Functional genomics highlights differential induction of antiviral pathways in the lungs of SARS-CoV-infected macaques

Abstract

The pathogenesis of severe acute respiratory syndrome coronavirus (SARS-CoV) is likely mediated by disproportional immune responses and the ability of the virus to circumvent innate immunity. Using functional genomics, we analyzed early host responses to SARS-CoV infection in the lungs of adolescent cynomolgus macaques (Macaca fascicularis) that show lung pathology similar to that observed in human adults with SARS. Analysis of gene signatures revealed induction of a strong innate immune response characterized by the stimulation of various cytokine and chemokine genes, including interleukin (IL)-6, IL-8, and IP-10, which corresponds to the host response seen in acute respiratory distress syndrome. As opposed to many in vitro experiments, SARS-CoV induced a wide range of type I interferons (IFNs) and nuclear translocation of phosphorylated signal transducer and activator of transcription 1 in the lungs of macaques. Using immunohistochemistry, we revealed that these antiviral signaling pathways were differentially regulated in distinctive subsets of cells. Our studies emphasize that the induction of early IFN signaling may be critical to confer protection against SARS-CoV infection and highlight the strength of combining functional genomics with immunohistochemistry to further unravel the pathogenesis of SARS.

Figure 1. SARS-CoV mRNA Levels and Global Gene Expression in Lungs of SARS-CoV–Infected Macaques

(A) RT-PCR was performed on all individual pulmonary samples from SARS-CoV–infected animals and mock-infected animals to determine SARS-CoV levels. Samples from SARS-CoV–infected animals that were pooled for microarray studies are indicated with circles. The three samples with lower levels of virus were hybridized separately. The 12 samples from mock-infected animals (i.e., PBS) were pooled and served as a reference sample. The number after PBS refers to the animal (i.e., PBS 1), while the number after the dash refers to the lung piece (i.e., PBS 1-1). (B) Using microarrays, gene expression in SARS-CoV–infected animals was compared to gene expression in mock-infected animals. The bar graph shows the total number of genes considered to be differentially expressed, defined as an absolute fold change > 2 with p < 0.0001. Samples 1A and 1B (day 1) and 4A–4D (day 4) are composed of the pooled individual pulmonary samples shown in (A). Samples 1A low, 4A low, and 4B low are the outliers from (A). Separate mock samples (i.e., PBS) were compared to the total mock pool.

Figure 2. Unsupervised Global Gene Expression Profile of SARS-CoV–Infected Macaques and Individual Variation in Mock-Infected Animals

Red corresponds to higher gene expression than that of the controls; green corresponds to lower gene expression. (A) Gene expression profiles result from comparing gene expression in lungs of experimental animals versus gene expression in lungs of mock-infected animals (pooled). Genes were included if they met the criteria of a 2-fold change or more (p ≤ 0.0001). A two-of-nine strategy allowed samples to cluster together if profile similarities existed based on timing of inoculation (n = 2 samples for day 1). (B) The number after PBS refers to the animal (i.e., PBS 1), while the number after the dash refers to the lung piece (i.e., PBS 1-1). Thirty-eight genes are displayed with an absolute fold change > 2 and p <0.0001 in at least two animal samples. Up-regulated genes are indicated in bold underline. Only one gene, HLA-DQA1, was down-regulated > 5. No up-regulated genes met these criteria in mock-infected animals. Separate mock samples (i.e., PBS 1-1) were compared to the total mock pool.

Figure 3. Common and Unique Temporal Gene Responses to SARS-CoV Infection

(A) The Venn diagram shows genes with an absolute fold change > 2 and p < 0.0001 in two out of the two day 1 samples (left circle) or two out of the four day 4 samples (right circle). (B) The heat map includes the 597 genes from the grey Venn diagram intersections. The right column highlights the 97 genes that are commonly regulated in all six samples with blue. All 97 genes are listed in Figure S1. All gene expression profiles are the results of comparing gene expression in lungs of experimental animals versus gene expression in lungs of mock-infected animals (pooled). (C) Functional annotation was used to categorize the 681 unique genes at day 1, and the 353 unique genes at day 4. The percentage of genes within the top functional categories is indicated in the day 1 and the day 4 bar graph, with a white line indicating the percentage of genes found at the alternative day. Cell-to-cell signaling and cell death, other top categories on day 1, and cell death, cellular growth, and proliferation, additional top categories on day 4, are not included in the bar graph, as these categories were observed to have more genes differentially expressed in the common signature.

Figure 4. Immune Response, Cell Cycle, and Lung Repair Genes with Strongly Induced or Reduced Expression

A selection of genes, involved in the immune response, cell cycle, or lung repair processes, that showed an absolute fold change > 5 and p < 0.0001 in both day 1 animals and/or in at least two of the four day 4 animals was made. Genes with an absolute fold change > 5 in both day 1 animals and in at least two of the four day 4 animals are indicated with bold, underlined text. A full summary of genes that show an absolute fold change > 5 and p < 0.0001 after SARS-CoV infection is given in Figure S2.

Figure 5. Innate Host Response Profile from Tissues Showing Presence or Absence of Viral mRNA

Using a bioset that contained a selection of genes involved in pathological and antiviral pathways, a heat map was created showing all genes with expression changes > 2 and p < 0.0001 in at least one of the samples with high SARS-CoV levels (A) or in at least one of the samples with low SARS-CoV levels (B). Although there is some functional overlap with genes, the heat map is segregated by functional annotations. Chemokines (yellow), classical antiviral genes (blue), interleukins (white), JAK/STAT pathway, interferons, or ISGs (red), and transcriptional factors (pink). Genes with an absolute fold change > 5 in two day 1 animals and in at least two of the four day 4 animals are indicated with bold, underlined text.

Figure 6. Confirmation of Microarray Results with RT-PCR and Correlation of Induced Genes with Presence of SARS-CoV

Quantitative RT-PCR for IL-6, IL-8, CXCL10 (IP-10), and IFN-β was performed on all separate lung samples. Expression levels of these genes were plotted against the presence of SARS-CoV in these samples (as detected by RT-PCR). Correlation coefficients were determined using Spearman's correlation test.

Figure 7. Detection of IFN-β and Phosphorylated STAT1 in Lung of SARS-CoV–Infected Macaques Using Immunohistochemistry

(A) Lack of IFN-β expression in lungs of mock-infected macaques (magnification 40×) and (B, C) expression of IFN-β (red) in lungs of SARS-CoV–infected macaques at day 1 post infection (40× [B] and 100× [C], arrowheads). (D) Lack of phosphorylated STAT1 in lungs of mock-infected macaques (40×) and (E) abundant presence of phosphorylated STAT1 (brown) in lungs of SARS-CoV–infected macaques at day 1 post infection (40×, arrowheads). (F) No detection of phosphorylated STAT1 (brown) in SARS-CoV–infected cells (red) (40×, arrowheads).

Figure 8. Inhibition of STAT1 Phosphorylation and Nuclear Translocation in SARS-CoV–Infected MA104 Cells

MA104 cells were infected with SARS-CoV for 24 h and then either fixed directly (left panel) or treated with type I IFN and then fixed. Subsequently, cells were stained for phosphorylated STAT1 and SARS-CoV. The far right panel shows a merge of the two middle panels, showing an inhibition of STAT1 phosphorylation and translocation in SARS-CoV–infected cells.