Pulmonary, cardiac and renal distribution of ACE2, furin, TMPRSS2 and ADAM17 in rats with heart failure: Potential implication for COVID‐19 disease

Abstract Congestive heart failure (CHF) is often associated with kidney and pulmonary dysfunction. Activation of the renin‐angiotensin‐aldosterone system (RAAS) contributes to avid sodium retention, cardiac hypertrophy and oedema formation, including lung congestion. While the status of the classic components of RAAS such as renin, angiotensin converting enzyme (ACE), angiotensin II (Ang II) and angiotensin II receptor AT‐1 is well studied in CHF, the expression of angiotensin converting enzyme‐2 (ACE2), a key enzyme of angiotensin 1‐7 (Ang 1‐7) generation in the pulmonary, cardiac and renal systems has not been studied thoroughly in this clinical setting. This issue is of a special interest as Ang 1‐7 counterbalance the vasoconstrictory, pro‐inflammatory and pro‐proliferative actions of Ang II. Furthermore, CHF predisposes to COVID‐19 disease severity, while ACE2 also serves as the binding domain of SARS‐CoV‐2 in human host‐cells, and acts in concert with furin, an important enzyme in the synthesis of BNP in CHF, in permeating viral functionality along TMPRSST2. ADAM17 governs ACE2 shedding from cell membranes. Therefore, the present study was designed to investigate the expression of ACE2, furin, TMPRSS2 and ADAM17 in the lung, heart and kidneys of rats with CHF to understand the exaggerated susceptibility of clinical CHF to COVID‐19 disease. Heart failure was induced in male Sprague Dawley rats by the creation of a surgical aorto‐caval fistula. Sham‐operated rats served as controls. One week after surgery, the animals were subdivided into compensated and decompensated CHF according to urinary sodium excretion. Both groups and their controls were sacrificed, and their hearts, lungs and kidneys were harvested for assessment of tissue remodelling and ACE2, furin, TMPRSS2 and ADAM17 immunoreactivity, expression and immunohistochemical staining. ACE2 immunoreactivity and mRNA levels increased in pulmonary, cardiac and renal tissues of compensated, but not in decompensated CHF. Furin immunoreactivity was increased in both compensated and decompensated CHF in the pulmonary, cardiac tissues and renal cortex but not in the medulla. Interestingly, both the expression and abundance of pulmonary, cardiac and renal TMPRSS2 decreased in CHF in correlation with the severity of the disease. Pulmonary, cardiac and renal ADAM17 mRNA levels were also downregulated in decompensated CHF. Circulating furin levels increased in proportion to CHF severity, whereas plasma ACE2 remained unchanged. In summary, ACE2 and furin are overexpressed in the pulmonary, cardiac and renal tissues of compensated and to a lesser extent of decompensated CHF as compared with their sham controls. The increased expression of the ACE2 in heart failure may serve as a compensatory mechanism, counterbalancing the over‐activity of the deleterious isoform, ACE. Downregulated ADAM17 might enhance membranal ACE2 in COVID‐19 disease, whereas the suppression of TMPRSS2 in CHF argues against its involvement in the exaggerated susceptibility of CHF patients to SARS‐CoV2.

