- Email: [email protected]
Decreased glomerular and tubular expression of ACE2 in patients with type 2 diabetes and kidney disease
original article
http://www.kidney-international.org & 2008 International Society of Nephrology
Decreased glomerular and tubular expression of ACE2 in patients with type 2 diabetes and kidney disease Heather N. Reich1, Gavin Y. Oudit2, Josef M. Penninger3, James W. Scholey1 and Andrew M. Herzenberg4 1
Division of Nephrology, Department of Medicine, University Health Network, University of Toronto, Toronto, Ontario, Canada; 2Division of Cardiology, Department of Medicine, University Health Network, University of Toronto, Toronto, Ontario, Canada; 3Institute for Molecular Biotechnology of the Austrian Academy of Sciences, Austria and 4Department of Laboratory Medicine and Pathology, University Health Network, University of Toronto, Toronto, Ontario, Canada
Angiotensin converting enzyme (ACE) generates angiotensin II from angiotensin I, which plays a critical role in the pathophysiology of diabetic nephropathy. However, ACE2 generates angiotensin 1–7, which may protect the kidney by attenuating the effects of angiotensin II, since deletion of the Ace2 gene leads to glomerulosclerosis in mice, and pharmacologic inhibition of ACE2 exacerbates experimental diabetic nephropathy. We measured ACE2 and ACE expression in renal biopsies of patients with kidney disease due to type 2 diabetes to determine if the expression pattern is specific to diabetic nephropathy. ACE2 and ACE mRNA levels were measured by real-time PCR in laser microdissected renal biopsies from 13 diabetic and 8 control patients. ACE2 mRNA was significantly reduced by more than half in both the glomeruli and proximal tubules of the diabetic patients compared to controls, but ACE mRNA was increased in both compartments. There was a significant parallel decrease in ACE2 protein expression, determined by immunohistochemistry, in proximal tubules, a pattern not found in 12 patients with focal glomerulosclerosis or 10 patients with chronic allograft nephropathy. Our results suggest that the kidney disease of patients with type 2 diabetes is associated with a reduction in ACE2 gene and protein expression and this may contribute to the progression of renal injury. Kidney International (2008) 74, 1610–1616; doi:10.1038/ki.2008.497; published online 1 October 2008 KEYWORDS: renin-angiotensin system; diabetes mellitus; glomerulosclerosis; real-time RT-PCR; immunohistochemistry; ACE2
Correspondence: Andrew M. Herzenberg, Department of Pathology, University Health Network, 11th floor, 200 Elizabeth Street, Toronto, Ontario, Canada M5G 2C4. E-mail: [email protected] Received 6 July 2008; revised 17 July 2008; accepted 29 July 2008; published online 1 October 2008 1610
The renin-angiotensin system (RAS) is important in the pathophysiology of many progressive renal diseases including diabetic nephropathy (DN), and blockade of the RAS system with either angiotensin-converting enzyme (ACE) inhibition or Ang II receptor blockade attenuates progression of glomerular disease including DN.1–3 Angiotensin-converting enzyme-2 (ACE2) is a recently discovered homologue of ACE that regulates activity of the RAS by reducing Ang II levels and increasing the generation of Ang 1–7, and therefore expression of ACE2 in the kidney may be an important determinant of injury.4 Microdissection studies of the rat kidney have shown that ACE2 mRNA levels can be detected in almost every segment of the nephron including the glomerulus and the proximal tubule (but not the medullary thick ascending limb).5 ACE2 is located along the apical surface of proximal tubules,6 and in the glomerulus ACE2 is expressed in podocytes and mesangial cells.7,8 Recent studies have shown that deletion of the Ace2 gene leads to the development of de novo glomerulosclerosis in mice and pharmacologic inhibition of ACE2 worsens experimental DN,7,9,10 but there are limited data on the expression of ACE2 in the kidneys of normal human subjects and subjects with DN.8,11 In a recent report, Lely et al.8 studied ACE2 protein expression by immunohistochemistry in the kidneys of patients with a variety of glomerular and non-glomerular diseases. They observed neo-expression of ACE2 protein in glomerular and peritubular capillary endothelial cells, but DN constituted only a small minority of the study patients, and disease-specific information on ACE2 expression was not reported.8 Konoshita et al.11 measured ACE2 mRNA levels in whole-kidney biopsy cores from subjects with diabetic and nondiabetic kidney disease. They found that ACE2 mRNA levels were similar in the two groups but no control subjects were included in the study. Accordingly, the goal of the current study was to determine if DN is associated with glomerular and tubular changes in ACE2 gene and protein expression in subjects with type 2 diabetes mellitus (DM) compared to control subjects with normal kidney function. To test this hypothesis, we Kidney International (2008) 74, 1610–1616
