Tissue-specific expression of the SARS-CoV-2 receptor, angiotensin-converting enzyme 2, in mouse models of chronic kidney disease

Elevated angiotensin-converting enzyme 2 (ACE2) expression in organs that are potential targets of severe acute respiratory syndrome coronavirus 2 may increase the risk of coronavirus disease 2019 (COVID-19) infection. Previous reports show that ACE2 alter its tissue-specific expression patterns under various pathological conditions, including renal diseases. Here, we examined changes in pulmonary ACE2 expression in two mouse chronic kidney disease (CKD) models: adenine-induced (adenine mice) and aristolochic acid-induced (AA mice). We also investigated changes in pulmonary ACE2 expression due to renin–angiotensin system (RAS) blocker (olmesartan) treatment in these mice. Adenine mice showed significant renal functional decline and elevated blood pressure, compared with controls. AA mice also showed significant renal functional decline, compared with vehicles; blood pressure did not differ between groups. Renal ACE2 expression was significantly reduced in adenine mice and AA mice; pulmonary expression was unaffected. Olmesartan attenuated urinary albumin excretion in adenine mice, but did not affect renal or pulmonary ACE2 expression levels. The results suggest that the risk of COVID-19 infection may not be elevated in patients with CKD because of their stable pulmonary ACE2 expression. Moreover, RAS blockers can be used safely in treatment of COVID-19 patients with CKD.

www.nature.com/scientificreports/ addition, nasal and branchial ACE2 expression is higher in adults than in children 14 . However, to the best of our knowledge, few studies have examined how pulmonary ACE2 expression is altered in other diseases. Patients with chronic kidney disease (CKD) have a risk of severe COVID-19 disease, but there is no evidence of increased prevalence of COVID-19 in CKD patients [15][16][17][18][19][20][21] . Considering the fact that ACE2 is required for COVID-19 infection, these findings may suggest that pulmonary ACE2 expression is not enhanced in patients with CKD. Notably, cellular ACE2 expression is presumably associated with susceptibility to the risk of infection by SARS-CoV, which is related to SARS-CoV-2 22 . The fact that children with low ACE2 expression are less susceptible to COVID-19 than adults 23 also supports these hypotheses. Therefore, we conducted experiments to examine changes in pulmonary ACE2 expression in two types of CKD model mice: adenine-induced (i.e., adenine mice) and aristolochic acid (AA)-induced (i.e., AA mice) to address this hypothesis.
The relationship of ACE2 expression with RAS blockers has also been investigated. Thus far, the use of RAS blockers has been shown to upregulate organ ACE2 expression in some instances [24][25][26][27] . Recent epidemiological investigations have shown that the use of RAS blockers does not increase the risk of COVID-19 [28][29][30] . Furthermore, ACE2 expression in the lungs was not enhanced when RAS blockers were administered to normal mice 31 . However, there have been no reports concerning the effects of RAS blockers on pulmonary ACE2 expression in the context of CKD; this point is essential for future management of COVID-19 and its sequelae. Here, we investigated changes in pulmonary ACE2 expression due to RAS blocker treatment in CKD model mice.
ACE2 expression in the kidneys was significantly reduced in adenine-induced CKD model mice, compared with controls, while ACE2 expression in the lungs did not differ between groups. Renal histology showed heterogeneous enlargement and atrophy of the tubules, as well as cellular infiltration into the renal interstitium, in adenine mice. ACE2 staining was observed mainly in the proximal tubules in both groups, but the degree of staining was reduced in the adenine group ( Fig. 2A). ACE2 protein levels in the kidneys (estimated by western blotting analysis) showed a significant reduction in the adenine group, compared with the control group, at both 2 and 4 weeks (2 weeks: adenine mice 0.55 ± 0.04 vs. control 1.00 ± 0.10, P < 0.001; 4 weeks: adenine mice 0.32 ± 0.07 vs. control 0.90 ± 0.07, P < 0.001; Fig. 2C). In addition, renal ACE2 mRNA expression showed a significant reduction in the adenine group at 2 weeks (adenine mice 0.61 ± 0.04 vs. control 1.00 ± 0.03, P < 0.001; Fig. 2E); a similar tendency for reduction was observed at 4 weeks, although the difference was not statistically significant (Fig. 2E). Lung histology findings did not differ between groups; ACE2 staining was observed in type 2 alveolar epithelial cells (Fig. 2B). There were no differences in ACE2 protein or mRNA expression levels in the lungs at 2 or 4 weeks (Fig. 2D, F). The difference in plasma ACE2 activity was not statistically significant (Fig. 2G).