The clinical presentation of patients with SARS-CoV-2 infection has stirred the interest in Ang II-Ang 1-7 balance, since the binding of SARS-CoV-2 spike protein to ACE2 was found to downregulate this organ-protective enzyme and decreasing Ang 1-7 production. 4,14,23 It is tempting to assume that this undesired Ang 1-7 / Ang II imbalance may underline many of the deleterious effects SARS-CoV-2 on the pulmonary, cardiac, renal and vascular systems, especially in patients with chronic underlying diseases. 4,23 Furthermore, blockade of the RAAS system by ACE inhibitors (ACEi) and angiotensin receptor blockers (ARBs) upregulates the expression of ACE2 in patients predisposed to severe SARS-CoV-2 infection, such as individuals with heart failure, already displaying abundant ACE2 in their COVID-19-target organs, thus potentially sensitizing patients on these medications to the virus. 21,24,25 The mechanisms underlying the high susceptibility of CHF patients to COVID-19 remain largely unknown. Therefore, alterations in ACE2 expression in patients with cardiovascular diseases should be taken into account when assessing their risk, morbidity and mortality related to the viral infection.
Unfortunately, the status of ACE2 has mostly been determined in the circulation, limited to ACE2 peptides detached from cell membranes. 4,26,27 There is limited post-mortem data regarding its expression on cell membranes in patients with COVID-19 infection, other than in cardiac tissues of patients with CHF, 8 and its expression in pulmonary and renal tissues is largely unknown. Moreover, the status of other principal participants in viral attachment, such as furin, has not been studied thoroughly in target organs in patients prone to severe COVID-19 infection, including those with CHF. Furin activates more than 150 substrates of mammalian, viral and bacterial origin, cleaving these substrates at basic residues with the typical recognition motif R-Xn-K/R-R, hence given the name PACE (paired basic amino acid cleaving enzyme). In this context, genomic characterization of SARS-CoV-2 revealed that furin cleaves the S spike of the viral envelope glycoproteins, enhancing its fusion to cellular membrane. [28][29][30] Noteworthy, furin is ubiquitously expressed in a broad range of cells within COVID-19-target organs, such as the lung, heart, gut and nasal mucosa, 31,32 where the mature and active form is present in the Golgi network and can be further transported to the cell membrane and back through the endosomal pathway. 31 Additionally, furin can be cleaved by an unknown mechanism and shed into the extracellular space. 33 In sum, so far, information concerning the expression and distribution of ACE2 and furin in target organs of SARS-CoV-2 in individuals with CHF is, unfortunately, limited. The impact of CHF on the expression of TMPRSS2 and ADAM17, Pulmonary, cardiac and renal ADAM17 mRNA levels were also downregulated in decompensated CHF. Circulating furin levels increased in proportion to CHF severity, whereas plasma ACE2 remained unchanged. In summary, ACE2 and furin are overexpressed in the pulmonary, cardiac and renal tissues of compensated and to a lesser extent of decompensated CHF as compared with their sham controls. The increased expression of the ACE2 in heart failure may serve as a compensatory mechanism, counterbalancing the over-activity of the deleterious isoform, ACE. Downregulated ADAM17 might enhance membranal ACE2 in COVID-19 disease, whereas the suppression of TMPRSS2 in CHF argues against its involvement in the exaggerated susceptibility of CHF patients to SARS-CoV2.
Elucidating these issues, may provide new insights into the physiology of CHF and regarding the pathologic mechanisms and susceptibility of patients with CHF to COVID-19 infection, and might eventually pave the way for a better assessment and treatment of such patients in general and while infected with SARS-CoV-2.

| Experimental groups
Studies were conducted on male Sprague Dawley rats weighing 300-350 g. The animals were kept in a temperature-controlled room and fed with standard rat chow containing 0.5% NaCl and tap water ad libitum. All experiments were approved by the Animal Care and Use Committee of the Technion-Israel Institute of Technology.

| Experimental model
Aorto-caval fistula (ACF) was induced in these rats by surgically creation of an incision (approximately 1.2-1.4 mm in length) between the abdominal aorta and inferior vena cava, distal to the origin of the renal arteries as described previously. 37 The opening in the vena cava is then closed by a continuous suture. A matched group of shamoperated rats served as controls. After surgery, rats were allowed to recover and placed in metabolic cages for one week. Seven days after the operation, rats with ACF were divided into two subgroups according to their daily absolute rate of sodium excretion (UNaV): rats with decompensated CHF (UNaV < 200 µEq/24 hours) and rats with compensated CHF (UNaV > 1200 µEq/24 hours). 37 The three subgroups of animals: sham-operated rats (n = 9), compensated (n = 9) and decompensated (n = 11) animals were anesthetized with Nembutal (60 mg/kg, i.p.), and their hearts, lungs and kidneys were washed via right ventricle perfusion with 120 mL phosphate buffered saline (0.01 mol/L PBS, pH 7.4) containing heparin (5 U/mL). The heart was removed and dissected, where the left atrium, left ventricle, right atrium and right ventricle were separated.
The kidney was longitudinally cut into two halves, where one half was homogenized for WB analyses either as a whole kidney or separated cortex and medulla, the second half for RT-PCR. The right lung was utilized for WB and the left one for RT-PCR analysis.

