Kidney and Lung ACE2 Expression After an ACE Inhibitor Or an Ang II Receptor Blocker: Implications for COVID-19

RESEARCH LETTERS www.jasn.org

Kidney and Lung ACE2 Expression after an ACE Inhibitor or an Ang II Receptor Blocker: Implications for COVID-19

Jan Wysocki,1 Enrique Lores,1 Minghao Ye,1 Maria Jose Soler,1,2 and Daniel Batlle1

1Division of Nephrology and Hypertension, Department of Medicine, The Feinberg School of Medicine, Northwestern University, Chicago, Illinois 2Department of Nephrology, Vall d’Hebron University Hospital, Barcelona, Spain

JASN 31: 1941–1943, 2020. doi: https://doi.org/10.1681/ASN.2020050667

After the recognition early in 2020 that the effect of captopril and telmisartan, associated with a significant increase in angiotensin-converting enzyme 2 (ACE2) administered for 2 weeks, to pharmaco- cytosolic ACE2 protein (Figure 1D) sug- is the main receptor for severe acute logically inhibit ACE activity and the gesting internalization of the protein. By respiratory syndrome coronavirus 2 AT1 receptor, respectively. confocal microscopy (Figure 1C), ACE2 (SARS-CoV-2),1 concerns were rap- We also studied the effect of kidney staining was mostly apical but could also idly raised about the use of renin- ACE deficiency on kidney ACE2 expres- be seen in the cytoplasm of tubular cells angiotensin system (RAS) blockers in sion in two genetic models of kidney from captopril-treated mice which was patients with coronavirus disease 2019 ACE ablation to investigate the effect of less evident in control mice. ACE staining, (COVID-19).2–4 The concerns were kidney ACE deficiency on kidney ACE2 by contrast, remained restricted to the largely based on previous studies showing expression. Our findings in the two apical membrane (Figure 1C). that angiotensin II type 1 (AT1) receptor models of kidney ACE genetic ablation, ACE2proteinintotalkidneylysatesfrom blockers (ARBs) and ACE inhibitors can global in the ACE.4 mice (Supplemental telmisartan-treated mice was also not signif- upregulate ACE2 in certain experimental Figure1)andrestrictedtothekidney icantly different from vehicle-treated mice conditions, although results have not al- in the ACE8/8 mice (Supplemental (114%616%). However, in isolated kidney ways been consistent.3–6 Nevertheless, Figure 2), revealed that lack of ACE pro- membranes, there was a significant decrease the question is important because certain tein was associated with a significant in ACE2 protein (76%69% of the vehicle- groups of patients at risk of severe reduction of kidney ACE2 protein that treated mice; P50.03) (Figure 1B). COVID-19—including those with hyper- was not accompanied by a reduction in In lung tissue, ACE2 protein is tension, heart disease, diabetes mellitus, kidney ACE2 mRNA. low.9,10 Consistent with this, attempts and the elderly—are often treated with Kidney cortex of mice treated with to perform Western blots with either to- RAS blockers.2 captopril or telmisartan for 2 weeks tal lysates or isolated membranes did not In polarized epithelia, like the lungs, and respective vehicle-treated controls yield a signal. Therefore, our results are kidneys and intestine, ACE2 in its full- were used to evaluate ACE2 mRNA, pro- limited to ACE2 activity that, although length form is anchored to the apical tein, and activity (see methods in low, was consistently detected and can plasma membrane.7,8 Although kidneys Supplemental Appendix). No significant be considered a surrogate for relative have abundant ACE2 in the proximal changes in mRNA levels were found be- protein abundance. Neither captopril tubule, lungs have a low level of ACE2 tween captopril-treated and control nor telmisartan had a significant effect expression.9,10 However, type 2 pneu- mice (99%621% of control). ACE2 ac- mocytes express ACE2 and, moreover, tivity and protein were lower in lysates Received May 16, 2020. Accepted June 27, 2020. possess Transmembrane protease, serine from captopril-treated mice (81%68% 2 (TMPRSS2), a protease critical for and 71%65% of control mice, respec- Published online ahead of print. Publication date priming of the ACE2–SARS-CoV-2 tively) but the difference did not reach available at www.jasn.org. complex, a step needed for cell entry.1,10 statistical significance. In isolated mem- Correspondence: Dr. Daniel Batlle, Division of Ne- The effect of ACE inhibitors and ARBs branes, however, the decrease in kidney phrology and Hypertension, Department of Medicine, Northwestern University, The Feinberg School of on ACE2 protein expression in the lungs, ACE2 protein was profound and statisti- Medicine, 320 E Superior, Tarry 14-727, Chicago, IL however, has not been previously report- cally significant (37%64% of the vehicle- 60611. Email: [email protected] P5 ed. To gain insight into this question, we treated mice; 0.0004) (Figure 1A). This Copyright © 2020 by the American Society of used kidney and lung lysates to examine decrease in membrane-bound ACE2 was Nephrology

