key: cord-1054846-gvrwkycm authors: Blume, Cornelia; Jackson, Claire L; Spalluto, Cosma Mirella; Legebeke, Jelmer; Nazlamova, Liliya; Conforti, Franco; Perotin-Collard, Jeanne-Marie; Frank, Martin; Crispin, Max; Coles, Janice; Thompson, James; Ridley, Robert A; Dean, Lareb S N; Loxham, Matthew; Azim, Adnan; Tariq, Kamran; Johnston, David; Skipp, Paul J; Djukanovic, Ratko; Baralle, Diana; McCormick, Chris; Davies, Donna E; Lucas, Jane S; Wheway, Gabrielle; Mennella, Vito title: A novel isoform of ACE2 is expressed in human nasal and bronchial respiratory epithelia and is upregulated in response to RNA respiratory virus infection date: 2020-07-31 journal: bioRxiv DOI: 10.1101/2020.07.31.230870 sha: 2d6552144e9b6f1f959838b94ab2f85b19fb5f6c doc_id: 1054846 cord_uid: gvrwkycm Angiotensin-converting enzyme 2 (ACE2) is the main entry point in the airways for SARS-CoV-2. ACE2 binding to SARS-CoV-2 protein Spike triggers viral fusion with the cell membrane, resulting in viral RNA genome delivery into the host. Despite ACE2’s critical role in SARS-CoV-2 infection, an understanding of ACE2 expression, including in response to viral infection, remains unclear. Until now ACE2 was thought to encode five transcripts and one 805 amino acid protein. Here we identify a novel short isoform of ACE2. Short ACE2 is expressed in the airway epithelium, the main site of SARS-CoV-2 infection; it is substantially upregulated in response to interferon stimulation and RV infection, but not in response to SARS-CoV-2 infection, and it shows differential regulation in asthma patients. This short isoform lacks SARS-CoV-2 spike glycoprotein high-affinity binding sites and altogether, our data are consistent with a model where short ACE2 may influence host susceptibility to SARS-CoV-2 infection. At the time of writing there have been more than 15 million confirmed cases of COVID-19 and more than 600,000 confirmed COVID-19 associated deaths worldwide (WHO; 26 th July 2020). There is therefore an urgent global need to understand the molecular mechanism of infection and disease to identify patients' susceptibility and targets for therapeutic intervention. A key molecule responsible for SARS-CoV-2 viral entry is the metalloprotease angiotensin-converting enzyme 2 (ACE2), a transmembrane protein encoded by the human ACE2 gene. ACE2 consists of 19 exons and encodes five annotated transcripts, two of which encode the same 805 amino acid protein ACE2 (UniProt Q9BYF1) which has a theoretical predicted molecular mass of 92.4 kDa and observed mass of ∼ 120 kDa due to multiple sites of glycosylation of the N-terminus region 1 . ACE2 consists of an N-terminal extracellular domain and a C-terminal 4 membrane anchor domain 1 . The extracellular domain contains a 17-amino acid signal peptide sequence, an N-terminal catalytic metallopeptidase domain with 41% amino acid sequence identity to ACE 1,2 and a C-terminal domain with 48% amino acid sequence identity to renal amino acid transporter collectrin (TMEM27) 3 . The extracellular domain can be shed via cleavage of specific residues in the ferredoxinlike fold domain (neck dimerization domain) of the protein by ADAM17, TMPRSS11D or TMPRSS2 proteases. TMPRSS2 expression and activity has been shown to increase SARS-CoV-2 viral entry 4, 5 . ACE2 is a carboxypeptidase with several known physiological functions. It catalyses the removal of the C-terminal residue from different vasoactive peptides, its key substrate being angiotensin II, which contribute to the renin-angiotensin system, a physiological feedback loop regulating blood pressure, salt and water balance in mammals 1, 2 . In the small intestine ACE2 is co-expressed and interacts with amino acid transporter B(0)AT1 (SLC6A19) at the brush border to form a catalytic complex required for amino acid uptake 6, 7 . By homology to collectrin, ACE2 is also required for glucose homeostasis and pancreatic beta-cell function 8,9 . Interestingly, ACE2 plays an important role in protection from acute lung injury. Ace2 expression is downregulated in mice models of acute lung injury and Ace2 knockout mice show a more severe acute lung injury phenotype 10 . Moreover, improved outcomes are seen in pig models of lung injury in which ACE2 is overexpressed 11 and an activator of ACE2, XNT, can protect against pulmonary hypertension in rat models 12, 13 . Although the molecular mechanism by which ACE2 protects against acute lung injury remains unclear, it is known that the carboxypeptidase function of 5 ACE2 is required to confer this protection and that AT2 (AngII receptor 2) also confers protection 10 . Importantly, ACE2 is the main viral entry point for coronavirus N63, SARS-CoV and SARS-CoV-2, which cause severe-acute respiratory syndromes, the latter being responsible for COVID-19 in humans [14] [15] [16] [17] . ACE2 binds to the S1 domain of trimeric SARS-CoV-2 Spike glycoprotein 18 , and viral entry is dependent upon the extracellular domain of ACE2 being cleaved by TMPRSS2 protease at arg697 and lys716 19 , and the transmembrane domain of ACE2 internalised with the virus