| RT-qPCR
Total RNA was isolated from snap-frozen tissue samples using TRIzol ® Reagent (Life Technologies), according to the manufacturer's instructions, and quantified by spectrophotometry using NanoDrop 2000. After oligo (dT)-primed reverse transcription of 1000 ng total RNA, the resulting single-stranded cDNA was used for PCR. PCR conditions were as follows: an initial denaturation step at 95°C for 3 minutes, 30 cycles of denaturation at 95°C for 30 seconds, and hybridization at 60°C for 30 seconds followed by elongation at 72°C for 1 minute. Finally, the PCR reaction was terminated by incubation at 72°C for 5 minutes. GAPDH was used as an internal standard. The following primers were used:

| Western Blot analysis
Lungs, heart chambers and kidneys were homogenized on ice and centrifuged at 4°C for 5 minutes at 1000 g. The homogenized tissue was then lysed in RIPA buffer (150 mmol/L NaCl, 1% NP40, 50 mmol/L Tris pH 8.0, 0.5% sodium deoxycholate and 0.1% SDS) supplemented with a cocktail of protease inhibitors (Roche) in rotation at 4°C for 20 minutes, and then centrifuged at 4°C for 10 min-

| Tissue fixation and immunofluorescence
Additional groups of rats with compensated and decompensated CHF as well as sham controls (n = 3-6) were anesthetized and their

| Plasma levels of ACE2 and furin
Plasma levels of ACE2 and furin of the various experimental groups were determined by commercial ELISA kits from Aviva Systems Biology, Corp.

| Serum creatinine and blood urea nitrogen
Serum creatinine (SCr) and blood urea nitrogen (BUN) in serum samples of the various experimental groups were determined using commercial kits (Siemens) with an auto-analyser dedicated instrument (Dimension RXL, Siemens).

| Statistical analysis
Data are presented as mean ± SEM One-way analysis of variance (ANOVA) for repeated measures, followed by Tukey's test for comparison of corresponding values between the groups. Comparisons between two groups were done using Student's t test. A value of P < .05 was considered statistically significant.
Rats with compensated and decompensated CHF displayed a significant increase in heart weight in comparison with their sham controls (1.465 ± 0.05 g; P < .001, 1.405 ± 0.044 g; P < .01 vs 1.175 ± 0.033 g, respectively) ( Figure 1A). In addition, rats with F I G U R E 1 Impact of aorto-caval placement on heart, lung and kidney weights as compared with sham-operated controls. Cardiac, lung, and kidney weights expressed as absolute values (A-C, respectively) of rats with compensated and decompensated CHF and their sham controls (n = 9-11). Values are means ± SEM decompensated CHF exhibited pulmonary congestion as was evident by a significant increase in lung weight as compared with sham controls (3.80 ± 0.43 g; P < .05 vs 2.08 ± 0.11 g, respectively) and even compensated CHF subgroup (2.28 ± 0.097 g, P < .05) ( Figure 1B). In contrast with cardiac hypertrophy and pulmonary congestion, kidney weight was significantly declined in correlation with CHF severity, reaching 1.016 ± 0.03 g (P < .01) in compensated CHF and 0.871 ± 0.02 g (P < .0001) in decompensated CHF animals, as compared to sham controls (1.154 ± 0.03 g) ( Figure 1C).
Expression and abundance of ACE2, Furin, TMPRSS2 and ADAM17 were determined in pulmonary, cardiac and renal tissues of compensated and decompensated CHF rats and sham control animals by using western blot, RT-PCR and immunohistochemical techniques.

