Angiotensin-converting enzyme 2 protects from lethal avian influenza A H5N1 infections

Abstract

The potential for avian influenza H5N1 outbreaks has increased in recent years. Thus, it is paramount to develop novel strategies to alleviate death rates. Here we show that avian influenza A H5N1-infected patients exhibit markedly increased serum levels of angiotensin II. High serum levels of angiotensin II appear to be linked to the severity and lethality of infection, at least in some patients. In experimental mouse models, infection with highly pathogenic avian influenza A H5N1 virus results in downregulation of angiotensin-converting enzyme 2 (ACE2) expression in the lung and increased serum angiotensin II levels. Genetic inactivation of ACE2 causes severe lung injury in H5N1-challenged mice, confirming a role of ACE2 in H5N1-induced lung pathologies. Administration of recombinant human ACE2 ameliorates avian influenza H5N1 virus-induced lung injury in mice. Our data link H5N1 virus-induced acute lung failure to ACE2 and provide a potential treatment strategy to address future flu pandemics.

Figure 1. Elevated serum angiotensin II levels in H5N1-infected patients.

(a) Angiotensin II levels in the sera of healthy volunteers (Controls, n=10) and H5N1-infected patients (n=26) divided into three groups according to plasma sampling times relative to the disease onset. Angiotensin II serum levels were determined using ELISA; bars represent the median value. **P<0.01 (Mann-Whitney U test). (b) Kinetics of angiotensin II serum levels in three H5N1-infected patients (patients 4, 13, and 14). Angiotensin II levels were determined using ELISA at the indicated time points after the onset of flu symptoms.

Figure 2. Downregulated ACE2 expression and increased serum angiotensin II levels in mice infected with live H5N1 virus.

(a) Western blots of total lung samples obtained 24 h after intratracheal instillation of vehicle control (allantoic fluid (AF)), live H1N1 viruses or live H5N1 viruses (10⁶ TCID₅₀ in both cases). Blots are representative of three different mice for each treatment (full blots are provided as Supplementary Fig. 7). Bar graphs present quantitative analyses of ACE and ACE2 protein levels, which are presented as the mean ACE- and ACE2-to-β-actin ratios±s.e.m. (n=3 mice per treatment). **P<0.01 (two-tailed t-test). (b) Angiotensin (Ang) II levels in the serum of vehicle control and virus-challenged mice 72 h after virus or vehicle instillation. Angiotensin II levels were determined using radioimmunoassays (data are shown as mean±s.e.m.; n=5-6 mice per group). **P<0.01 (two-tailed t-test). (c) Angiotensin II levels in sera on day 3 (72 h) or day 5 (120 hrs) after infection with live H5N1 virus (10⁶ TCID₅₀). Angiotensin II levels were determined using radioimmunoassays. Death group, n=14, Survival group, n=4, data are shown as mean±s.e.m., *P<0.05 (two-tailed t-test). Each experiment was repeated at least three times.

Figure 3. ACE2 deficiency increases the severity of H5N1-induced acute lung injury.

Wild-type (WT) and ACE2-knockout (ACE2 KO) mice were intratracheally instilled with vehicle control (allantoic fluid, AF) or live H5N1 virus (10⁶ TCID₅₀). (a) Kaplan-Meier survival curves were recorded. n=9–10 mice per group. *P<0.05 (log-rank test) when comparing the (ACE2 KO+H5N1) group with the (WT+H5N1) group. (b) Virus titers in wild-type or ACE2 KO mouse lung were assessed 1, 3 and 5 days after H5N1 infection, n=3–5 mice per time point (mean±s.e.m.). *P<0.05, **P<0.01 (two-tailed t-test). (c) Relative mRNA expression levels of the influenza A matrix 2 (M2) gene in the lungs of wild type and ACE2 knockout mice at an indicated time were determined by real-time PCR, n=3–5 mice per time point (mean±s.e.m.). **P<0.01 (two-tailed t-test). (d) Wet to dry weight ratios of the lungs analyzed 3 days after intratracheal instillation of AF as a control or live H5N1 virus. n=4–6 mice per group (mean±s.e.m.). **P<0.01 (two-tailed t-test). (e) Representative images of the lung pathology of WT and ACE2 KO mice 3 days after intratracheal instillation of AF or live H5N1 virus. Scale bar=100 μm. The numbers of infiltrating cells per microscopic field (mean±s.e.m.; top panel) and lung injury scores (mean±s.e.m., bottom panel) are shown in the bar graphs. n=100 fields analyzed for three mice for each treatment. *P<0.05, **P<0.01 (two-tailed t-test). Each experiment was repeated at least twice.

Figure 4. Recombinant hACE2 reduces the severity of H5N1-induced acute lung injury.

BALB/c mice were injected with recombinant hACE2 (i.p., 100 μg kg⁻¹) or allantoic fluid (AF) as a vehicle control (+BSA i.p., 100 μg kg⁻¹) 3 h prior to, as well as 8 h and 3 days after intratracheal instillation of AF or live H5N1 virus (10⁶ TCID₅₀). (a) Kaplan-Meier survival curves. n=21 for the Control+H5N1 and ACE2+H5N1 groups, n=5 for the Control+AF and ACE2+AF groups. *P<0.05 when comparing the Control+H5N1 group with the ACE2+H5N1 group (Log-rank test). (b) Relative mRNA expression levels of influenza A M2 in the lungs of mice at the indicated time points using real-time PCR (mean±s.e.m.). n=3–5 mice per time point. **P<0.01 (two-tailed t-test). (c) Angiotensin II levels in the serum of mice were determined using radioimmunoassays at the indicated time points (mean±s.e.m.). n=3-4 mice per time point. *P<0.05 (two-tailed t-test). (d) Wet to dry lung weight ratios of the mice were assessed 4 days after AF or H5N1 instillation in the presence or absence of recombinant hACE2 (mean±s.e.m.). n=4-5 mice per group. *P<0.05, **P<0.01 (two-tailed t-test). (e) Representative lung histopathology of AF and H5N1-challenged BALB/c mice left untreated or treated with recombinant ACE2 (i.p., 100 μg kg⁻¹). Scale bar=100 μm. The numbers of infiltrating cells per microscopic field (mean±s.e.m.; top panel) and lung injury scores (mean±s.e.m.; bottom panel) are shown for day 4 after infection in the bar graphs. n=100 fields analyzed for three mice for each treatment. *P<0.05, **P<0.01 (two-tailed t-test). Each experiment was repeated at least three times.