent fusions with a 5' joint point at position 1073 of SARS-CoV-2 genome were detected mainly in the patients with death outcome, suggesting its potential indication of clinical outcome. Finally, we identified the motifs in TRS of the viral leader sequence-dependent fusion events of SARS-CoV-2 and compared with that in SARS-CoV, suggesting its evolutionary trajectory. These results implicated potential roles and predictive features of viral transcripts in the pathogenesis of COVID-19 moderate and severe patients. Such features and evolutionary patterns require more data to validate in future.

Keywords: COVID-19; ORF10; SARS-CoV-2; cell infection rate; viral fusion evolution; virulence; virus evolution; virus gene fusion.

FIGURE 1

SARS-CoV-2 infected cells and differential gene expression of infected vs. uninfected epithelial cells in BALF samples. (A) UMAP (Uniform Manifold Approximation and Projection) view of the cell types for all SARS-CoV-2 infected BALF samples. Color labels each cell cluster identified by scRNA-seq data analysis (see “Materials and Methods”). (B) The heatmap of differentially expressed genes (DEGs) of infected epithelial cells vs. uninfected epithelial cells in moderate (mild) cases and in severe cases. One gene, BPIFA1, was observed in both moderate and severe cases.

FIGURE 2

SARS-CoV-2 gene expression and RNA fusions in BALF samples. (A) Differential SARS-CoV-2 gene expression (measured by normalized average gene expression) between the moderate and severe groups. (B) Expression of four SARS-CoV-2 genes (ORF1ab, ORF3a, N, and ORF10) in each infected cell. Each dot represents an infected cell. (C) Levels of read coverages in different viral regions in each sample (left) and ratio of coverage of ORF10 vs. N (right, ORF10/N). The unpaired t test was used for all the statistical tests.

FIGURE 3

Major viral RNA fusions detected in infected cells. (A) TRS-L dependent RNA fusions. (B) The most frequent TRS-L independent RNA fusions with a 5’ position at 1073 in SARS-CoV-2 genome.