AA-induced CKD model mice showed significant weight loss and renal functional decline, compared with the vehicle group, while BP did not differ between groups. Baseline BW and systolic BP were identical between vehicle and AA groups. BW increased over time in the vehicle group, while it decreased in the AA group; this BW gain was significantly different between groups (Fig. 3A). However, there was no difference in systolic BP between groups (Fig. 3B). Compared with the vehicle group, there were significant enhancements of plasma creatinine and BUN levels, as well as urinary albumin excretion, in the AA group (creatinine: AA mice 0.40 ± 0.03 mg/dL vs. vehicle 0.14 ± 0.01 mg/dL, P < 0.001; BUN: AA mice 62.8 ± 4.6 mg/ dL vs. vehicle 29.0 ± 1.4 mg/dL, P < 0.001; urinary albumin excretion: AA mice 61.3 ± 7.3 μg/day vs. vehicle 9.5 ± 0.9 μg/day, P < 0.001; Fig. 3C-E). Creatinine clearance also showed a significant reduction in the AA group (AA mice 72 ± 12 μL/min vs. vehicle 159 ± 13 μL/min, P < 0.001; Fig. 3F).
ACE2 expression in the kidneys was significantly reduced in AA-induced CKD model mice, compared with controls, while pulmonary ACE2 protein levels did not differ between groups. Renal histology analysis showed tubular atrophy and cellular infiltration into the renal interstitium in the AA group. ACE2 staining was observed mainly in the proximal tubules in both groups, but the degree of staining was reduced in the AA group (Fig. 4A). ACE2 protein levels in the kidneys showed a significant reduction in the AA group, compared with the vehicle group (AA mice 0.22 ± 0.03 vs. vehicle 1.00 ± 0.08, P < 0.001; Fig. 4C). In addition, renal ACE2 mRNA expression was significantly reduced in the AA group (AA mice 0.30 ± 0.02 vs. vehicle 1.00 ± 0.06, P < 0.001; Fig. 4E). Lung histology findings did not differ between groups; www.nature.com/scientificreports/ ACE2 staining was observed in type 2 alveolar epithelial cells (Fig. 4B) and there was no difference in pulmonary ACE2 protein level between groups (Fig. 4D). ACE2 mRNA expression in the lungs was significantly lower in the AA group than in the vehicle group (AA mice 0.76 ± 0.04 vs. vehicle 1.00 ± 0.04, P < 0.01; Fig. 4F). Moreover, there was a significant reduction in plasma ACE2 activity in the AA group (AA mice 84.8 ± 2.6 RFU/min vs. vehicle 98.3 ± 3.0 RFU/min, P < 0.01; Fig. 4G).
In adenine mice, olmesartan treatment improved plasma creatinine level (adenine mice 0.19 ± 0.02 vs. control 0.13 ± 0.01, P < 0.01; adenine + olmesartan mice 0.14 ± 0.01 vs. adenine mice 0.19 ± 0.02, P < 0.01; Fig. 5C) and suppressed urinary albumin excretion (adenine mice 67.9 ± 6.7 μg/day vs. control 18.0 ± 3.6 μg/day, P < 0.001;  Olmesartan did not affect ACE2 expression. In the control and adenine group, neither mRNA nor protein expression levels of ACE2 in the kidney were altered by olmesartan treatment (Fig. 6A, C). There were no changes in pulmonary ACE2 mRNA or protein expression levels between the control and adenine groups (Fig. 6B, D). There was also no significant difference in plasma ACE2 activity (Fig. 6E). ACE2 mRNA expression in the upper respiratory tract (pharynx) was also examined, but no significant difference was observed between the groups (Supplementary Figure S1).

