Diet-Induced Obesity and NASH Impair Disease Recovery in SARS-CoV-2-Infected Golden Hamsters

Abstract

Obese patients with non-alcoholic steatohepatitis (NASH) are prone to severe forms of COVID-19. There is an urgent need for new treatments that lower the severity of COVID-19 in this vulnerable population. To better replicate the human context, we set up a diet-induced model of obesity associated with dyslipidemia and NASH in the golden hamster (known to be a relevant preclinical model of COVID-19). A 20-week, free-choice diet induces obesity, dyslipidemia, and NASH (liver inflammation and fibrosis) in golden hamsters. Obese NASH hamsters have higher blood and pulmonary levels of inflammatory cytokines. In the early stages of a SARS-CoV-2 infection, the lung viral load and inflammation levels were similar in lean hamsters and obese NASH hamsters. However, obese NASH hamsters showed worse recovery (i.e., less resolution of lung inflammation 10 days post-infection (dpi) and lower body weight recovery on dpi 25). Obese NASH hamsters also exhibited higher levels of pulmonary fibrosis on dpi 25. Unlike lean animals, obese NASH hamsters infected with SARS-CoV-2 presented long-lasting dyslipidemia and systemic inflammation. Relative to lean controls, obese NASH hamsters had lower serum levels of angiotensin-converting enzyme 2 activity and higher serum levels of angiotensin II-a component known to favor inflammation and fibrosis. Even though the SARS-CoV-2 infection resulted in early weight loss and incomplete body weight recovery, obese NASH hamsters showed sustained liver steatosis, inflammation, hepatocyte ballooning, and marked liver fibrosis on dpi 25. We conclude that diet-induced obesity and NASH impair disease recovery in SARS-CoV-2-infected hamsters. This model might be of value for characterizing the pathophysiologic mechanisms of COVID-19 and evaluating the efficacy of treatments for the severe forms of COVID-19 observed in obese patients with NASH.

Keywords: SARS-CoV-2; coronavirus disease 2019; hamster; non-alcoholic steatohepatitis; obesity.

Figure 1

Free-choice-diet-induced obesity, NASH/liver fibrosis, and lung inflammation. (A–D) Representative liver histology images after H&E staining (A,B) and Sirius Red staining (C,D) of lean hamsters (A,C) and obese NASH hamsters (B,D), respectively. Blue dashed squares indicate inflammatory foci and hepatocyte ballooning, and green arrows indicate liver fibrosis. (E) Cytokeratin-18 (CK18) immunostaining of liver samples from obese NASH hamsters. Black dashed squares indicate clusters of ballooned hepatocytes lacking cytokeratin-18. (F) Hepatocyte ballooning is shown at a higher magnification. (G) Serum levels of MCP-1 and IL-6 and (H) lung levels of IL-6 protein in lean hamsters and obese NASH hamsters. Data are shown as the mean ± SD (n = 10 animals per group). Significant differences were determined using an unpaired, two-tailed Student’s t-test or a Mann–Whitney U test (* p < 0.05; ** p < 0.01; *** p < 0.001). Scale bars: 250 µm (A–F).