via the clathrin-mediated 20, 21 and clathrin-independent 22 endocytosis pathways. ACE2 expression in different tissues is controlled by multiple promoter elements 23 . ACE2 expression is under the control of Ikaros homology activating elements around -516/-481 in the heart 24 and under the control of estrogen responsive elements in adipose tissue 25 . ACE2 expression in human nasal epithelia and lung tissue is under the control of interferon-responsive promoters, with STAT1, STAT3, IRF8, and IRF1 binding sites at −1500-500 bp 26 . Activation of interferon (IFN) responsive genes is an important antiviral defence pathway in humans, and both interferon and influenza exposure lead to an increase in ACE2 expression in human airway 26 most likely reflecting its anti-inflammatory role which serves to protect against acute lung injury following viral infection. Bulk RNA sequencing data 27 detects low-level expression of ACE2 in testis, small intestine, thyroid, colon, kidney, heart left ventricle and atrial appendage, and visceral adipose. Single cell RNA sequencing (scRNAseq) studies show ACE2 6 expression at low levels in airway, cornea, esophagus, ileum, colon, liver, gallbladder, heart, kidney and testis 28 . Using scRNAseq and RNA in situ hybridisation, ACE2 expression in the airways has been observed to be relatively high in nasal epithelium and progressively lower in the bronchial and alveolar regions 29 . Highest expression is seen in goblet and ciliated cells of the nasal epithelium 28 , and ACE2 protein localises to the membrane of motile cilia of respiratory tract epithelia 30 Here we detail the identification of a novel isoform of ACE2, which we name short ACE2, that is expressed in human nasal and bronchial respiratory epithelia, the main site of SARS-CoV-2 infection, and is preferentially expressed in asthmatic bronchial epithelium relative to full length ACE2 (long ACE2). In airway primary cells, short ACE2 is upregulated in response to IFN-beta treatment and RNA respiratory rhinovirus infection, but not SARS-CoV-2. 7 We analysed the expression of ACE2 in airway epithelia in our existing RNAseq datasets from nasal brushings and nasal epithelia cultured at air-liquid interface (ALI) by aligning reads to human genome build 38 using STAR 2-pass mapping 31 and GENCODE v33 gene annotations. We visually analysed mappings to ACE2 using Integrative Genomics Viewer 32 , which identified multiple reads mapping to a genomic region between exon 9 and 10 of the constitutive ACE2 gene build (Figure 1a) . These mappings showed a discrete 3' junction at GRCh38 chrX:15580281, but variable 5' length suggesting a splice junction with downstream exon 10, but no splicing upstream to exon 8. This suggests that a novel unannotated exon exists between exon 9 and 10, and that this exon is the beginning of a novel transcript distinct from full-length ACE2 transcripts ACE2-202 (ENST00000427411.1) or ACE2-201 (ENST00000252519.8) in the airway. Visual assessment of total read mappings to ACE2 gene show approximately double the number of read support to exons 10-19 compared to exons 1-9, further suggesting that a novel shorter transcript of ACE2, which includes a novel exon plus exons 10-19, is expressed at equal or higher levels than longer ACE2 transcripts including exons 1-19 (ACE2-202) or 2-19 (ACE2-201) (Figure 1a) . Assembly of transcriptomes from all samples using SCALLOP tool identified novel transcripts including this novel exon to exon 19 (Figure 1b) . Sashimi plot analysis confirmed splicing between this new exon and downstream exon 10, but complete absence of splicing at the 5' end of the new exon ( Figure 1c) . Analysis of these RNAseq data to identify and filter novel splice junctions with code developed by Cummings et al. 33 also independently detected a novel splice junction at chrX:15580281. Analysis of splice junctions identified by STAR aligner confirmed multiple uniquely-mapped reads to a novel exon/exon boundary removing a novel intron of coordinates GRCh38 chrX:15578316-8 15580280. We call this novel exon 9a. Study of the sequence of the exon 9a/intron boundary showed a strong U1-dependent consensus splice site sequence (AG|GTAAGTA) suggesting that it is a strong splice donor site (Figure 1d ). This splicing event introduces a new in-frame ATG start codon 29 nucleotides upstream of the splice site (Figure 1d) , and a TATA box 148 nucleotides upstream of the splice site (Figure 1d) suggesting that this transcript is protein-coding. Furthermore, a promoter flanking region has been identified at GRCh38 chrX:15581200-15579724 (ENSR00000902026), suggesting active transcription upstream of exon 9a (approx. This suggests that the short form of ACE2 is under independent transcriptional control from full-length ACE2 expression, and that this may be controlled by IFN, AP-1 and NF-kB elements. To confirm expression of this novel transcript we performed RT-PCR using primers specific to exon 1, exon 9a and exon 19 using cDNA from both nasal brushings and differentiated immortalized bronchial epithelial cells BCi-NS1.1 37 (Figure 2a, b) . We performed Sanger sequencing to confirm the identity of these PCR amplicons and confirmed sequences spanning exon 9 and 10, and exon 9a and 10 in the amplicons from the constitutive transcript and novel transcript, respectively (Figure 2c) . To investigate expression of this novel ACE2 transcript relative to full-length ACE2 transcripts (ACE-202 and ACE-201) we extracted the number of reads mapped to exon9a/exon10 junction and reads mapping to exon9/exon10 from the STAR alignment output file and calculated exon 9a inclusion rates relative to inclusion of 9 exon 9 using the following calculation: PSI_new = reads[9a to 10]/(reads[9a to 10] + reads [9 to 10] ). This analysis showed that in nasal epithelia the mean expression level of short and long ACE2 transcripts was 0.745 (reads mapped to exon9a/10 or exon9/10 per million mapped reads) and relative inclusion of exon 9a was 0.763 (st err 0.0829). This is a statistically significant higher level of expression of short ACE2 than long ACE2 in nasal epithelial cells (p<0.05, Student's t-test, n=6). We then designed specific qPCR primers to amplify the short and long transcripts of ACE2 individually, as well as a pair of common primers to amplify both transcripts and quantify total levels of ACE2 expression. Expression of long ACE2 was confirmed in a number of cell lines and primary airway cells, with highest expression being observed in the Caco2 cell line and nasal epithelial cells grown at ALI whose expression was comparable to that observed ex vivo (Figure 3a,b) . With the exception of Caco2 cells, expression of short ACE2 was very low in the cell lines studied, and this contrasted with the airway cells which exhibited high expression of this novel isoform (Figure 3a,b) . As we observed high expression of both ACE2 isoforms in differentiated airway cultures, we assessed its induction during differentiation of nasal epithelial cells grown at ALI in vitro. We observed very low 38 . At all time-points the level of short ACE2 transcript expression was higher than expression of the long ACE2 transcripts, although this was not 1 0 statistically significant. Given reports that bronchial epithelial cells express lower levels of ACE2 than nasal cells 26 , we also compared expression of the long and short isoforms of ACE2 in these two cell types. Consistent with previous reports, total levels of ACE2 were lower in bronchial epithelial cells, which was due to reduced expression of both long and short forms of ACE2 (Figure 3d) . We then investigated whether the shorter transcript of ACE2 is expressed in tissues outside the airway by performing transcript-specific qPCR on cDNA from a multiple tissue control panel. This showed expression of the long transcript of ACE2 in all tissues tested except whole brain (Figure 3e,f) , however the short transcript of ACE2 was only detected robustly in lung, liver and kidney (Figure 3e ,f) and was below detection within the dynamic range of the qPCR (Supplementary Figure 2) in all other tissue types tested. In the lung, short and long ACE2 are expressed at approximately equivalent levels, but in liver and kidney long ACE2 is expressed at a higher level than short ACE2 (Figure 3e ,f). Together these data show that we identified a novel ACE2 transcript that is expressed in liver, kidney, and particularly in the lung and airway suggesting a significant role in this compartment. Having confirmed expression of this novel transcript in multiple cell types we sought to investigate whether this transcript is translated into a protein product. Having identified a TATA box and in-frame ATG start codon in the new exon we predicted that this novel transcript would produce a 459 amino acid protein consisting of Arg357 -Phe805 of the full-length long ACE2 protein isoform plus an additional 10 novel amino acids (aa) before Arg357 (M-R-E-A-G-W-D-K-G-G). This is predicted to To investigate whether an ACE2 protein isoform of the expected size is expressed in airway epithelial cells and other cell types, we performed western blotting cells using multiple antibodies to ACE2 recognising epitopes on different regions of the protein (Figure 4a ). We anticipated that short ACE2 would be detected by antibodies raised to the C-terminal domain (CTD) of ACE2 but not the N-terminal domain (NTD) of ACE2. Using a CTD ACE2 antibody (Abcam 15348), with Vero cells as a control, we observed two bands at 100 and 120kDa, consistent with the presence of glycosylated and non-glycosylated forms of full length ACE2 protein 14 . In nasal epithelial cells, we also observed these two isoforms, and an additional band at ~50kDa, the expected molecular weight of short ACE2. (Figure 4b) . Furthermore, this 50kDa band was at least as intense as full length ACE2, consistent with our qPCR analyses of nasal epithelium. To orthogonally validate antibody specificity, we used an antibody raised against the NTD of ACE2 (Abcam 108252) in nasal epithelial cells (Figure 4b) . Our data show that the 50kDa band is only recognized by the antibody raised against anti-ACE2 CTD domain antibody and not by anti-ACE2 NTD domain antibody as expected. Using a validated ACE2 antibody which recognises epitopes from aa 18-740 (R&D AF933) we also observed the 50kDa band in bronchial epithelial cells grown at ALI (Figure 4b ). 