Discussion
The COVID-19 pandemic has had a profound impact worldwide. For elderly individuals and people with comorbidities, information concerning the infection risk and severity of COVID-19 is a critical consideration. There have been numerous epidemiological studies of COVID-19 worldwide, which have shown that patients with CKD are more likely to develop severe COVID-19 17 but the morbidity of COVID-19 may not be elevated in patients with CKD [15][16][17][18][19][20][21] . Notably, the results of several meta-analysis revealed that the prevalence of CKD was 1.7-5.2% in patients with COVID-19 [15][16][17]21 . In addition, an analysis of 5,700 patients with COVID-19 in New    19 . Given that the estimated prevalence of CKD is 9.1% worldwide 18 and approximately 15% in the United States 20 , the risk of SARS-CoV-2 infection may not increase in relation to the presence of CKD. ACE2 plays an important role in the transmission of COVID-19 1,2 . In addition, the level of ACE2 expression in airway cells is associated with the risk of infection by SARS-CoV, a similar coronavirus to the causative agent of COVID-19 22 . Because COVID-19 is a respiratory disease, ACE2 expression in the lungs is an important component of disease transmission. In the present study, we hypothesized that the lack of increase in pulmonary ACE2 expression in CKD is linked to the absence of elevated COVID-19 morbidity. This hypothesis is supported by the fact that children with low ACE2 expression are less susceptible to COVID-19 than adults.
We used two types of CKD model mice in our experiments to examine changes in pulmonary ACE2 expression in connection with CKD pathogenesis. Adenine-induced CKD is a representative animal model of disease 32 . Excessive adenine administration can induce CKD by causing renal tubular obstruction and degeneration, as well as renal interstitial fibrosis, due to the deposition of 2-8 dihydroxyadenine 32,33 . In contrast, AA-induced CKD is caused by inducing fibrosis from AA-induced tubular damage 34 . These two models were chosen because BP . P values, versus control: *P < 0.05; **P < 0.01; ***P < 0.001; versus without olmesartan: † P < 0.05; † † P < 0.01; † † † P < 0.001. BP, blood pressure; BUN, blood urea nitrogen. www.nature.com/scientificreports/ has been shown to increase in the adenine-induced CKD model 35 , while it does not increase in the AA-induced CKD model 36 . Notably, hypertension is a common complication observed in patients with CKD; we presumed that it would be useful to investigate whether the presence or absence of hypertension influences pulmonary ACE2 expression in CKD pathogenesis. The results showed that the renal ACE2 protein levels were reduced in both CKD models. In adenine mice, there was no significant difference in mRNA levels at 4 weeks, but there was a decrease in protein levels. One of the reasons for this discrepancy may be the involvement of translational repression. Lambert DW, et al. showed that miR-421 repressed the translation of ACE2 protein without affecting ACE2 transcript levels 37 . Another study also showed that serum miR-421 was increased in CKD or hypertension patients 38,39 . Thus, in the adenine mice group with a longer period of CKD and hypertension (4 weeks), miR-421 www.nature.com/scientificreports/ might be increased and suppressed protein translation, resulting in a discrepancy between mRNA and protein expression. Pulmonary ACE2 protein levels did not differ in both CKD models compared to the controls and vehicles. Although it is intriguing that pulmonary ACE2 mRNA expression was reduced in the AA-induced CKD model, our findings were more notable because we found no difference in pulmonary ACE2 protein levels, which directly influence SARS-CoV-2 infection. In addition, plasma ACE2 activity did not differ in both CKD models compared to the controls and vehicles. Soluble ACE2 in plasma is considered to be cleaved and released from tissue cells. Increased plasma ACE2 activity has been reported to be associated with increase in release of ACE2 protein from tissues, suggesting a decrease in ACE2 protein in the tissues 40 . Thus, the results of the present study that plasma ACE2 activity was not changed in the CKD state indicates that the release of ACE2 protein expressed in tissues is not changed in the CKD state, which supports the findings that ACE2 protein expression in tissues is unchanged in the context of CKD.