original article
HN Reich et al.: ACE2 and diabetic nephropathy
RESULTS Patient and biopsy characteristics
The clinical parameters of study subjects are provided in Table 1. Subjects with DN (n ¼ 13) tended to be older, and there was a higher ratio of male subjects in comparison with control subjects (n ¼ 8). As expected, subjects with DN had higher hemoglobin A1C levels, and impaired renal function. Pathologic review confirmed that all of the biopsies from subjects with diabetes showed diffuse or diffuse and nodular diabetic glomerulosclerosis with marked mesangial matrix expansion, diffuse glomerular basement membrane thickening, and mild to moderate interstitial fibrosis. All of the kidney biopsies from the potential living donors were normal at the light microscopic level but five subjects had evidence of thin glomerular basement membranes by electron microscopy. Subjects did not have uncontrolled hypertension at the time of biopsy. None of the control subjects had a history of hypertension (data available for 6 of 8 subjects), and all of the diabetic subjects had a history of hypertension (data available for 7 of 13 subjects).
the glomerular compartment (Po0.05) of biopsies from patients with DN vs controls (Figure 1a). The abundance of ACE2 mRNA was generally lower in the glomerular compartment than in the tubular compartment in both groups. As illustrated in Figure 1b, the expression of ACE was twice as high in both the tubular and glomerular compartment of subjects with DN vs control (Po0.05). Effect of ACE inhibitor or angiotensin receptor blocker treatment on ACE2 expression in the diabetic kidney
Seven of the thirteen diabetic subjects were treated with ACE inhibitors or angiotensin receptor blocker at the time of the renal biopsy, and therefore kidney ACE2 expression was compared in the treated and untreated diabetic subjects (Table 2). Median values for creatinine and proteinuria were similar in the two groups of diabetic subjects and there were no differences in glomerular or tubular expression of ACE2. Fibrosis scores and gene expression
ACE2 and ACE expression were related to semiquantitative scores of segmental glomerulosclerosis, global glomerulosclerosis, and tubulointerstitial fibrosis in DM biopsies. We did not find a relationship between gene expression and these scores (data not shown). Immunohistochemistry
Immunohistochemical analysis of ACE2 protein expression was assessed in the same renal biopsies and mirrored the 1.2
ACE2 and ACE gene expression
Table 1 | Clinical parameters for the control and diabetic subjects Control (n=8) Age (years) M:F HbA1C (%)a Median SCr (mmol/l)a Median eGFR (ml/min/1.73 M2)a Median proteinuria (g/24 h)a % treated with ACEI (n)
44±3 1:7 5.1±0.1 78 (52–115) 87 (46–122) 0.125 (0.1–1) 13 (1)
2.12 (0.19–12)* 54 (7)*
ACEI, angiotensin-converting enzyme inhibitor; eGFR, estimated glomerular filtration rate based on the MDRD equation; F, female; HbA1C, hemoglobin A1C; M, male; proteinuria, 24-h urine protein excretion; SCr, serum creatinine. Data are presented as mean±s.e.m. or median and range. * Po0.05 vs the control group as indicated; #P=0.06 vs the control group. a Based on available data.
Kidney International (2008) 74, 1610–1616
Controls DN
0.6
*
3
Controls DN 0.3
Diabetic nephropathy (n=13) 50±3 6:7 9.2±0.7* 234 (45–530)* 29 (8–145)#
*
0.9
ACE2 mRNA (AU)
Laser capture microdissection of glomeruli and tubules was performed on frozen sections. There was a 58% decrease in the expression of ACE2 in the proximal tubules of subjects with DN vs controls as quantified by real-time PCR (Po0.05), and a 56% decrease in ACE2 gene expression in
6
ACE mRNA (AU)
measured ACE2 mRNA levels with real-time PCR in lasercaptured glomeruli and proximal tubules from kidney biopsy samples of control subjects and subjects with type 2 DM and DN, and we assessed ACE2 protein expression in the kidney with immunohistochemistry and computer-assisted image analysis. To determine if the pattern of ACE2 expression is specific to diabetes, we also compared ACE2 mRNA and protein levels in microdissected kidney biopsies of subjects with nondiabetic kidney disease: primary focal segmental glomerulosclerosis (FSGS) and chronic allograft nephropathy (CAN).