1 2 To examine the localisation of the ACE2 isoforms in airway epithelial cells, we performed immunofluorescent staining of differentiated ALI cultures of primary bronchial epithelial cells with anti-ACE2 antibodies visualised by confocal microscopy (Figure 4c, Supplementary Figure 3) . The antibodies which recognise common epitopes in short and long ACE2 localised mainly to the apical portions of the cells and to motile cilia. Interestingly, ACE2 staining extended further up the cilium than the microtubular axoneme, giving the impression of staining the ciliary tip, whereas, in fact, it stained the full length (Figure 4c, Supplementary Figure 3) . Staining with the third antibody that detects only long ACE2 was too weak to interpret. However, staining with the third antibody that detected only long ACE2 was too weak to interpret. Thus, to further investigate whether short ACE2 localises to cilia as has been reported for full-length ACE2 40 we extracted total protein from BCi-NS1.1 cells hTERT immortalized bronchial cells which differentiate robustly into airway multiciliated cells 37 , deciliated BCi-NS1.1 cells and cilia purified from BCi-NS1.1 cells and analysed by western blot with antibodies specific to the C-terminus of full-length ACE2 (Ab15348, aa788-805). Consistent with our previous experiments, this analysis showed a distinct band around 50kDa detected by the C-terminal ACE2 antibody (Figure 4d) . Notably, the band corresponding to short ACE2 was not enriched in the cilia fraction albeit still present in detectable amounts (Figure 4d) suggesting that it is predominantly localised to the apical membranes of the cells. Densitometry analysis of western blot images confirmed enrichment of long ACE2 in cilia fraction relative to short ACE2 (Figure 4d) . Furthermore, long and short ACE2 were both found in the deciliated cells in a ratio of approximately 1:1 -2:1, suggesting the possibility of heterodimer formation between the two isoforms. To better understand the possible functions of the short isoform of ACE2 we modelled the structure of this short isoform, based on the full-length structure of long ACE2 resolved by cryo-EM (PDB 6M18) 17 . This analysis highlighted the extent of loss of the SARS-CoV-2 binding region in short ACE2, with many residues previously shown to be important for viral binding not present in short ACE2 sequence (Figure 5a -c) 17 . In particular, short ACE2 lacks two entire regions involved in interaction with SARS-CoV-2 spike glycoprotein (aa 30-41 and aa 82-84), including a high-affinity binding site (aa 30-41), but retains part of a third region involved in this interaction 17 . This latter sequence is replaced by the N-terminal specific sequence of short ACE2, which is predicted to form a disordered/helical secondary structure by PEP-fold, compared to the beta sheet present in long ACE2 further modifying the third binding interface to Spike. Short ACE2 however retains the sequences required for cleavage by ADAM17, TMPRSS11D and TMPRSS2, suggesting that can be a substrate of these proteases. Interestingly, the ACE2 residue critical for substrate selectivity to Angiotensin II, Arg514, is missing from short ACE2 suggesting that it may lack catalytic activity toward Angiotensin II. Altogether, this analysis suggests that short ACE is not competent for high affinity binding to SARS-CoV-2 spike protein, but it can be a substrate of host proteases acting on long ACE2 during viral entry. While the majority of the spike binding domain of ACE2 is missing from short ACE2, the neck domain (also called the ferredoxin-like fold domain (residues 616 to 726) which is the most important dimerization interface 17 and transmembrane region are present in short ACE2. This suggests that ACE2 may exist not only as the 1 4 conventional full length ACE2 homodimer, but also as a short ACE2 homodimer and a heterodimer of full length ACE2 and short ACE2 (Supplementary Figure 4) . Since a substantial portion of ACE2 is missing in short ACE2, we then undertook molecular dynamic simulation of short ACE2 dimer to assess its dimerization and stability. Assuming that the short form folds in the homologous parts in the same way as the full length ACE2, molecular dynamic simulation of short ACE2 dimer and analysis of secondary structure elements changes show that this structure is likely a To begin investigating the functional relevance of short ACE2 transcript expression, we first assessed whether it was an IFN stimulated gene. As expected, treatment of bronchial epithelial cells with IFN-beta resulted in upregulation of the IFN-responsive genes, MxA and IP10. We also observed upregulation of total ACE2, which to our surprise was largely due to an effect on short ACE2 rather than long ACE2 ( Figure 6a ). This led us to evaluate the response to viral infection, as it has been reported