A Time-Series Metabolomic Analysis of SARS-CoV-2 Infection in a Ferret Model

Abstract

The global threat of COVID-19 has led to an increased use of metabolomics to study SARS-CoV-2 infections in animals and humans. In spite of these efforts, however, understanding the metabolome of SARS-CoV-2 during an infection remains difficult and incomplete. In this study, metabolic responses to a SAS-CoV-2 challenge experiment were studied in nasal washes collected from an asymptomatic ferret model (n = 20) at different time points before and after infection using an LC-MS-based metabolomics approach. A multivariate analysis of the nasal wash metabolome data revealed several statistically significant features. Despite no effects of sex or interaction between sex and time on the time course of SARS-CoV-2 infection, 16 metabolites were significantly different at all time points post-infection. Among these altered metabolites, the relative abundance of taurine was elevated post-infection, which could be an indication of hepatotoxicity, while the accumulation of sialic acids could indicate SARS-CoV-2 invasion. Enrichment analysis identified several pathways influenced by SARS-CoV-2 infection. Of these, sugar, glycan, and amino acid metabolisms were the key altered pathways in the upper respiratory channel during infection. These findings provide some new insights into the progression of SARS-CoV-2 infection in ferrets at the metabolic level, which could be useful for the development of early clinical diagnosis tools and new or repurposed drug therapies.

Keywords: COVID19; SARS-CoV-2; animal models; ferret; host metabolic responses; metabolomics; omics.

Figure 1

Viral RNA shedding in nasal washes collected from SARS-CoV-2-infected ferrets (n = 20). Red and green dots represent female (n = 10) and male ferrets (n = 10), respectively, while all ferrets (both males and females) are annotated in blue. SARS-CoV-2 RNA was detected in nasal wash samples from ferrets at 3, 5, 7, and 9 dpi. The dotted line represents the limit of detection (LOD) of the reverse-transcription qPCR assay.

Figure 2

Orthogonal partial least square-discriminant analysis (OPLS-DA) of the central carbon metabolism metabolite dataset of nasal wash samples collected from ferrets. (A) OPLS-DA scatter plot and (B) OPLS-DA loadings plot. For this plot, R²X (cum) = 0.453, R²Y (cum) = 0.304, Q² = 0.108. The ellipse presented in panel (A) represents Hotelling’s T2 confidence limit (95%). The colored circles in panel (A) represent each analyzed sample, while the black crossed circles in panel (B) indicate the average group position for each sample cluster, with the white circles representing the distribution of metabolite features between these groups.

Figure 3

Relevant pathways identified using the central carbon metabolism dataset in nasal washes collected from ferrets. The red-colored circles represent significantly relevant pathways, while the white circles represent the non-significant pathways. Noting, 1: Pentose phosphate pathway, 2: Pentose and glucuronate interconversions, 3: Arginine biosynthesis, 4: Starch and sucrose metabolism, 5: D-Glutamine and D-glutamate metabolism, 6: Alanine, aspartate and glutamate metabolism, 7: Citrate cycle (TCA cycle), 8: Butanoate metabolism, 9: Valine, leucine and isoleucine biosynthesis, 10: Amino sugar and nucleotide sugar metabolism, 11: Phenylalanine, tyrosine and tryptophan biosynthesis, and 12: Glyoxylate and dicarboxylate metabolism.

Figure 4

Time-series observed for metabolites related to the 13 significantly relevant central carbon metabolic pathways. (A) Pentose phosphate pathway, (B) Pentose and glucuronate interconversions, (C) arginine biosynthesis, (D) starch and sucrose metabolism, (E) D-glutamine and D-glutamate metabolism, (F) glyoxylate and dicarboxylate metabolism, (G) citrate cycle, (H) phenylalanine, tyrosine and tryptophan biosynthesis, (I) amino sugar and nucleotide sugar metabolism, (J) alanine, aspartate and glutamate metabolism, (K) butanoate metabolism, (L) valine, leucine and isoleucine biosynthesis, and (M) nicotinate and nicotinamide metabolism.

Figure 5

Heat map representing the significant metabolites that were found to significantly vary identified using two-way analysis of variance.