* * 0
0 Proximal Glomerular tubular compartment compartment
Proximal Glomerular tubular compartment compartment
Figure 1 | ACE2 and ACE mRNA expression in human diabetic nephropathy (DN). Real-time RT-PCR from glomerular and proximal tubular compartments, following laser capture microdissection, from kidney biopsies of human subjects with DN (n ¼ 13) and controls (n ¼ 8). (a) ACE2 mRNA expression. (b) ACE mRNA expression. All real-time PCR data are expressed in arbitrary units (AUs) corrected for GAPDH expression. *Po0.05 vs nondiabetic controls. 1611
original article
HN Reich et al.: ACE2 and diabetic nephropathy
Table 2 | The effect of ACE inhibitor or angiotensin receptor blocker treatment on ACE2 expression in diabetic subjects a
Median SCr (mmol/l) Median eGFR (ml/min/1.73 M2)a Median proteinuria (g/24 h)a Glomerular ACE2, mRNA (AU) Tubular ACE2, mRNA (AU)
No ACEI (n=6)
ACEI (n=7)
220 (45–274) 30 (17–135) 2.18 (0.19–3.2) 0.070±0.040 0.52±0.22
321 (56–530) 18 (8–145) 2.17 (1.36–12) 0.075±0.060 0.32±0.37
ACEI, angiotensin-converting enzyme inhibitor; AU, arbitrary unit; eGFR, estimated glomerular filtration rate based on the MDRD equation; SCr, serum creatinine. AU corrected for GAPDH expression. Data are presented as mean±s.e.m. or median and range as indicated. There were no significant differences between the two groups of diabetic subjects. a Based on available data; proteinuria (24-h urine protein excretion).
Table 3 | Clinical parameters for the control, FSGS, and CAN subjects
Age (years) M:F Median S*Cr Median eGFR* Median proteinuria* % Rx ACEI (n)*
Control (n=10)
FSGS (n=12)
CAN (n=10)
43.6±2.6 2:8 69 (53–97) 93.5 (56–124) n/a 0
42.0±5.2 7:5 131 (45–273) 49.5 (22–132) 3.65 (0.8–10.4) 45.5 (5)
50.3±4.3 6:4 199 (122–270) 30 (18–50) o0.30 (o0.3–1) 20 (2)
CAN, chronic allograft nephropathy; eGFR, estimated glomerular filtration rate based on the MDRD equation; F, female; FSGS, focal segmental glomerulosclerosis; M, male; proteinuria, 24-h urine protein excretion; Rx ACEI, treated with an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker treated; SCr, serum creatinine. Data are presented as mean±s.e.m. or median and range as indicated. *Po0.05 across groups.
Control
DN
60 Control DN
40
*
20
Control 1
FSGS CAN
0.5 0
0 Proximal Glomerular tubular compartment compartment
Figure 2 | Immunohistochemical staining for ACE2 protein in renal biopsies in diabetic nephropathy (DN) and controls. (a) A representative study of patient with DN and a normal live human kidney donor, and (b) quantitative computer image analysis of ACE2 immunohistochemistry staining intensity of diabetic humans (n ¼ 13) and controls (n ¼ 8) (*Po0.05 vs nondiabetic controls).