| Angiotensin converting enzyme-2
Higher immunoreactive levels of ACE2 were detected in the lungs of compensated (1.82 ± 0.245, P < .05), but not decompensated rats (0.71 ± 0.267), as compared to sham controls (1.0 ± 0.01) ( Figure 2B). In line with these findings, pulmonary ACE2 expression was ~3-fold higher in compensated rats (2.96 ± 0.47, P < .05) as compared with control animals (1.03 ± 0.04) ( Figure 2C). In contrast, F I G U R E 2 Immunoreactive levels and expression of ACE2 in the lungs, LV, RV and kidneys of compensated, decompensated CHF and sham controls as were determined with western blots analysis and RT-qPCR. A, Representative western-blot analysis of tissue lysates with antibody for ACE2. Western-blot analysis quantification of ACE2 in the lungs, LV, RV and kidneys (B, D, F and H, respectively) of sham, compensated CHF and decompensated CHF (n = 3-10), where GAPDH was used as loading control. Quantification of RT-qPCR analysis for ACE2 mRNA normalized to GAPDH are depicted in C, E, G and I.
Interestingly, the expression of ACE2 in the LV and RV was upregulated in compensated CHF but downregulated in decompensated animals ( Figure 2E,G). No differences were found in atrial immunoreactive levels of ACE2 between sham controls, compensated and decompensated CHF, although ACE2 mRNA was significantly higher in compensated in both atria as compared with decompensated subgroup and sham controls (Data not shown).
Immunofluorescence staining revealed intense immunostaining of ACE2 in the pulmonary, cardiac and renal tissues ( Figure 6A).
Specifically, ACE2 immunofluorescence was detected mainly in the endothelium of pulmonary small blood vessels and alveolar epithelial cells of all studied subgroups. In agreement with WB analysis, immunofluorescence of pulmonary ACE2 was explicitly enhanced in lung tissue of compensated CHF group, but not in decompensated CHF animals ( Figure 6A). Similar pattern was observed in the cardiac tissue, where ACE2 immunofluorescence was enhanced in compensated CHF and to a lesser extent in the decompensated subgroup.
Similarly, renal immunofluorescence of ACE2 was more abundant in compensated subgroup but not decompensated animals ( Figure 6A).
It should be emphasized that ACE2 was mainly localized to the peritubular capillaries and almost absent in tubular epithelial cells. In the next step, we analysed regional ACE2 immunofluorescence in the kidney. ACE2 immunofluorescence was localized mainly to the renal cortex (vasculature) and to lesser extent in the medulla (vasa recta) ( Figure S2A). Immunofluorescence staining revealed intense immunostaining of furin in the pulmonary, cardiac and renal tissues ( Figure 6B).

| Furin
Specifically, furin immunofluorescence was detected in the alveolar epithelial cells of all studied groups. Noteworthy, immunostaining was more intense in rats with compensated and decompensated CHF ( Figure 6B). In agreement with the WB and RT-PCR analysis, immunofluorescence of furin was enhanced in cardiac tissue of both compensated and decompensated CHF subgroups ( Figure 6B). Renal immunofluorescence of furin was abundant in the tubular epithelial cells and vascular endothelium, where it is elevated in the compensated group and to a lesser extent in the decompensated group, compared to sham rats. ( Figure 6B). Analysis of regional furin immunofluorescence showed abundance of this enzyme in both renal cortex and medulla, mainly in the tubule. Induction of CHF slightly enhanced furin immunofluorescence in the cortical and medullary tissues ( Figure S2B).