The relationship between COVID-19 and RAS blockers is an ongoing focus of research discussions. Previous animal studies showed that RAS blockers upregulate tissue ACE2 expression, which led to fears that the use of RAS blockers may increase the risk of SARS-CoV-2 infection. However, the use of RAS blockers in those prior studies only led to upregulation of tissue ACE2 under specific pathological conditions 24,27 ; this upregulation did not exceed the normal range. Another study examining pulmonary and renal ACE2 expression in normal mice treated with RAS blockers showed no increase, compared with the control group 31 . On the other hand, there was a report that RAS blockers such as candesartan and captopril increased pulmonary ACE2 expression in normal rats 41 . In real-world clinical practice, some epidemiological studies concerning the use of RAS blockers in patients with COVID-19 have demonstrated that the use of those drugs does not increase the risk of COVID-19 [28][29][30] . However, the effect of RAS blockers on pulmonary ACE2 expression is still controversial. No previous studies have examined the effects of RAS blockers on pulmonary ACE2 expression in the context of CKD, although RAS blockers are essential for the treatment of patients with CKD. The present study was performed to address this gap in the literature. To more closely model real-world clinical practice, we used the adenine-induced CKD model, which exhibits concurrent hypertension. The results showed that administration of olmesartan, an RAS blocker, did not enhance pulmonary ACE2 expression, compared with controls. In addition, ACE2 mRNA expression in the upper respiratory tract, the initial target of SARS-CoV-2 infection, was not enhanced in response to olmesartan treatment. Renal ACE2 protein levels were reduced in the CKD model; administration of olmesartan did not increase renal ACE2 protein levels in the control and adenine mice. The lack of restoration in renal ACE2 in olmesartan-treated adenine mice might have been due to the use of a different CKD model, compared with a previous study 27 , as well as the relatively short duration of angiotensin receptor blocker treatment 25,27 . However, olmesartan treatment improved several renal failure parameters, including urinary albumin excretion. These results support the continued use of RAS blockers in patients with COVID-19 who have CKD.
This study had some limitations. First, it only examined changes in pulmonary ACE2 expression in CKD model mice, whereas it did not directly investigate SARS-CoV-2 infection. Second, we observed the adenine mice for 4 weeks and the AA mice for 8 weeks in this study, but the results may be different in a longer period of observation. Third, ACE2 expression in the lungs may differ between mice and humans. Fourth, in patients with CKD, COVID-19 severity is likely to be greater, but we could not refer to the underlying mechanism of severity. Fifth, it is not only the expression level of pulmonary ACE2 that determines the susceptibility to COVID-19, but it may also be affected by many other factors.
In conclusion, we found no change in pulmonary ACE2 protein expression in CKD model mice, regardless of hypertension status. In addition, the use of angiotensin receptor blocker treatment in CKD model mice did not increase pulmonary ACE2 expression. These results suggest the hypothesis that the risk of COVID-19 morbidity may not be elevated in patients with CKD because of their stable pulmonary ACE2 expression. The findings also provide important basic scientific evidence that RAS blockers can be used safely in treatment of patients with COVID-19 who have CKD.

Material and methods
Animals. This study was performed in accordance with the National Institutes of Health guidelines for the use of experimental animals. All animal experiments were reviewed and approved by the Animal Studies Committee of Yokohama City University (Approval Number: FA20-027), which were in compliance with the ARRIVE guidelines. Efforts were made to minimize the number of animals used and ensure minimal sufferings. The mice were housed in a controlled environment with a 12-h light-dark cycle at a temperature of 25 °C. The mice were allowed free access to food and water.
In the olmesartan treatment experiment, mice were randomly divided into four experimental groups: (1) control group; (2) adenine group, fed a diet mixed with 0.2% adenine; (3) control-olmesartan group, treated with the AT1 receptor antagonist olmesartan in drinking water (4 mg/kg/day; Daiichi Sankyo Chemical Pharma Co., Ltd, Tokyo, Japan); and (4) adenine-olmesartan group, which doses identical to those in groups 2 and 3. Kikai Co., Tokyo, Japan), as described previously 42,43 . All measurements were performed between 9:00 and 14:00 h. At least 10 measurements were performed in each mouse and the mean value was used for analysis.