measures of mRNA expression (Figure 2a). Image analysis confirmed a 60% reduction in immunoperoxidase staining for ACE2 protein in the proximal tubules of the diabetic kidney biopsies compared with the nondiabetic control biopsies (Figure 2b). There was no significant relationship between the tubular expression of ACE2 protein and the level of proteinuria in the diabetic subjects by linear regression analysis. The ACE2 protein levels were lower than the limit of immunohistochemical detection in the glomerular compartment in both groups. ACE2 and ACE expression in FSGS and CAN
To determine if these changes in ACE2 and ACE expression are unique to DN, gene expression was also measured in the biopsies of subjects with FSGS and CAN. The clinical characteristics of these subjects are provided in Table 3. There was a significant difference in estimated glomerular filtration rate (eGFR), serum creatinine, proteinuria, and ACE inhibitor/angiotensin receptor blocker use across groups (Po0.05). The results of measurement of ACE2 and ACE mRNA expression are shown in Figure 3. There was no significant difference in glomerular or tubular ACE2, or 1612
1.5
Proximal Glomerular tubular compartment compartment
ACE mRNA (AU)
2
80 ACE2 mRNA (AU)
Immunohistochemistry score for ACE2 expression
100
10 9 8 7 6 5 4 3 2 1 0
Control FSGS CAN
Proximal Glomerular tubular compartment compartment
Figure 3 | ACE2 and ACE mRNA expression in focal segmental glomerulosclerosis (FSGS), chronic allograft nephropathy (CAN), and control biopsies. (a) ACE2 mRNA expression. (b) ACE mRNA expression. Real-time RT-PCR from proximal tubular and glomerular compartments, following laser capture microdissection from patients with FSGS (n ¼ 12), CAN (n ¼ 10), and controls (n ¼ 10). Results are expressed as arbitrary units (AUs) corrected for GAPDH. There were no significant differences across or between the three groups.
glomerular ACE expression, in FSGS or CAN compared with control biopsies (Figure 3). There was a trend toward increased tubular ACE expression in FSGS but this did not reach statistical significance. Similarly, there was no significant difference in ACE2 protein expression as measured by immunohistochemistry in either the tubular or glomerular compartment (Figure 4). When subjects with either FSGS or CAN were divided according to ACE inhibitor/ angiotensin receptor blocker use, there was an increase in the tubular expression of ACE2 noted in subjects taking these medications (2.15±1.34, n ¼ 7 users vs 0.67±0.82, n ¼ 14 nonusers, P ¼ 0.03). There was no difference in glomerular ACE2, tubular or glomerular ACE, when subjects were analyzed according to medication use. DISCUSSION
In this study we observed decreased ACE2 mRNA levels and protein expression in kidneys of human subjects with DN compared to nondiabetic healthy control subjects. The decrease in ACE2 expression occurred in both glomeruli and proximal tubule cells and the effect of DM on mRNA Kidney International (2008) 74, 1610–1616
original article
HN Reich et al.: ACE2 and diabetic nephropathy
Immunohistochemistry score for ACE2 expression
100
80
60
20
0 Control
FSGS
CAN
Control FSGS CAN
40
Proximal Glomerular tubular compartment compartment
Figure 4 | Immunohistochemical staining for ACE2 protein in renal biopsies in focal segmental glomerulosclerosis (FSGS), chronic allograft nephropathy (CAN), and controls. (a) Representative study patients with idiopathic FSGS, CAN, and a normal live kidney donor. (b) Quantitative computer image analysis of ACE2 immunohistochemistry staining intensity in biopsies of patients with FSGS (n ¼ 12), CAN (n ¼ 10), and in controls (n ¼ 10). There were no significant differences across or between the three groups.
Table 4 | Summary of available literature describing ACE2 expression in glomeruli and renal tubules in diabetic nephropathy Glomerulus
Proximal tubules/cortex
ACE2 mRNA
ACE2 protein
ACE2 mRNA
ACE2 protein
ACE2 activity
Human Reich et al. Lely et al.8 Konoshita et al.11
k NA NA
BD m NA
k NA 2
k 2 NA
NA NA NA
Animal models Tikellis et al.17 Streptozotocin rats (type 1 DM) Ye et al.7,19; Wyscocki et al.18db/db mice (type 2 DM) Wysocki et al.18 Streptozotocin mice (type 1 DM) Wong et al.13 Akita mice (type 1 DM)
m NA NA NA
m k NA NA
k 2 2 m
k m m NA
NA m m NA
k, decreased ACE2 expression; m, increased ACE2 expression; 2, no significant difference in expression; BD, below detection limit; DM, diabetes mellitus; NA, not analyzed. The results of the study reported herein and other human and rodent studies of renal ACE2 expression in diabetic nephropathy are summarized. Studies reporting mRNA expression of whole cortex renal tissue were considered to represent the proximal tubular compartment for the purposes of this summary.
