The ACE2 gene, mapped to chromosome locus Xp22.2, encodes angiotensin-converting enzyme 2 (ACE2, UniProt ID: Q9BYF1), comprising 805 amino acid residues (~120 kDa mass) and functioning as a zinc-metalloproteinase type 1 transmembrane protein1–4. ACE2 is a critical regulator of the renin-angiotensin-system in the cardiovascular system, where it counteracts increases in blood pressure through metabolism of angiotensin peptides3,5−7. Additionally, ACE2 has been shown to cleave the apelin peptide [Pyr1]apelin-13 to [Pyr1]apelin-13(1−12), which is expressed and functional as a potent vasoactive agent and positive cardiac inotrope in the cardiovascular system8. ACE2 also acts as a chaperone for epithelial and brush-border expression of the neutral amino acid transporter, B0AT1 (SLC6A19), in proximal tubules of the kidney and the small intestine respectively9–11.
Surprisingly, ACE2 is exploited as a host cell surface receptor for entry of viruses such as the severe acute respiratory syndrome coronavirus (SARS-CoV) and human corona virus NL63 (HCoV-NL63), where ACE2 has been shown to efficiently bind the S1 domain of coronavirus spike (S) proteins3,12−14. The ongoing corona virus disease 2019 (COVID-19) pandemic, that has put huge strain on health care services across the globe, is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). As is the case for SARS-CoV and HCoV-NL63, ACE2 is able to bind the SARS-CoV-2 spike (S) protein with high affinity (~15 nM) to facilitate host cell infection, and remains a critical therapeutic target that may need to be exploited to combat the virus and its emerging variants15–19. ACE2, long seen as a beneficial protein in the protective arm of the renin-angiotensin-system, is therefore a negative mediator of viral infection that has been reviewed extensively in the COVID-19 pandemic20–22. Interestingly, the ACE2-B0AT1 heterodimer complex was shown using cryogenic electron microscopy to be able to bind two SARS-CoV-2 spike proteins simultaneously16. Recently, both ACE2 and B0AT1 have been targeted pharmacologically, with the compounds DX600 and benztropine respectively, to reduce SARS-CoV-2 spike dependent viral infection in a human embryonic stem cell derived cardiomyocyte model23.
Structural modelling identifies several key residues in the extracellular domain of ACE2 that interact with SARS-CoV-2 spike protein17. The authors suggest that G474, Q498, T500, and N501 of the receptor binding domain of spike forms a network of H-bonds with G24, Y41, G42, K353, and R357 of ACE2. K417 and Y453 of the spike receptor binding domain also interact with D30 and H34 of ACE2. F486 of the spike protein may also interact with M82 through van der Waals forces. Due to a lack of selection pressure prior to the COVID-19 pandemic, ACE2 population variants are surprisingly rare4,24. Several ACE2 polymorphisms have been identified that are associated with hypertension25–28, but it remains unclear whether these have an impact on SARS-CoV-2 interaction. Several genomic studies confirm few natural resistance mutations in ACE2 exist, with many of the identified variants exhibiting similar binding affinity for SARS-CoV-2 spike and some ACE2 variants (such as I21V, E23K, K26R, T27A, N64K, T92I, Q102P, and H378R) that are even predicted to confer increased susceptibility to SARS-CoV-224,29. Interestingly however, the alternate allele (T or A) of ACE2 rs2285666 correlated with lower infection and case-fatality rate among Indian populations30 and ACE2 alleles rs73635825 (S19P) and rs143936283 (E329G) showed noticeable variations in their intermolecular interactions with the viral spike protein29. Host cell receptor variation, such as the Δ32 variant in C-C chemokine receptor type 5 (CCR5) that confers resistance to strains of HIV31, is a critical concept in understanding viral entry and pharmacological intervention – and ACE2 variation will need to be studied further in relation to SARS-CoV-2 infection.
ACE2 shows relatively wide tissue distribution. In original work, particularly high mRNA expression was observed in the kidney, testis, and heart1,2, where ACE2 was localised to the vascular endothelium, smooth muscle, myofibroblasts, and the myocytes themselves32,33. QRT-PCR performed in 72 tissue types confirmed expression in cardiovascular tissues but also showed high abundance of mRNA in the gastrointestinal system, particularly the intestines34. Protein expression has been identified in the vascular endothelium and smooth muscle, lung alveolar epithelial cells, intestinal enterocytes35, and also in cardiomyocytes, gall bladder, and renal tubules36. Interestingly, expression in the respiratory tract has been shown to be relatively low overall, and restricted to certain structures and subsets of cells – chiefly cells of the sinonasal cavity and alveolar type II cells35–38. The localisation of ACE2 at nasopharyngeal, lung, and gastrointestinal tract epithelia falls in line with the proposed entry routes for SARS-CoV-2 infection37,39−41. Importantly, as ACE2 resides in the X chromosome, females show higher overall expression or ‘gene dosing’ of the protein42, which paradoxically may contribute to the reduced susceptibility to SARS-CoV-2 symptoms and mortality versus males observed globally in COVID-19 cases43,44.
To date, two recent reports provide further insight into the spectrum of responses of individuals, ranging from those that are asymptomatic, to those with severe illness and long-term effects such as Long COVID that affects the heart. The papers describe a novel isoform deltaACE2 (herein referred to as dACE2) that is upregulated by interferon stimulation and rhinovirus infection but not SARS-CoV-245,46. This short isoform, comprising amino acids 357-805 of ACE2 with a 10 amino acid insert at the N-terminus, lacks both a fully functional enzyme catalytic site and high affinity spike S1 binding sites, and is observed in airway epithelia and squamous tumours of the respiratory, gastrointestinal, and urogenital tracts. Both studies conclude the short isoform is unlikely to confer host susceptibility to infection by SARS-CoV-2, and that it is this isoform, not regular ACE2, that is upregulated in response to interferon or rhinovirus infection. Whilst these studies focused primarily on ACE2 and dACE2 mRNA, the protein distribution has not been extensively mapped.
We hypothesise that the tissue distribution of the full length ACE2 versus non-virus binding dACE2 will be distinct between organs and specific tissue beds, and may give insight as to why susceptibility to infection varies. Our aim was to use immunohistochemistry to compare the protein distribution of the ACE2 isoforms in the major organs in humans that show SARS-CoV-2 dependent viral damage in post-mortem assessment. We find both isoforms are localised in multiple tissue types in humans but protein expression of dACE2 is enriched in the epithelial cells of the lungs and bile duct. We speculate that these tissues could be used as systems to identify mechanisms that alter short-term dACE2 expression and, more widely, potentially reduce infection. Additionally, we confirm low binding of a fluorescently tagged SARS-CoV-2 spike protein monomer in lung airway and bile duct epithelia where dACE2 was enriched.