JASN 31: 1941–1943, 2020 ISSN : 1046-6673/3109-1941 1941 RESEARCH LETTERS www.jasn.org

A B 250 ** 250 *

200 200

150 150

100 100

50 50 % mean of the Control % mean of the Control

0 0 Control Captopril Control Telm

C ACE2 ACE Merge D 250 * 200 150 Control 100 Captopril 50

% mean of the Control 0 Cytosol ACE2 protein

E F 1.5 1.5 ns ns

1.0 1.0

0.5 0.5 Lung ACE2 act (RFU/ug/hr) 0.0 Lung ACE2 act (RFU/ug/hr) 0.0 Control (8) Capto (8) Control (6) Telm (6)

Figure 1. ACE2 expression in kidney and lung membranes from captopril and telmisartan treated mice. (A and B) ACE2 protein in kidney membranes from (A) captopril- and (B) telmisartan-treated mice. *P,0.05; **P,0.01. (C) Confocal microscopy of kidney proximal tubule showing ACE2 (red), ACE (green), and merged image (yellow). ACE2 and ACE are mainly in the apical site in both control (upper panel) and captopril treated mice (lower panel). ACE2 staining is also seen in the cytoplasm (double arrow) of a captopril-treated mouse which is not as evident in control mice. The inset (D) (bar graphs) shows a significant increase in cytosolic ACE2 protein in cytosolic membranes from captopril treated as compared with vehicle-treated mice, *P,0.05. Data are expressed as percentage of control. (E and F) ACE2 enzymatic activity in isolated membrane preparations from lungs from (E) captopril- and (F) telmisartan-treated mice. In lungs, no significant differences were detected as compared to their respective controls. Act, activity; Telm, telmisartan. on ACE2 activity in lung total lysates protein. Within lung tissue, ACE2 pro- its full-length form is anchored to the (Supplemental Figure 3). tein is detectable only in type 2 pneumo- apical plasma membrane.7,8,10 Here we Likewise, in lung membranes, ACE2 cytes. In this cell type, Transmembrane show that captopril and telmisartan activity was unaffected by either captopril protease, serine 2 (TMPRSS2), a prote- both decrease kidney ACE2 protein in (Figure 1E) or telmisartan (Figure 1F). ase critical for activation and fusion of kidney membranes without signifi- In assessing the relative quantities of the SARS-CoV-2 and ACE2 complex, is cantly affecting protein abundance in ACE2 that can act as the SARS-CoV-2 re- also present.1,10 total kidney lysates (Figure 1). Capto- ceptor, what matters most is the abundance In polarized epithelia like in the pril, in particular, produced a marked of full-length, membrane-bound ACE2 lungs, kidneys, and intestine, ACE2 in decline in ACE2 protein in kidney