that ACE2 expression is upregulated in this condition 26 . We exposed nasal epithelial cells grown at ALI to rhinovirus (RV) and harvested cells 24hr after infection. qPCR analysis showed a significant upregulation of both long and short ACE2 RNA expression relative to UV-RV treated control (*: p=<0.001, non-parametric Wilcoxon Signed Rank test, control vs. RV16 n=11) (Figure 6b) . While this is consistent with previously published work showing that ACE2 is upregulated in response to influenza exposure 26 , it is notable that we found that it was the short ACE2 transcript 1 5 which was upregulated more robustly than long ACE2 transcript (around 9-fold increase in expression of short ACE2 compared to around 2.5-fold increase in long ACE2 expression) (Figure 6b) . Parallel experiments using bronchial epithelial cells infected with RV, confirmed induction of short ACE2 but no significant effect on long ACE2 (Figure 6b) . As long ACE2 has been described as a point of entry for SARS-CoV-2, and that It has been reported that patients with asthma have reduced susceptibility to SARS-CoV-2, and that asthma symptoms are not exacerbated by SARS-CoV-2 infection [43] [44] [45] . However, it has been suggested that a subset of asthma patients (Type-2 low asthmatics) who are more at risk of SARS-CoV-2 show higher ACE2 expression in bronchial epithelium, associated with upregulation of viral response genes 46 . 1 6 To explore whether expression of short ACE2 might have a role in SARS-CoV-2 infection in asthma, we investigated expression of the long and short transcripts of ACE2. Transcript-specific qPCR on cDNA from bronchial brushings from healthy controls and patients with severe asthma showed statistically significantly lower expression of total ACE2 and long ACE2 in patients with severe asthma, but no significant difference in short ACE2 expression between groups (Figure 7a) . This difference in the profile of ACE2 expression may be a feature of the disease itself, an effect of treatment, or a combination of both. For example, in asthma there is goblet cell metaplasia resulting in a reduction in the number of ciliated cells 47 which are the site of ACE2 expression while elevated levels of the type 2 cytokine, IL-13, are reported to suppress ACE2 expression 48 . Since all of the severe asthmatic subjects received inhaled or oral corticosteroids as part of the regular care, it is noteworthy that oral or intravenous dexamethasone treatment of mechanically ventilated Covid-19 patients resulted in improved 28-day mortality 49 . While corticosteroids are well known as anti-inflammatory agents, whether they also directly modulate ACE2 expression is unknown. Asthma patients have also been shown to mount a reduced antiviral IFN response following infection with respiratory viruses resulting in disease exacerbation with a risk of hospitalisation or even death [50] [51] [52] [53] [54] [55] . As we observed that short ACE2 was not induced following SARS-CoV-2 infection where IFN expression and signalling are compromised, we examined its induction in differentiated bronchial epithelial cultures from severe asthmatic donors in response to RV infection. When compared to cultures from healthy control donors, bronchial epithelial cells from patients with severe asthma showed less upregulation of total ACE2 and long ACE2, whilst short ACE2 was induced to almost the same levels as in control cultures. (Figure 5c ). As 1 7 we observed previously 54 , IFN lambda was also reduced in cultures from severe asthmatic donors. Together, these data are consistent with a hypothesis that suppression of long ACE2, alongside maintenance of short ACE2 expression is protective against SARS-CoV-2 infection in severe asthmatics. Here we present identification and characterisation of a short 11 exon transcript of human ACE2, consisting of a novel previously unannotated first exon, which we name exon 9a, and exons 10-19 of the long ACE2 transcript. We show that, whilst the long transcript of ACE2 is expressed in multiple tissues, this short ACE2 transcript is expressed exclusively in airways, liver and kidney. We show highest expression in primary respiratory epithelia, most notably in the nasal epithelium where the level of short transcript expression is higher than long ACE2 transcript expression. Expression of both transcripts of ACE2 is dependent on differentiation of epithelial cells, with levels of expression comparable with primary nasal epithelia from days 4 -63 of differentiation at ALI culture. We show that transcription of this short transcript is regulated independently of the long transcript of ACE2, with putative promoter elements identified upstream of the transcriptional start site of the short ACE2 transcript. We confirm that this short transcript is translated into a protein product of around 52kDa, a 459 amino acid isoform of ACE2 which lacks a signal peptide. Modelling suggests that this isoform contains a transmembrane domain, collectrin homology domain and portions of the ACE homology domain but lacks the 356 most N-terminal amino acids