Fig.5: Similarities in gene expression and pathways in MHV-1
infected obese vs lean mice and severe vs mild COVID-19 patients.
RNAseq was performed on the RNA collected from blood of MHV-1 and
mock-infected lean and obese C3H/HeJ mice (n = 8 per group) at 2 dpi.(A) Principal component analysis (PCA) of MHV-1 and
mock-infected lean and obese mice. PC1 and PC2 were plotted to identify
clustering of the MHV-1 and mock-infected samples. For identification of
DEGs, genes with p-adjusted values < 0.1 and
log2FoldChange > 1, were selected by
integrated Differential Expression and Pathway (iDEP .96) analysis tool.Upregulated (B) and downregulated (C) pathways unique to MHV-1
infected obese vs lean mice. For this comparison, gene expression in
mock-infected lean or obese mice was removed from their MHV-1 infected
counterparts and the resulting data was compared between obese and lean
mice to identify DEGs unique to obese infected mice. For pathway
analysis, DEGs with p-adjusted values < 0.1 and
log2FoldChange > 1 were selected and
pathways were generated using the ShinyGO 0.77 analysis tool.
The pathways obtained were then
compared to data from the blood of COVID-19 patients with severe
disease; orange arrows indicate common upregulated and blue arrows
indicate common downregulated pathways between MHV-1 infected obese vs
lean mice and severe vs mild COVID-19 patients.
TNF has been shown to damage the respiratory epithelium and stimulate
the synthesis of fibrin and collagen in the respiratory
system57,58. Although the direct role of TNF in severe
COVID-19 has not yet been established, it has been associated with heart
failure59, blood clotting60-63, and
is elevated in patients with obesity64 and
hypertension65, all of which are risk factors for
severe COVID-19. The NLR signaling pathway could potentially be linked
to COVID-19 related multiple sclerosis (MS)66considering that the NLRP3 inflammasome has been implicated in COVID-19
pathogenesis67. Platelet abnormalities in COVID-19
patients have been associated with disease severity and
mortality68 related to organ
failure69. The underlying mechanisms leading to severe
COVID-19 symptoms, such as acute respiratory distress syndrome (ARDS),
cytokine storm and thrombotic events, have been associated with
neutrophils, particularly increased formation of neutrophil
extracellular traps (NETs)70. Excessive NET formation
in SARS-CoV-2 infection has been associated with the onset of acute lung
injury (ALI)71. The Rap1 signaling pathway has been
implicated in the pathogenesis of respiratory, cardiovascular, and
nervous system and some studies have shown that activation of the Rap1
signaling pathway can improve the symptoms of pulmonary
fibrosis72.
T cells are recognized as potent immune modulators against various
infections by activating cytotoxic or humoral immune-mediated reactions
and are mainly divided into T helper type 1 (Th1), T helper type 2
(Th2), T helper type 17 (Th17) and regulatory T (Treg)
cells73. The equilibrium between Th1 and Th2 responses
has been linked to COVID-19 outcomes, with an adequate Th1 response
being associated with good prognosis and Th2 activation being linked to
worse disease outcomes74,75. Th17 has been associated
with chronic inflammation and autoimmune diseases and may be involved in
SARS-CoV-2 induced pneumonia76. Similarly, the other
downregulated pathways involved hematopoietic cell lineage, NK
cell-mediated cytotoxicity and primary immunodeficiency, suggesting
lowered lymphocyte immune responses in MHV-1-infected mice and COVID-19
patients.
We also compared expression of individual DEGs unique to MHV-1 infected
obese vs. lean mice with expression patterns in severe vs mild COVID-19
patients and found 20 upregulated and 23 downregulated genes shared
between MHV-1 and COVID-19 (Supplementary Table 1). In particular, MHV-1
infected obese mice exhibited upregulation of
B4galt577, Flt178,
Stx379, and Ass180 and
downregulation of Crip277 and Rora81changes (Supplementary Fig.S2) which have also been associated with
severe disease outcomes in COVID-19 patients. Strikingly, the
similarities observed in the immune pathways and gene expression
patterns in our MHV-1 infected obese vs lean mice suggest that our
coronavirus infection model mimics COVID-19 in humans and could be a
potentially relevant model to identify biomarkers of severe disease
outcome in SARS-CoV-2 infected individuals. However, future mechanistic
studies are needed to elucidate the precise roles of these biomarkers in
coronavirus disease outcomes.
Gene Set Variation Analysis (GSVA) showed significant
differences between MHV-1 infected lean and obese C3H/HeJ mice. We
further utilized our RNAseq data to carry out Gene Set Variation
Analysis (GSVA) and compare the impact of MHV-1 infection and obesity on
differential enrichment of disease-relevant immune cell, inflammatory,
and cellular pathway gene signatures12,33,82 (Fig.6A
& B). Several immune cell type signatures associated with COVID-19
patient disease pathology were also changed in MHV-1 infected mice,
including enrichment of inflammatory neutrophils, low density
granulocytes (LDGs), monocytes, and activated B cells as well as
de-enrichment of T cells (Fig.6A). Pathway gene signatures representing
response to inflammatory cytokines, such as interferons (IFN) and TNF,
as well as metabolic pathways, including glycolysis and lipid metabolism
were also elevated in infected mice (Fig.6B). Gene expression changes
associated with obesity in the context of infection were much more
subtle. However, we did find that the unfolded protein and lipid
metabolism gene signatures were significantly increased suggesting that
dysregulated cell stress and metabolism pathways could be involved in
exacerbating disease outcomes in obese mice (Fig.6B).
In order to relate disease pathology with changes in gene expression, we
carried out linear regression analysis using the inflammatory neutrophil
gene signature, which was derived from gene expression studies of severe
COVID-19 patients83, and clinical features. This
revealed a significant correlation between the inflammatory neutrophil
GSVA enrichment score and weight loss as well as lung and liver tissue
pathology scores (Fig.6C-E) suggesting inflammatory neutrophils may also
be contributing to disease severity in MHV-1 infected mice. In contrast,
the activated B cell gene signature was significantly correlated with
overall weight loss (Fig.6F), but not tissue pathology scores (Fig.6G-H)
indicating that inflammatory neutrophils and not B cells were associated
with tissue damage in infected mice. Overall, we have demonstrated the
utility of MHV-1 infected C3H/HeJ mice as a model for human COVID-19 as
well as the effect of diet-induced obesity on disease pathogenesis.