Comparative analysis of ACE2 protein expression in rodent, non-human primate, and human respiratory tract at baseline and after injury: A conundrum for COVID-19 pathogenesis

Abstract

Angiotensin converting enzyme 2 (ACE2) is the putative functional receptor for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Current literature on the abundance and distribution of ACE2 protein in the human respiratory tract is controversial. We examined the effect of age and lung injury on ACE2 protein expression in rodent and non-human primate (NHP) models. We also examined ACE2 expression in human tissues with and without coronavirus disease 19 (COVID-19). ACE2 expression was detected at very low levels in preterm, but was absent in full-term and adult NHP lung homogenates. This pattern of ACE2 expression contrasted with that of transmembrane protease serine type 2 (TMPRSS2), which was significantly increased in full-term newborn and adult NHP lungs compared to preterm NHP lungs. ACE2 expression was not detected in NHP lungs with cigarette smoke-induced airway disease or bronchopulmonary dysplasia. Murine lungs lacked basal ACE2 immunoreactivity, but responded to hyperoxia, bacterial infection, and allergen exposure with new ACE2 expression in bronchial epithelial cells. In human specimens, robust ACE2 immunoreactivity was detected in ciliated epithelial cells in paranasal sinus specimens, while ACE2 expression was detected only in rare type 2 alveolar epithelial cells in control lungs. In autopsy specimens from patients with COVID-19 pneumonia, ACE2 was detected in rare ciliated epithelial and endothelial cells in the trachea, but not in the lung. There was robust expression of ACE2 expression in F344/N rat nasal mucosa and lung specimens, which authentically recapitulated the ACE2 expression pattern in human paranasal sinus specimens. Thus, ACE2 protein expression demonstrates a significant gradient between upper and lower respiratory tract in humans and is scarce in the lung. This pattern of ACE2 expression supports the notion of sinonasal epithelium being the main entry site for SARS-CoV-2 but raises further questions on the pathogenesis and cellular targets of SARS-CoV-2 in COVID-19 pneumonia.

Fig 1. ACE2 expression is not detected at baseline and not induced by cigarette-smoke exposure or bronchopulmonary dysplasia in adult NHP lungs.

Representative immunoblots with kidney and lung homogenates (30 mg/lane) from C. macaque model of cigarette-smoke-induced airway injury; identical membranes were probed with 3 different ACE2 antibodies as indicated on the image. K, kidney; L, lung (A). Representative image of lung section from C. macaque model of cigarette-smoke-induced airway injury stained with Alcian Blue (AB) and H&E demonstrating abundant goblet cells in the bronchial epithelium (B), but lack of ACE2 immunoreactivity in bronchial (C) and alveolar epithelial cells (D). A human kidney section was used as a positive control and demonstrates abundant ACE2 expression in proximal tubular epithelial cells (E).

Fig 2. Expression of ACE2 is higher and TMPRSS2 is lower in preterm lungs compared to full-term and adult lungs.

Representative immunoblots and densitometry of NHP (P. papio) lung homogenates from preterm (125d and 140d) and full-term newborn and adult animals demonstrate faint ACE2 expression in the 125d and 140d groups, but not in adult samples (A & B), whereas TMPRSS2 expression is significantly higher in adult lungs compared to other groups (A & C). Representative images of IHC for ACE2 in NHP (P. papio) lung tissues from preterm (125d) (D) and adult animals (E) as well as NHP lungs with BPD (F) demonstrate lack of ACE2 immunoreactivity. * p < 0.05, ** p < 0.01.

Fig 3. Robust ACE2 detection in ciliated epithelial cells in the upper but not lower airways in human specimens.

Representative images of ACE2 IHC on control human paranasal sinus section showing ACE2 immunoreactivity in ciliated epithelial cells (A) and submucosal glandular epithelial cells (B). Double IF analysis confirmed co-localization of ACE2 in the ciliated epithelial cells with α-tubulin (C). Representative IHC for pro-SPC (D) and ACE2 (E) on control human lung sections; black arrows indicate pro-SPC positive AT2s in panel “D” and ACE2-positive AT2s in panel “E”. Representative IHC for ACE2 on an autopsy lung section (F) and trachea (G) from patients who died of COVID-19. Black arrows indicate ACE2-positive ciliated epithelial cells and black arrowhead indicates a capillary blood vessel with ACE2 immunoreactivity in the lamina propria. Representative IHC for ACE2 (H) and double IF for ACE2 and CD31 (I) on control heart sections; yellow arrow indicates a capillary blood vessel with ACE2 and CD31 co-localization and white arrows indicate a larger blood vessel without ACE2 signal.

Fig 4. ACE2 is constitutively expressed in F433/N rat nasal mucosa and lungs, and induced in murine lungs with injury.

Representative IHC for ACE2 in adult murine kidney (A), intestine (B), and lung (C). Representative IHC for ACE2 (J) and double IF for ACE2 and α-tubulin on F433/N rat nasal mucosa sections (K) and IHC for ACE2 on F433/N lung section (L). Immunoblotting and densitometry of murine whole lung homogenates from newborn and adult mice demonstrate TMPRSS2, but not ACE2 expression in adult samples (D & E). Representative adult murine lung sections with acute lung injury induced by hyperoxia exposure (F) and P. aeruginosa pneumonia (G), and ovalbumin-induced allargic airway inflammation (H) demonstrate diffuse ACE2 immunoreactivity in bronchial epithelial cells. Representative IHC for ACE2 (I) and double IF for ACE2 and α-tubulin on F433/N rat nasal mucosa sections (J) and IHC for ACE2 on F433/N lung section (K).