of full-length ACE2, with 10 novel amino acids in their place. Most of the key residues required for SARS-CoV-2 spike binding are absent from this isoform and we hypothesise that short ACE2 is not a viral entry 1 8 point for SARS-CoV-2. Molecular dynamic simulation suggests that this short isoform can form thermodynamically stable dimers, but the function of this short form of ACE2 remains unclear. Given that the short isoform of ACE2 retains its transmembrane domain and collectrin domain it seems possible that the short isoform may be functionally involved in regulation of amino acid transport in the airway. The short form also retains the catalytic residues conferring carboxypeptidase function, suggesting that the short isoform may retain some catalytic activity, albeit likely with different substrate specificity. Structural studies of full-length ACE2 catalytic domain suggest that lack of the N-terminal residues of ACE2 (especially R273) may affect substrate specificity, suggesting that angiotensin-II may not be the primary substrate of the short ACE2 isoform. We further show that the short transcript is an interferon-regulated gene and is more strongly induced by IFN-beta and viral infection than long ACE2. Although the function of the short ACE2 isoform remains unclear, our data clearly show that this isoform is expressed in the airways, particularly in ciliated cells of the nasal and bronchial epithelium. Of note, ACE2 is an homologue of ACE which also utilizes different promoters to produce two distinct isoforms, the full length molecule and a lower molecular mass variant which is found only in testis (tACE) 56 where its expression regulates sperm capacitation 57 . Like the motile cilia of the airways, sperm flagella have a characteristic 9+2 axoneme structure suggesting some important function of the shorter ACE2 and ACE isoforms, respectively, in relation to these structures. However, we did not find evidence that short ACE2 was enriched in cilia which contrasts with the long form of ACE2. Instead, we found that short ACE2 was retained within the cell body along with a fraction of long ACE2. Therefore, it is 1 9 interesting that short ACE2 is preferentially upregulated in response to viral infection, independently of long transcript ACE2 expression. We hypothesise that this short isoform of ACE2 plays an important physiological role in the airway and, in addition that it may influence host susceptibility to SARS-CoV-2 infection either by heterodimerising with long ACE2 to prevent its trafficking to the exposed tips of the cilia and/or influencing the spike binding interaction; alternatively it may compete for other membrane proteases required for viral entry. Nasal epithelial cells were isolated by brushing the inferior turbinate with a sterile 3.0 mm cytology brush (Conmed). Cells were processed into RNAlater for subsequent RNAseq analysis or were stored in liquid nitrogen prior to cell culture. Bronchial epithelial cells were harvested by bronchoscopic brushings for primary bronchial epithelial cell culture. Airway samples for the study were collected following approval Raw FASTQ reads were subjected to adapter trimming and quality filtering (reads containing N > 10%, reads where >50% of read has Qscore<= 5) by Novogene Inc. A Mendelian RNA-seq method for identifying and filtering splice junctions developed by Cummings et al. 33 was used to detect aberrant and novel splice events. No changes were made to this code. The individual sample splice junction discovery output files were combined into an overall splice junction discovery file used for splice junction normalisation. From nasal and bronchial brushings, cDNA was synthesised from excess RNA purified for RNAseq using High Capacity cDNA Reverse Transcription kit (Thermo Fisher Scientific) following manufacturer's instructions. From cell lines, RNA was isolated from cell lysates using standard phenol-chloroform extraction, and reverse transcribed to cDNA using a Precision Reverse Transcription HRP-conjugated anti-β-actin antibody (Sigma) was used as a loading control. ImageJ was used for densitometry. After apical wash with HBSS, cells were fixed with 4% PFA, permeabilized with 0. Alpha tubulin mouse monoclonal Alpha tubulin Sigma T9026 Human rhinovirus (HRV16; ATCC VR-283™, Teddington, UK) was amplified using H1 HeLa cells as previously described 76, 77 . Infectivity of stocks and release of infective virions in cell culture supernatants was determined using a HeLa titration assay and 50% tissue culture infective dose assay (TCID 50 /ml). Ultraviolet-irradiated virus controls (UV-RV16) were prepared by exposure of virus stocks to UV light at 1200 mJ/cm 2 on ice for 50 min. Fully differentiated nasal and bronchial epithelial cells (28 or 21 days after ALI) were apically infected with human rhinovirus 16 (RV16) at a multiplicity of Infection (MOI) of 1 for 6h, washed apically 3x using HBSS and incubated for additional 18h at the air-liquid interface (24h in total). Cells were washed 3x with HBSS and lysed using TriZol (Invitrogen) for RNA and protein extraction. We are grateful to healthy volunteers and respiratory patients who donated airway cells and to Synairgen Research Ltd who provided cells and reagents to support these studies. Data and materials availability 3 2 All sequence files to be deposited on Sequence Read Archive. Accession numbers will be provided upon publication. 