Figure 2

Impact of SARS-CoV-2 infection on lean hamsters vs. obese NASH hamsters. Hamsters were inoculated intranasally with 2 × 104 TCID50 of the clinical SARS-CoV-2 isolate hCoV-19/France/lDF0372/2020. (A) Changes in body weight (as the mean ± SD percentage of the initial body weight). Sidak’s correction for multiple comparisons was used to detect significant differences (*** p < 0.001). (B) Determination of the infectious viral load in the lungs on dpi 4 (TCID50 assay). The data are expressed as the number of infectious virus particles per lung. (C) Viral RNA-dependent RNA polymerase (RdRp) transcript levels in the lungs, as quantified in a qRT-PCR (dpi 4 and 7). The data are expressed as the delta Ct (normalized against γ actin). The bottom line corresponds to the detection threshold. The results are expressed as the mean ± SD. (D) Isg15 mRNA transcripts were quantified by qRT-PCR (lung). The data are expressed as the mean ± SD fold increase over the level observed in the mock-treated lean hamsters. It should be noted that there were no significant differences between the mock-treated lean hamsters and mock-treated obese NASH hamsters (day 0) (not shown). (A–D) n = 5–8 animals/group. (E,F) Histopathologic assessment of lung sections from lean hamsters (E) and obese NASH hamsters (F) on dpi 10. Representative images after H&E staining) are depicted. (G,H) Representative lung histology images after Sirius Red staining (25 dpi). (I) The lung histopathologic score following SARS-CoV-2 infection (on dpi 4, 7, 10, and 25) (n = 5–8 animals/group). (J) The percentage of lung tissue labeled by Sirius Red on day 0 and on dpi 25. The data are quoted as the mean ± SD or (panel I) the median (n = 6–11 animals/group). Significant differences were determined using the Mann–Whitney U test (* p < 0.05; ** p < 0.01, *** p < 0.001, **** p <0.0001). The values of the various groups were also compared with those of mock-infected animals (### p < 0.001). Scale bars: 500 µm (E–H).

Figure 3

Effects of SARS-CoV-2 infection on serum variables in lean hamsters vs. obese NASH hamsters. Blood from the mock-infected and SARS-CoV-2-infected hamsters was collected at different timepoints post-infection. Serum levels of (A) ALT, (B) free fatty acids, (C) triglycerides, (D) HDL cholesterol, (E) LDL cholesterol, and (F) MCP-1 on dpi 4, 7, and 25 are shown. The data are expressed as the mean ± SD (n = 6–8 animals/group). Significant differences between lean hamsters and obese NASH hamsters were determined using the Mann–Whitney U test (* p < 0.05; ** p < 0.01, *** p < 0.001, **** p < 0.0001). The effect of SARS-CoV-2 infection is also depicted. The values were compared with those of mock-infected animals using a Kruskal–Wallis test and Dunn’s post-test (# p < 0.05, ## p < 0.01, and ### p < 0.001).

Figure 4

Effects of SARS-CoV-2 infection on serum ACE2, ACE, and the renin-angiotensin system activity levels in lean hamsters vs. obese NASH hamsters. Blood from mock-infected and SARS-CoV-2-infected lean hamsters and obese NASH hamsters were collected at the indicated timepoints post-infection. Serum ACE2 (A), ACE (B), and renin angiotensin (C) activities are depicted. Angiotensin II levels are shown in (D). The data are expressed as the mean ± SD (n = 6–8 animals/group). Significant differences were determined using the Mann–Whitney U test (* p < 0.05; ** p < 0.01, *** p < 0.001). The effect of SARS-CoV-2 infection is also depicted. The values were compared with those of mock-infected animals using a Kruskal–Wallis test and Dunn’s post-test (#p < 0.05, ## p < 0.01, and ### p < 0.001). ACE-S, angiotensin-converting enzyme activity (Ang II/Ang I); PRA-S, serum renin activity (Ang I + Ang II).

Figure 5

Effects of SARS-CoV-2 infection in terms of liver damage and fibrosis in lean hamsters vs. obese NASH hamsters. The livers of mock-infected and SARS-CoV-2-infected hamsters were collected on dpi 4, 10, and 25. (A–F) mRNA copy numbers were quantified by qRT-PCR. The data were expressed as the fold increase over the expression level in mock-treated lean animals. (G–J) Representative liver histology images after H&E staining (G,H) or after Sirius Red staining (I,J) of lean hamsters and obese NASH hamsters. (K) The total NAFLD activity score. (L) The percentage of the liver section stained by Sirius Red on day 0 and dpi 25. Data are shown as the mean ± SD (n = 6–8 animals/group). Significant differences were determined using the Mann–Whitney U test (* p < 0.05; ** p < 0.01, *** p < 0.001). The effect of SARS-CoV-2 infection is also depicted. The values of SARS-CoV-2-infected hamsters were compared with those of mock-infected animals using a Kruskal–Wallis test and Dunn’s post-test (# p < 0.05, ## p < 0.01, and ### p < 0.001). Scale bars: 250 µm (G–J).