| TMPRSS2
The immunoreactive levels and expression of TMPRSS2 in the pulmonary, cardiac and renal tissues are presented in Figure 4. The immunoreactive levels of TMPRSS2 declined in the pulmonary tissue of CHF animals in correlation with the severity of the disease, where it reached 48.3 ± 6% (P < .001) in decompensated subgroup ( Figure 4B). In parallel to these results, pulmonary TMPRSS2 mRNA levels were significantly decreased in both compensated and decompensated CHF animals (0.38 ± 0.08; P < .01, 0.12 ± 0.03; P < .0001 vs 1.0 ± 0.07, respectively), where the decline was more profound in the latter subgroup ( Figure 4C). While LV immunoreactive levels of TMPRSS2 was not significantly changed following CHF induction ( Figure 4D), RV TMPRSS2 abundance was decreased in decompensated CHF animals by ~ 53±9% (P < .01) ( Figure 4F). In RV, and in similarity with its behaviour in the lung and LV tissues, TMPRSS2 expression was downregulated in CHF animals, especially in the de- Immunofluorescence staining revealed immunostaining of TMPRSS2 in the pulmonary, cardiac and renal tissues ( Figure 6C).
Specifically, TMPRSS2 immunofluorescence was detected in the

| ADAM17
The immunoreactive levels and expression of ADAM17 in the pulmonary, cardiac and renal tissues are presented in Figure 5. While F I G U R E 6 Histological changes in lung, heart and kidney tissues from rats with compensated and decompensated CHF as compared to their sham controls. Slices from lung, heart and kidneys of the various experimental groups were stained with primary antibodies of ACE2 (A), Furin (B) and TMPRSS2 (C). Cy™3 Donkey Anti-Rabbit IgG, Cy™3 Donkey Anti-Goat IgG and Cy™3 Donkey Anti-Mouse IgG were used as secondary antibodies together with DAPI Fluoromount-G ® for nuclear staining. Images were captured using a Widefield Zeiss Upright microscope and analysed with Zen software. Representative images of the cardiac, pulmonary and renal tissues were obtained at 40×, 20× and 20× magnification, respectively compensated animals displayed no differences in pulmonary immunoreactive levels of ADAM17 as compared with sham controls, a non-significant upregulation was observed in decompensated rats ( Figure 5B). In contrast, pulmonary ADAM17 expression exhibited distinct pattern, where this enzyme was increased by ~2 fold (1.97 ± 0.39; P = .09) in compensated rats and declined by ~40% (0.6 ± 0.07; P < .01) in decompensated CHF subgroup ( Figure 5C).
While no change in ADAM17 immunoreactivity was detected in LV tissue of the different CHF subgroups, lower levels were observed in the RV chamber of decompensated CHF animals (0.75 ± 0.07; P < .05) ( Figure 5D,F). The expression of ADAM17 in LV was significantly higher in compensated CHF rats as compared with sham (1.47 ± 0.08 vs

| Plasma levels of ACE2 and furin
Plasma levels of ACE2 and furin of the various experimental groups are shown in Figure 7. No change in ACE2 levels was detected in CHF compensated (0.39 ± 0.05 ng/mL) and decompensated (0.45 ± 0.11 ng/mL) subgroups as compared with their sham controls (0.39 ± 0.09 ng/mL). In contrast, circulatory levels of furin increased in CHF animals in correlation with the disease severity: 0.03 ± 0.01 ng/mL (Compensated) and 0.03 ± 0.02 (Decompensated) vs 0.01 ± 0.01 (Sham), but did not reach statistical significance.  38 Similarly, overexpression of ACE2 in cardiac tissues of spontaneously hypertensive rats decreased cardiac remodelling as was evident by reduced left ventricular wall thickness and perivascular fibrosis, 39 probably via reduction of collagen synthesis. 40 Plausible ACE2/Ang 1-7/Mas-R -mediated protective mechanisms might involve myocardial conservation, afterload-reduction, natriuresis and diuresis. 8,16,20,41,42 In these perspectives, our current findings, showing that enhanced myocardial ACE2 expression in compensated CHF is blunted and even suppressed in rats with decompensated CHF, suggest a role for ACE2 in maintaining viability in the setup of cardiac dysfunction. Support for this concept emerges from previous experimental reports demonstrating cardiac contractility defects in rats with reduced X chromosomal-derived ACE2 expression and heart failure with pulmonary congestion in ACE2 knockout mice (ACE2 KO). 43,44 Interestingly, the hearts of animals depleted of ACE2 exhibited similar changes that occur after coronary artery disease or  Heart failure is also associated with increased expression of furin. 28 Furin presents mainly intracellularly and to a lesser extent in the circulation, 33 where it converts ventricular proBNP to active BNP, an important physiological process in heart failure subjects.