7a. Bronchial epithelial brushes from healthy controls (n=13) or severe asthmatic (n=11) donors were harvested and RNA extracted for analysis of ACE2 transcript expression by transcript-specific RT-qPCR. Data were analysed using nonparametric Mann-Whitney test. 7b. Bronchial epithelial cells from healthy (n=11) or severe asthmatic (n=7) donors were grown at ALI and infected with rhinovirus (RV16) (MOI of 1) or mock-infected using a UV-irradiated control. After 24h, induction of ACE2 transcripts was assessed by RT-qPCR with transcript-specific primers in duplicate. The fold-induction of each transcript by RV16 was quantified using the ddCt method using UV-RV exposed 3 7 cells as control. Activation of anti-viral response in RV16-infected ALI cultures was demonstrated by detection of IL29/IL28 in basolateral supernatants by ELISA (healthy n=14, severe n=8). Data were analysed using Student's t-test. 4a. Long ACE2 homodimer (left), long and short ACE2 heterodimer (middle) and short ACE2 homodimer (teal, from pdb 6M18) with neck dimerization domains shown in green. SARS-Cov-2 spike protein is shown in orange. Figure 6 . Short ACE2 is upregulated in response to IFNβ and Rhinovirus infection but not SARS-Cov-2 In vitro differentiated PBEC In vitro differentiated PBEC c b Figure 7 . Long ACE2 is expressed at lower levels in bronchial brushings from severe asthmatic donors and RV-induced expression of long ACE2 is reduced in bronchial ALI cultures from severe asthmatic donors a A human homolog of angiotensin-converting enzyme. Cloning and functional expression as a captopril-insensitive carboxypeptidase A novel angiotensin-converting enzyme-related carboxypeptidase (ACE2) converts angiotensin I to angiotensin 1-9 Collectrin, a collecting duct-specific transmembrane glycoprotein, is a novel homolog of ACE2 and is developmentally regulated in embryonic kidneys A transmembrane serine protease is linked to the severe acute respiratory syndrome coronavirus receptor and activates virus entry TMPRSS2 and ADAM17 cleave ACE2 differentially and only proteolysis by TMPRSS2 augments entry driven by the severe acute respiratory syndrome coronavirus spike protein Tissue-specific amino acid transporter partners ACE2 and collectrin differentially interact with hartnup mutations A protein complex in the brush-border membrane explains a Hartnup disorder allele Loss of angiotensinconverting enzyme 2 leads to impaired glucose homeostasis in mice Angiotensin Iconverting enzyme type 2 (ACE2) gene therapy improves glycemic control in diabetic mice Angiotensin-converting enzyme 2 protects from severe acute lung failure Recombinant angiotensin-converting enzyme 2 improves pulmonary blood flow and oxygenation in lipopolysaccharide-induced lung injury in piglets Evidence for angiotensin-converting enzyme 2 as a therapeutic target for the prevention of pulmonary hypertension Prevention of pulmonary hypertension by Angiotensin-converting enzyme 2 gene transfer Angiotensin-converting enzyme 2 is a functional receptor for the SARS coronavirus Efficient Activation of the Severe Acute Respiratory Syndrome Coronavirus Spike Protein by the Transmembrane Protease TMPRSS2 A crucial role of angiotensin converting enzyme 2 (ACE2) in SARS coronavirus-induced lung injury Structural basis for the recognition of SARS-CoV-2 by full-length human ACE2 Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor Human coronavirus HCoV-229E enters susceptible cells via the endocytic pathway Clathrin-dependent entry of severe acute respiratory syndrome coronavirus into target cells expressing ACE2 with the cytoplasmic tail deleted SARS coronavirus entry into host cells through a novel clathrin-and caveolae-independent endocytic pathway The transcription factor HNF1α induces expression of angiotensin-converting enzyme 2 (ACE2) in pancreatic islets from evolutionarily conserved promoter motifs Identifying the regulatory element for human angiotensin-converting enzyme 2 (ACE2) expression in human cardiofibroblasts Administration of 17β-estradiol to ovariectomized obese female mice reverses obesity-hypertension through an ACE2-dependent mechanism SARS-CoV-2 Receptor ACE2 Is an Interferon-Stimulated Gene in Human Airway Epithelial Cells and Is Detected in Specific Cell Subsets across Tissues The Genotype-Tissue Expression (GTEx) project SARS-CoV-2 entry factors are highly expressed in nasal epithelial cells together with innate immune genes SARS-CoV-2 Reverse Genetics Reveals a Variable Infection Gradient in the Respiratory Tract Severe acute respiratory syndrome coronavirus infection of human ciliated airway epithelia: role of ciliated cells in viral spread in the conducting airways of the lungs STAR: ultrafast universal RNA-seq aligner Integrative