1942 JASN JASN 31: 1941–1943, 2020 www.jasn.org RESEARCH LETTERS membranes while increasing cytosolic FUNDING SARS-CoV-2 cell entry depends on ACE2 ACE2 (Figure 1). and TMPRSS2 and is blocked by a clinically proven protease inhibitor. Cell 181: 271–280.e8, In conclusion, genetic ablation and in- D. Batlle was funded by National Institute of 2020 Diabetes and Digestive Kidney Diseases grant hibition of kidney ACE are both ac- 2. Fang L, Karakiulakis G, Roth M: Are patients RO1DK104785. companied by a reduction of kidney with hypertension and diabetes mellitus at ACE2 expression. This suggests changes increased risk for COVID-19 infection? [pub- intheexpressioninoneenzymemayelicit lished correction appears in Lancet Respir Med 8: e54, 2020]. Lancet Respir Med 8: e21, similarly directional changes in the ACKNOWLEDGMENTS other homolog such that formation 2020 3. Danser AJ, Epstein M, Batlle D: Renin- and degradation of angiotensin II can This work was supported by the Joseph and angiotensin system blockers and the be regulated. The administration of an Bessie Feinberg Foundation (Dr. Daniel COVID-19 pandemic: At present there is no ACE inhibitor, captopril, and an ARB, tel- Batlle). We thank Dr. Hong D. Xiao (Provi- evidence to abandon renin-angiotensin sys- – misartan, decreased ACE2 protein expres- dence Portland Medical Center, Portland, tem blockers. Hypertension 75: 1382 1385, 2020 sion in kidney-isolated membranes. Each OR)andDr.KennethE.Bernstein(Cedars type of RAS blocker had no detectable ef- 4. Vaduganathan M, Vardeny O, Michel T, Sinai, Los Angeles, CA) for generously pro- McMurray JJV, Pfeffer MA, Solomon SD: fect on ACE2 activity in lung-isolated viding kidneys from ACE deficiency models. – – fi Renin angiotensin aldosterone system in- membranes. These ndings altogether Dr. Daniel Batlle reports nonfinancial sup- hibitors in patients with Covid-19. NEngl show that ACE2 is not increased in two port from Angiotensin Therapeutics Inc., JMed382: 1653–1659, 2020 organs that are potential target sites for outside the submitted work. Dr. Maria Jose 5. Soler MJ, Ye M, Wysocki J, William J, SARS-CoV-2 infection. In fact, in kidney Lloveras J, Batlle D: Localization of ACE2 in Soler reports personal fees from AstraZeneca, fi apical membranes, ACE2 protein is de- fi the renal vasculature: Ampli cation by an- non nancial support from Boehringer In- giotensin II type 1 receptor blockade using creased after both captopril and telmisartan fi gelheim, non nancial support from Eli Lilly, telmisartan. Am J Physiol Renal Physiol 296: administration. Therefore, we conclude personal fees and nonfinancial support from F398–F405, 2009 that RAS blockers do not increase ACE2 Esteve, personal fees from FMC, personal fees 6. Ferrario CM, Jessup J, Chappell MC, Averill in either lung or kidney epithelia. If from Janssen, personal fees from Mundi- DB, Brosnihan KB, Tallant EA, et al.: Effect changes in full-length ACE2 are indeed pharma, and personal fees from NovoNordisk, of angiotensin-converting enzyme inhibition fi and angiotensin II receptor blockers on car- suf cient to affect SARS-CoV-2 infectiv- outside the submitted work. ity, the risk cannot be increased by ACE diac angiotensin-converting enzyme 2. Cir- culation 111: 2605–2610, 2005 inhibitors or ARBs. This experimental 7. Ye M, Wysocki J, William J, Soler MJ, Cokic finding in mouse organs support the po- SUPPLEMENTAL MATERIAL I, Batlle D: Glomerular localization and sition of many medical societies and re- expression of angiotensin-converting enzyme cent publications expressing the view 2 and angiotensin-converting enzyme: Impli- This article contains the following sup- that the use of RAS blockers should be cations for albuminuria in diabetes. JAmSoc plemental material online at http://jasn. – continued in patients at risk of contract- Nephrol 17: 3067 3075, 2006 asnjournals.org/lookup/suppl/doi:10.1681/ 8. Hamming I, Timens W, Bulthuis MLC, Lely ing COVID-19. Ongoing clinical trials ASN.2020050667/-/DCSupplemental. AT,NavisGJ,etalH:Tissuedistributionof may or may not show benefitofusing Supplemental Appendix. ACE2 protein, the functional receptor for RAS blockers in patients with COVID- SARS coronavirus. A first step in under- Supplemental Figure 1. ACE2 protein, 19, but what it is clear is that there is no standing SARS pathogenesis. J Pathol 203: enzymatic activity and mRNA in kidneys increased risk for infectivity by using 631–637, 2004 from ACE.4 mice. RAS blockers. 9. Serfozo P, Wysocki J, Gulua G, Schulze A, Supplemental Figure 2. ACE2 protein, Ye M, Liu P, et al.: Ang II (angiotensin activity and mRNA in kidneys from ACE 8/8 II) conversion to angiotensin-(1-7) in the mice and wild type controls. circulation is POP (prolyloligopeptidase)- DISCLOSURES Supplemental Figure 3. ACE2 activity in dependent and ACE2 (angiotensin-converting enzyme 2)-independent. Hypertension 75: total cell lysates from lungs of captopril (A) – D. Batlle is a coinventor of the patent “Active 173 182, 2020 and telmisartan (B) treated mice. Low Molecular Weight Variants of Angiotensin 10. Ziegler CGK, Allon SJ, Nyquist SK, Mbano Converting Enzyme 2,” founder of “Angioten- IM, Miao VN, Tzouanas CN, et al.: HCA Lung sin Therapeutics Inc.”J. Wysocki is a coinventor Biological Network: SARS-CoV-2 receptor of the patent “Active Low Molecular Weight REFERENCES ACE2 is an interferon-stimulated gene in Variants of Angiotensin Converting Enzyme human airway epithelial cells and is detected 2.” All remaining authors have nothing to 1. Hoffmann M, Kleine-Weber H, Schroeder S, in specific cell subsets across tissues. Cell disclose. Krüger N, Herrler T, Erichsen S, et al.: 181: 1016–1035.e19, 2020

