Hypoxia induces expression of angiotensin-converting enzyme II in alveolar epithelial cells: Implications for the pathogenesis of acute lung injury in COVID-19

Abstract

SARS-CoV-2 uptake by lung epithelial cells is a critical step in the pathogenesis of COVID-19. Viral entry is dependent on the binding of the viral spike protein to the angiotensin converting enzyme II protein (ACE2) on the host cell surface, followed by proteolytic cleavage by a host serine protease such as TMPRSS2. Infection of alveolar epithelial cells (AEC) in the distal lung is a key feature in progression to the acute respiratory distress syndrome (ARDS). We hypothesized that AEC expression of ACE2 is induced by hypoxia. In a murine model of hypoxic stress (12% FiO2), the total lung Ace2 mRNA and protein expression was significantly increased after 24 hours in hypoxia compared to normoxia (21% FiO2). In experiments with primary murine type II AEC, we found that exposure to hypoxia either in vivo (prior to isolation) or in vitro resulted in greatly increased AEC expression of both Ace2 (mRNA and protein) and of Tmprss2. However, when isolated type II AEC were maintained in culture over 5 days, with loss of type II cell characteristics and induction of type I cell features, Ace2 expression was greatly reduced, suggesting that this expression was a feature of only this subset of AEC. Finally, in primary human small airway epithelial cells (SAEC), ACE2 mRNA and protein expression were also induced by hypoxia, as was binding to purified spike protein. Hypoxia-induced increase in ACE2 expression in type II AEC may provide an explanation of the extended temporal course of human patients who develop ARDS in COVID-19.

FIGURE 1

Effect of in vivo hypoxic exposure on lung ACE2 and TMPRSS2 expression. Mice were exposed to room air (21% oxygen) or hypoxia (12% oxygen) for 48 h. Relative mRNA expression in whole lung homogenates (Figure 1a) or in type II AEC ex vivo (Figure 1c) was determined by RT‐PCR and are expressed relative to the normoxic control. Ace2 protein expression in lung homogenates was determined by ELISA (Figure 1b). Boxplots (McGill et al., 1978) show data representative of two independent (n = 3) experiments; p‐values show differences between paired boxplots

FIGURE 2

Effect of in vitro hypoxic exposure in vitro on mouse primary AEC expression of ACE2 and TMPRSS2. AEC cultures were exposed to 1% oxygen, 21% oxygen, or DMOG (1 µM, in 21% oxygen) for 24 h. Relative mRNA expression of Ace2 and Tmprss2 was determined by RT‐PCR and expressed relative to the normoxic control (Figure 2a). Ace2 protein expression was determined in AEC extracts by ELISA (Figure 2b). Boxplots (McGill et al., 1978) show data representative of 3 independent (n = 3) experiments; p‐values from multivariable linear regressions are from comparisons with control conditions

FIGURE 3

Changes in expression of AEC phenotypic markers, Ace2 and Tmprss2 over time in culture. Relative mRNA expression of the pan‐epithelial marker, E‐cadherin (Figure 3a) and primary murine AEC phenotypic markers for type II (Figure 3b‐d) and type I (Figure 3e‐f) AEC was determined by RT‐PCR in cells maintained in 21% oxygen over 5 days in culture. Expression of Ace2 and Tmprss2 mRNA are shown in Figure 3g, h, respectively. Boxplots (McGill et al., 1978) show data representative of two independent (n = 3) experiments. Linear regression slopes and p‐values are shown with each transcript

FIGURE 4

Effect of hypoxia exposure on expression of ACE2 in human SAEC. Relative mRNA expression for ACE2 and TMPRSS2 was determined by RT‐PCR in SAEC exposed to 1% oxygen, 21% oxygen or DMOG (1 µM, in 21% oxygen) for 24 h (Figure 4a). Boxplots (McGill et al., 1978) show data representative of three independent (n = 3) experiments, each using cells derived from a different donor; p‐values are from multivariable linear regressions with control conditions. ACE2 protein levels (Figure 4b) were determined by ELISA (R&D Systems) using whole cell extracts of SAEC (10⁶ cells/sample) after 24 h exposure to 1% or 21% oxygen. Boxplots show data representative of two independent (n = 3) experiments; p‐values are from t‐test comparisons of paired groups. Following 24 h exposure to 1% or 21% oxygen, the amount of ACE2 able to bind immobilized COVID‐19 spike protein was determined in whole cell extracts from SAEC (Figure 4c). Boxplots show data from an average of two samples per condition; the t‐test p‐value is shown. A similar result was obtained using SAEC cells from a second donor