levels and protein levels was similar in both of these compartments. ACE2 is a monocarboxypeptidase that cleaves a single amino acid from the C terminus of Ang I and Ang II, which leads to the generation of Ang 1–9 and Ang 1–7, respectively. Ang 1–7 has vasodilatory and antiproliferative effects, and it antagonizes Ang-II-mediated cell signaling including transforming growth factor-b 1 expression.12 Together with the reports of gene deletion studies and pharmacologic ACE2 blockade,7,9,10,13,14 our findings on ACE2 expression in the diabetic kidney are consistent with the notion that ACE2 serves a protective role in the kidney, and support the hypothesis that reduced ACE2 expression (and activity) may contribute to the development and progression of kidney injury, including DN. However, the data do not establish a causal relationship between ACE2 expression and DN nor the mechanism(s) responsible for the decrease in ACE2. In contrast to the change in ACE2 expression, ACE expression increased in both the glomerular and proximal tubule cells of the diabetic kidneys compared to the control kidneys, in accord with previous observations by Kidney International (2008) 74, 1610–1616
Zimpelmann et al.15 in the kidneys of streptozotocin-induced diabetic rats, and the observations of Ye et al.7 in the glomeruli of diabetic db/db mice. It is possible that these relative changes might lead to a shift in angiotensin peptide processing toward generation of Ang II and away from the generation of Ang 1–7 in the diabetic kidney, an effect that is predicted to be deleterious. In support of this hypothesis, Huang et al.16 reported that there was a relationship between the development of DN and levels of ACE expression in transgenic mice. We could not measure angiotensin peptides in our kidney biopsy samples so the effect of the changes in ACE2 and ACE expression on angiotensin peptide processing remains speculative. Table 4 summarizes the current literature regarding ACE2 expression and activity in experimental and clinical DN. Tikellis et al.17 were the first to study ACE2 expression in rats with streptozotocin-induced diabetes, a model of type 1 DM. They reported that ACE2 mRNA and protein levels were reduced by 50% in the renal tubules of diabetic rats, similar to our findings, and they also observed that ACE2 expression was mainly in the tubular compartment. In contrast to our 1613
original article
results in human kidney biopsies, DM was associated with an increase in glomerular ACE2 expression. Interestingly, Wysocki et al.18 studied ACE2 expression and activity in the kidney cortex of mice with streptozotocin-induced DM and they reported that ACE2 protein expression and ACE2 activity were increased compared to nondiabetic mice. Similar findings were reported in the db/db mouse, a model of type 2 DM.18,19 Recently, the same group reported that ACE2 immunostaining was reduced in the glomeruli of the db/db mice compared to nondiabetic db/m mice.7 Taken together, it is difficult to reconcile these observations on ACE2 expression in experimental DM, and the effect of diabetes may be specific for each model and dependent on the stage of injury. There are limited but incomplete data on ACE2 expression in human DN. A study comparing ACE2 mRNA levels in the biopsies of patients with diabetic and nondiabetic kidney disease demonstrated no differences in ACE2 expression across a wide range of disease categories.11 However, diabetic subjects consisted only a small subset of the study subjects (8 of 74 subjects), gene expression was not compared with that of a normal control group, and the mRNA levels were not correlated with protein expression. Lely et al.8 reported that ACE2 protein immunostaining was uniformly increased in the glomeruli and peritubular capillaries in kidney biopsies representing a variety of kidney diseases, including some subjects with DN, however this was assessed semiquantitatively, and mRNA levels, which may be more sensitive given the relatively low abundance of the protein, were not assessed. Furthermore, the relative difference in ACE2 expression between DN and normal kidney biopsies was not a specific focus of this study.8 The effect of blockade of the RAS on ACE2 expression in the human kidney has not been well studied although Tikellis et al.17 have examined the effect of ACE inhibitor treatment on ACE2 expression in the kidneys of rats with streptozotocin-induced DM. They reported that ACE2 mRNA levels were low in the diabetic kidney and unaffected by treatment with ramipril, in accord with our observations in the kidneys of subjects with type 2 DM. They did report that ACE2 protein levels were markedly increased in association with ramipril treatment (in contrast to the effect on mRNA levels) in diabetic rats, but we did not observe any discordance between measures of ACE2 mRNA expression and protein expression in our kidney biopsy samples. Ferrario et al.20 reported studies of the effect of ACE inhibition, angiotensin receptor blockade, and combination therapy on ACE2 mRNA levels in the kidneys of normal rats. ACE2 mRNA levels were not affected by any of the treatments. To determine if the pattern of ACE2 and ACE expression that we observed in the kidneys of subjects with DN was specific for DN or might represent a generalized response to kidney injury, we studied two separate groups of subjects with kidney disease: FSGS and CAN. We chose FSGS because like DN, it is also a glomerular disease marked by proteinuria; we chose CAN as it represents a chronic progressive kidney 1614