JASN 31: 1941–1943, 2020 RAS Blockers and ACE2 Expression 1943 Supplement to Research Letter

‘Kidney and Lung ACE2 expression after an ACE inhibitor or an Ang II receptor blocker: implications for COVID-19’

Jan Wysocki1, Enrique Lores1, Minghao Ye1, Maria Jose Soler 1,2 and Daniel Batlle1

1Division of Nephrology & Hypertension, Department of Medicine The Feinberg School of Medicine, Northwestern University Chicago, IL 60611

2 Hospital Universitari Vall d'Hebron, Department of Nephrology Hospital Vall d'Hebron,

Correspondence: Daniel Batlle, MD

Division of Nephrology & Hypertension, Department of Medicine The Feinberg School of Medicine, Northwestern University 320 E Superior Chicago, IL 60611 Phone: (312) 908-8342 Fax: (312) 503-0622 e-mail: [email protected]

METHODS

Animal Models

All studies were conducted with the review and approval of the Institutional Animal Care and Use

Committee of Northwestern University. Organs from two genetic models of ACE kidney ablation

(ACE.4 and ACE 8/8) were generously provided by Drs Hong D. Xiao and Kenneth Bernstein.

ACE.4 mice have the somatic ACE promoter replaced by the kidney androgen-regulated protein promoter which in these mice is essentially non-functional[1]; in the absence of exogenous androgens the levels of kidney ACE are less than 1% of normal and no ACE is detected in other organs ; this model is overtly hypotensive [1]. The effect of localized kidney ACE deficiency on kidney ACE2 levels was also examined in ACE8/8 mouse which is a model lacking ACE in the kidney or vascular endothelium, but with 100-fold normal cardiac ACE levels[2]. In addition,

ACE8/8 mice have significant levels of ACE activity in the lung and in the blood plasma [2]. In this animal model, kidney represents a major change from wild-type mice as this organ normally expresses a substantial amount of ACE activity in both vascular endothelium and proximal tubular epithelium[2]. The ACE8/8 mice have a near normal blood pressure and do not exhibit any gross abnormalities in kidney function[2]. Both ACE models used in this study have a mixed C57/129 background[1, 2].