| D ISCUSS I ON
Patients with heart failure are specifically characterized by upregulation of cardiac furin, providing an additional potential explanation for their vulnerability COVID-19 infection. 63  Our study is descriptive and does not address mechanisms involved in the changes in the transcription and expression of ACE2 and of the molecules linked to ACE2 regulation caused by CHF.
Further studies are needed to assess their status in humans with CHF. Furthermore, there is no data regarding their expression and role in humans during COVID-19 disease, and our relevant discussion is to large extent speculative. Yet these assumptions and suggestions should be further evaluated, since they may bear clinical F I G U R E 8 SARS-CoV-2 binding, activation, invasion and replication in target cells. The initial step after the invasion of SARS-CoV-2 is binding to membranal ACE2 widely expressed in vital organs including lung, heart and kidney. ACE2 is responsible for the conversion of Ang II to Ang 1-7 which exerts beneficial effects on the cardiac tissue such as vasodilation, anti-fibrosis and anti-inflammation via Mas receptor (MasR). The binding of SARS-CoV-2 to ACE2 is preceded by TMPRSS2/furin-mediated exposure of the viral receptor binding protein (RBP) localized to S-glycoprotein (S1 domain of the viral spike) and revealing the viral effusion site on S2 domain. Furin is expressed in these tissues both intracellularly and in the circulation as a free enzyme, making it a key factor along TMPRSS2 in the uncovering of RBP and eventually in SARS-CoV-2 transmission. In addition, furin enhances the affinity of the virus to ACE2, not only by exposing the viral binding site on S1 domain but also by revealing the effusion site on the S2 domain in the viral spike. Consequently, the virus undergoes endocytosis and massive replication accompanied by profound activation by the abundant intracellular furin and Cathepsin L (Cat-L). The activated intracellular SARS-CoV-2 undergoes exocytosis where it binds again to ACE2 elsewhere, thus creating a vicious feed-forward devastating cycle. According to the current study, compensated, but not decompensated congestive heart failure (CHF) is characterized by enhanced expression of myocardial ACE2 and downregulation of TMPRSS2. ADAM17 is responsible for shedding of ACE2, a process stimulated by AT1 receptor (AT1-R) and may explain why renin angiotensin aldosterone system (RAAS) inhibitors augment ACE2 expression.
relevance with plausible interventional implications in this devastating disease. Additional limitation of the current study is the short follow up period (1 week), which represents acute CHF, rather than a chronic condition that usually develops in humans in many years.
In summary, ACE2 and furin are overexpressed in pulmonary, cardiac and renal tissues of compensated and to a lesser extent of decompensated CHF as compared with their sham controls ( Figure 8). While upregulation of myocardial ACE2 in heart failure has been reported in animals and humans, 8,16,20,25 its enhanced expression in pulmonary and renal tissues, reported here, is a novel finding. The increased expression of the beneficial ACE2 and of furin in compensated heart failure may serve as a compensatory response to the over-activity of the deleterious isoform, namely, ACE. Failure to increase ACE2 might play a role in the development of decompensated CHF, through augmented afterload, ROS-mediated myocardial injury, cardiac remodelling and fibrosis and fluid retention (Figure 8).
In the perspective of the role ACE2 plays in SARS-CoV-2 invasion and the likely impact of its depletion in the clinical manifestations of COVID-19 disease, the observed additional changes in TMPRSS2 and ADAM17 tissue expression in experimental CHF may have pathophysiologic significance.

ACK N OWLED G EM ENTS
We wish to thank Mrs Aviva Kabala for her technical support.

CO N FLI C T O F I NTE R E S T
Authors have no conflicting interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.