genomics viewer Improving genetic diagnosis in Mendelian disease with transcriptome sequencing ISGF3, the transcriptional activator induced by interferon alpha, consists of multiple interacting polypeptide chains The activator protein 1 binding motifs within the human cytomegalovirus major immediate-early enhancer are functionally redundant and act in a cooperative manner with the NF-{kappa}B sites during acute infection Specification of DNA binding activity of NF-kappaB proteins Generation of a human airway epithelium derived basal cell line with multipotent differentiation capacity ACE2 receptor expression and severe acute respiratory syndrome coronavirus infection depend on differentiation of human airway epithelia PepQuery enables fast, accurate, and convenient proteomic validation of novel genomic alterations Robust ACE2 protein expression localizes to the motile cilia of the respiratory tract epithelia and is not increased by ACE inhibitors or angiotensin receptor blockers Interaction of SARS and MERS Coronaviruses with the Antiviral Interferon Response SARS-CoV-2 nsp13, nsp14, nsp15 and orf6 function as potent interferon antagonists Asthma prevalence in patients with SARS-CoV-2 infection detected by RT-PCR not requiring hospitalization SARS-CoV-2 pneumonia in hospitalized asthmatic patients did not induce severe exacerbation Prevalence and characterization of asthma in hospitalized and non-hospitalized patients with COVID-19 Expression of SARS-CoV-2 Receptor ACE2 and Coincident Host Response Signature Varies by Asthma Inflammatory Phenotype Relationship of epidermal growth factor receptors to goblet cell production in human bronchi Type 2 inflammation modulates ACE2 and TMPRSS2 in airway epithelial cells Dexamethasone in Hospitalized Patients with Covid-19 -Preliminary Report Rhinovirus-16 induced release of IP-10 and IL-8 is augmented by Th2 cytokines in a pediatric bronchial epithelial cell model The innate antiviral response upregulates IL-13 receptor α 2 in bronchial fibroblasts Barrier responses of human bronchial epithelial cells to grass pollen exposure Transforming growth factor-beta promotes rhinovirus replication in bronchial epithelial cells by suppressing the innate immune response Role of deficient type III interferon-lambda production in asthma exacerbations Asthmatic bronchial epithelial cells have a deficient innate immune response to infection with rhinovirus Transcription of testicular angiotensin-converting enzyme (ACE) is initiated within the 12th intron of the somatic ACE gene Testis-specific isoform of angiotensinconverting enzyme (tACE) is involved in the regulation of bovine sperm capacitation Perspectives on the development of neutralizing antibodies against SARS-CoV-2 Decoy ACE2-expressing extracellular vesicles that competitively bind SARS-CoV-2 as a possible COVID-19 therapy Airways Expression of SARS-CoV-2 Receptor, ACE2, and TMPRSS2 Is Lower in Children Than Adults and Increases with Smoking and COPD ACE2 Expression is Increased in the Lungs of Patients with Comorbidities Associated with Severe COVID-19 ACE2 levels are altered in comorbidities linked to severe outcome in COVID-19. medRxiv ACE2 and TMPRSS2 are expressed on the human ocular surface, suggesting susceptibility to SARS-CoV-2 infection Co-expression of SARS-CoV-2 entry genes in the superficial adult human conjunctival, limbal and corneal epithelium suggests an additional route of entry via the ocular surface Expression of the COVID-19 receptor ACE2 in the human conjunctiva Expression of SARS-CoV-2 receptor ACE2 and TMPRSS2 in human primary conjunctival and pterygium cell lines and in mouse cornea Incomplete annotation has a disproportionate impact on our understanding of Mendelian and complex neurogenetic disorders Culture of primary ciliary dyskinesia epithelial cells at air-liquid interface can alter ciliary phenotype but remains a robust and informative diagnostic aid Defective epithelial barrier function in asthma Identification of an Immortalized Human Airway Epithelial Cell Line with Dyskinetic Cilia RSeQC: quality control of RNA-seq experiments Accurate assembly of transcripts through phase-preserving graph decomposition Salmon provides fast and bias-aware quantification of transcript expression A Proteomic Analysis of Human Cilia Isolation of cilia from porcine tracheal epithelium and extraction of dynein arms Viral stimuli trigger exaggerated thymic stromal lymphopoietin expression by chronic obstructive pulmonary disease epithelium: role of endosomal TLR3 and cytosolic RIG-I-like helicases Peroxisome proliferator-activated receptor gamma negatively regulates IFN-beta production in Toll-like receptor (TLR) 3-and TLR4-stimulated macrophages by preventing interferon regulatory factor 3 binding to the IFN-beta promoter UCSF Chimera--a visualization system for exploratory research and analysis Improving physical realism, stereochemistry, and side-chain accuracy in homology modeling: Four approaches that performed well in CASP8 New ways to boost molecular dynamics simulations VMD: visual molecular dynamics