To examine the effect of pharmacological ACE inhibition on ACE2, two groups of 12-14 weeks old C57BLKS/J mice were assigned to drink either tap water (n=8) or tap water with an ACE inhibitor, captopril, (n=8) at a dose of 120 mg*kg-1*day-1 for 14 days. Two other experimental groups of 12-14 weeks old C57BLKS/J mice were randomly assigned to drink either tap water

(vehicle, n=6) or tap water with an angiotensin II receptor antagonist, telmisartan (Boehringer

Ingelheim), at a dose of 2 mg*kg-1*day-1 (n=6) for 14 days. Before captopril and telmisartan administration, mice were weighted and the daily fluid intake per mouse was recorded to estimate the concentration of the compound needed to be added to the drinking water. Also during the drug administration, water consumption and body weights were controlled to make appropriate adjustments.

RNA isolation and reverse transcriptase real-time PCR

RNA was isolated from kidney cortex with Trizol reagent (GIBCO Invitrogen). Quantitative real- time PCR was performed using the TaqMan Gold RT-PCR kit and ABI Prism 7700 (Applied

Biosystems) sequence-detection system. Primers and probes for ACE were designed using Primer

Express software (Applied Biosystems). The sequences of forward, reverse primer, and probe for

ACE, ACE2 and Glyceraldehyde-3-phosphate dehydrogenase (G3PDH) were the same as described before[3]. Reverse transcription was carried out for 30 min at 48°C. The ACE and ACE2 mRNA levels of the samples were normalized to their G3PDH contents. Experiments were carried out in triplicate for each data point.

Protein extraction and measurement of enzymatic activity for ACE2

For total cell lysates, tissues (kidney cortex and lungs) were homogenized in a buffer consisting of (in mmol/l) 50 HEPES, pH 7.4, 150 NaCl, 0.5% Triton X-100, 0.025 ZnCl2, and 1.0 phenylmethylsulphonyl fluoride and then clarified by centrifugation at 6,000g for 15 min. For membrane fraction isolation, a previously employed protocol was used[4] but without EDTA addition. After measuring protein concentration, tissue samples were diluted in a buffer (50 mmol/l

4-morpholineethanesulfonic acid, 300 mmol/l NaCl, 10 µmol/l ZnCl2, and 0.01% Triton-X-100, pH 6.5), containing EDTA-free tablets. The plates were read using a fluorescence plate reader

FLX800 (BIOTEK Instruments) at an excitation wavelength of 320 nm and an emission wavelength of 400 nm. All reactions were performed at ambient temperature in microtiter plates with a 100 µl total volume [5, 6].

Western blot analysis

Total cell lysates and membrane preparations from kidney cortices and lungs were isolated as specified above and subjected to Western blot analysis as previously described[6]. For detection of ACE, nitrocellulose membranes were incubated with mouse monoclonal antibody (5C4, kind gift from Dr. Sergei Danilov). ACE2 protein was detected using our affinity purified rabbit anti-

ACE2 antibody[5, 6]. Signals on Western blots were quantified by densitometry (Eagle Eye,

Stratagene) and corrected for GAPDH (total cell lysates) or β-actin (membrane fraction) loading control.

Statistical analyses

The data were expressed as relative to the control (vehicle or wild-type) animal groups, assigning a value of 100% to the control baseline mean. The data were reported as mean ± standard error

(SE). Significance was defined as p< 0.05. For comparisons between two independent means, the t-test was used.

Results

In ACE.4 mice, kidney ACE mRNA was about 7±4% of the mRNA levels in WT mice as previously reported [1]. Membrane kidney lysates from ACE.4 mice kidney cortex were examined for the presence of ACE and ACE2 protein (Figure S1A). The signal for ACE protein in kidneys from WT mice was detectable as a single band at the expected molecular weight of about 170 kD.

In ACE.4 mice, ACE protein was barely detectable (equivalent of 3±1% of the WT) (Figure S1A).

In ACE.4 mice, ACE2 protein abundance in kidney cortex total cell lysates and in isolated membranes was reduced to 70 % and to 42% of the WT, p<0.05, respectively (Figure S1B and

S1C).