Real-time quantitative reverse transcription polymerase chain reaction analysis. Total RNA
was extracted from lung, renal and pharynx tissue using ISOGEN (Nippon Gene, Tokyo, Japan); cDNA was synthesized using the SuperScript III First-Strand System (Invitrogen, Carlsbad, CA, USA). Real-time quantitative reverse transcription polymerase chain reaction analysis was performed using an ABI PRISM 7000 Sequence Detection System; reverse transcription products were incubated with TaqMan PCR Master Mix and a custom TaqMan probe (Applied Biosystems, Foster City, CA, USA), as described previously 44,45 . The following TaqMan probe was used: ACE2 (Mn01159003_m1). mRNA levels were normalized to those of 18S rRNA.
Western blotting analysis. Protein expression was analyzed by western blotting using tissue homogenates, as described previously 45,46 . Briefly, total protein extract was prepared from tissues with sodium dodecyl sulfate-containing sample buffer. The protein concentration of each sample was measured with NanoDrop One (Thermo Fisher Scientific), using bovine serum albumin as the standard. Equal amounts of protein extract from each tissue samples (lung tissue: 24 μg, kidney tissue: 10 μg) were fractionated on a 5-20% polyacrylamide gel (Atto, Tokyo, Japan). The separated proteins were then transferred to a polyvinylidene difluoride membrane using the iBlot Dry Blotting System (Invitrogen  Figure S3).

Immunohistochemical analysis.
Lung and renal tissues from mice were fixed with 4% paraformaldehyde and subsequently embedded in paraffin. Four-micrometer-thick sections were dewaxed and rehydrated; antigen retrieval was performed by microwave heating. The sections were blocked to reduce endogenous biotin activity using peroxidase blocking reagent (Dako, Carpinteria, CA, USA) and treated for 60 min with 10% normal goat serum in phosphate-buffered saline. The sections were then incubated with anti-ACE2 antibody diluted to 1:500 (Ab15348, Abcam). To examine the specificity of the antibody, immunostaining was also examined by omission of primary Ab and using an ACE2-selective blocking peptide (ab15352). ACE2 staining in lung and renal tissues was observed, which was not observed when the antibody was preabsorbed with an ACE2-selective blocking peptide (ab15352) or omission of ACE2 antibody. (Supplementary Figure S4).

Biochemical analysis.
After inhaling 5% isoflurane anaesthesia, Blood samples were collected by cardiac puncture in the fed state by heart punctures. Experimental animals were killed humanely after anaesthesia. Whole blood samples were centrifuged at 3000 rpm (MR-150, Tomy Seiko Co., Ltd., Tokyo, Japan) at 4 °C for 10 min to separate the plasma. The resulting plasma samples were stored at − 80 °C until use. Plasma creatinine, BUN, urinary creatinine, and urinary albumin levels were measured using a Hitachi 7180 autoanalyzer (Hitachi, Tokyo, Japan).
Plasma ACE2 activity. ACE2 activity was measured by TechnoPro R&D Company (Tokyo, Japan) using the ACE2 Activity Assay Kit (SensoLyte 390) at a temperature of 27 °C. Samples were diluted 1:10 using the assay buffer provided with the ACE2 Activity Assay Kit. The assay was performed in 384-well plates. Plasma samples, assay buffer only (background), or 4-methyl-coumaryl-7-amide (0.16-5 μM) as a reference standard were added to an OptiPlate-384 F reaction plate at 10 μL/well. Then, ACE2 substrate solution (0.05 mM 4-methylcoumaryl-7-amide/Dnp in assay buffer) was added at 10 μL/well. After the reaction, fluorescence intensity (Ex/ Em = 330 nm/390 nm) was measured at 5-min intervals for 3 h using an EnSpire plate reader (PerkinElmer, Waltham, MA, USA). The fluorescence intensity of samples at each time point was calculated from the fluorescence intensity of the samples. The slope of each sample (RFU/min) was determined from a linear approximation of the plot during the period from 30 to 90 min after the start of the measurement.
Statistical analysis. Data are expressed as mean ± standard error of the mean. Differences were analyzed as follows. Two-way analysis of variance, followed by Bonferroni post hoc analysis, was performed to determine differences over time between adenine and control groups (Figs. 1, 2), AA and vehicle groups (Fig. 3A, B) or adenine and olmesartan groups (Figs. 5, 6 and S1). Unpaired t-tests were used to determine differences between AA and vehicle mice (Figs. 3C-F, 4). P values < 0.05 were considered statistically significant.