Concordant with ACE2 protein expression, ACE2 activity was also reduced significantly in kidneys from ACE.4 mice as compared to WT controls (58.3±12% of the WT) (Figure S1D). By contrast , kidney cortex ACE2 mRNA levels in ACE.4 mice were not significantly different from those of their respective WT littermates (1.24±0.71 vs. 0.69±0.18 gene expression units, respectively) (Figure S1E).

ACE2 in kidneys from ACE8/8 mice, a model of kidney ACE deficiency and cardiac ACE over-expression

Consistent with results from Xiao et al.[2], in kidney cortex from ACE8/8 mice there was no detectable ACE protein expression (Figure S2A), whereas it was clearly present in WT kidneys

(Figure S2A). ACE2 protein expression in isolated kidney membranes from ACE8/8 was significantly reduced (38±9% of WT mice) (Figure S2B). In ACE8/8 mice, ACE2 activity likewise was lower than in the wild-type but this difference was small and not statistically significant

(90±17% of the WT) (Figure S2C). ACE2 mRNA levels in kidney cortex from ACE8/8 mice were not different from those of their respective WT littermates (1.02±0.04 vs. 1.00±0.07, respectively).

References

1. Campbell, D.J., et al., Effect of reduced angiotensin-converting enzyme gene expression and angiotensin-converting enzyme inhibition on angiotensin and bradykinin peptide levels in mice. Hypertension, 2004. 43(4): p. 854-859. 2. Xiao, H.D., et al., Mice with cardiac-restricted angiotensin-converting enzyme (ACE) have atrial enlargement, cardiac arrhythmia, and sudden death. The American journal of pathology, 2004. 165(3): p. 1019-1032. 3. Soler, M., et al., ACE2 inhibition worsens glomerular injury in association with increased ACE expression in streptozotocin-induced diabetic mice. Kidney international, 2007. 72(5): p. 614-623. 4. Ye, M., et al., Increased ACE 2 and decreased ACE protein in renal tubules from diabetic mice: a renoprotective combination? Hypertension, 2004. 43(5): p. 1120-5. 5. Ye, M., et al., Murine recombinant angiotensin-converting enzyme 2: effect on angiotensin II-dependent hypertension and distinctive angiotensin-converting enzyme 2 inhibitor characteristics on rodent and human angiotensin-converting enzyme 2. Hypertension, 2012. 60(3): p. 730-40. 6. Wysocki, J., et al., ACE and ACE2 activity in diabetic mice. Diabetes, 2006. 55(7): p. 2132- 9.

Figure S1. ACE2 protein, enzymatic activity and mRNA in kidneys from ACE.4 mice and wild type controls.

Figure S2. ACE2 protein, activity and m RNA in kidneys from ACE8-8 mice and wild type controls.

Figure S3. ACE2 activity in total cell lysates from lungs of captopril and telmisartan treated mice.

A B ) )

r 1.5 r 1.5

h h ns / ns / g g u u / / U U

F 1.0 F 1.0 R R ( (

t t c c a a

2 2

E 0.5 E 0.5 C C A A

g g n n u u

L 0.0 L 0.0 ) ) ) ) 8 8 6 6 ( ( ( ( l o l o t o lm tr p tr e n a n T o C o C C

Supplemental Figure legends

Figure S1. ACE2 protein, enzymatic activity and mRNA in kidneys from ACE.4 mice and wild type controls. Panel A shows a representative Western blot for ACE, ACE2 and β-actin for three mice from each respective group in kidney membrane preparations. ACE2 protein in kidney total cell lysates (panel B), ACE2 protein in isolated membranes (panel C), ACE2 enzymatic activity (panel D) and ACE2 mRNA (panel E) * p<0.05.).

Figure S2. ACE2 protein, activity and mRNA in kidneys from ACE8-8 mice and wild type controls. Panel A shows a representative Western blot for ACE, ACE2 and β-actin for three mice from each respective group. Panel B shows ACE2 protein in isolated membranes . Panel C shows ACE2 enzymatic activity level and panel D ACE2 mRNA in cell lysates ** p<0.01.

Figure S3 ACE2 activity in total cell lysates from lungs of captopril (A) and telmisartan (B) treated mice. No significant differences were detected.