HN Reich et al.: ACE2 and diabetic nephropathy
disease that is not glomerular in origin, and it is characterized by prominent tubulointerstitial injury. We found that ACE2 and ACE mRNA levels were not different in the kidney biopsies of subjects with FSGS or CAN compared to controls, suggesting that decline in ACE2 expression is not a general feature of kidney injury and that there is some specificity of this response for DN. Interestingly, blockade of the RAS (with ACE inhibitor/angiotensin receptor blocker) in subjects with either FSGS or CAN was associated with an increase in the tubular expression of ACE2, and it is tempting to speculate that ACE inhibitor/angiotensin receptor blocker use may alter ACE2 levels in subjects without diabetes. The strengths of the current study are that we compared ACE2 and ACE expression in the kidney of subjects with type 2 DM with the kidneys of healthy control subjects, and we localized expression at the mRNA level by performing laser capture microscopy of the glomeruli and proximal tubule cells. We also related ACE2 mRNA levels with protein expression by immunohistochemistry and image analysis. However, this report is descriptive, and we did not correlate measures of mRNA and protein expression levels with ACE2 activity in the kidney or with measures of angiotensin peptides. In addition, gender and hypertension may be confounding variables in our analysis. Women were overrepresented in our control group, largely because of the fact that our control subjects were potential living kidney donors, which tend to be women at our center.21 A history of hypertension was prominent in our diabetic subjects but we could not correct our analysis for blood pressure levels. In conclusion, ACE2 mRNA levels and protein expression decrease in the proximal tubules of patients with DN compared with nondiabetic healthy control subjects, and a similar trend is observed in the glomeruli. ACE2 expression is relatively low in human glomeruli compared to the proximal tubule cells in both normal and diabetic kidneys, and treatment with an ACE inhibitor does not affect ACE2 expression in the diabetic kidney. Finally, the pattern of decreased ACE2 and increased ACE expression was not observed in the biopsies of subjects with FSGS or CAN. In subjects with diabetes, the combination of effects of altered ACE2 and ACE expression may contribute to the progression of DN. MATERIALS AND METHODS Renal biopsies
Renal biopsies had been performed as part of routine clinical diagnostic investigation. Informed consent was obtained from patients for use of archived human biopsy material and chart review in accord with institutional research ethics board review. The diabetes samples were obtained from patients with documented type 2 diabetes and a pathologic diagnosis of DN. Control tissues were obtained from healthy potential live transplant donors who were referred for biopsy because of microscopic hematuria, who had normal renal function and no proteinuria. Control renal biopsies were either normal or showed thin basement membranes accordKidney International (2008) 74, 1610–1616
original article
HN Reich et al.: ACE2 and diabetic nephropathy
ing to pathologist review. Routine immunofluorescence microscopy of these biopsy samples showed negative staining for immunoglobulins and complement. Laser capture microdissection and real-time PCR
Snap-frozen renal biopsies stored at 80 1C were cut at 8 mm and stained with HistoGene (Arcturus, Mountainview, CA, USA) and laser microdissected using the Pixcell IIe (Arcturus). Glomerular profiles and kidney proximal tubules were identified by an experienced renal pathologist (AMH), separately dissected, and collected for mRNA extraction. The mRNA was extracted using the PicoPure kit (Arcturus) and first-strand cDNA synthesis was performed using the Sensiscript RT Kit (Qiagen, Valencia, CA, USA). The cDNA was used undiluted for TaqMan real-time PCR on an Applied Biosystems 7900 machine. The ACE2 gene expression was expressed as a ratio of ACE2/GAPDH (glyceraldehyde-3phosphate dehydrogenase) housekeeping gene. There was no significant difference in GAPDH levels in disease samples compared to control samples, and we have found that GAPDH is a more consistent reference mRNA than 18S or bactin. Primers and probe for amplification were as follows— ACE2: CATTGGAGCAAGTGTTGGATCTT (forward), GAG CTAATGCATGCCATTCTCA (reverse), CTTGCAACACCAG TTCCCAGGCA (probe); ACE: CCGAAATACGTGGAACT CATCAA (forward), CACGAGTCCCCTGCATCTACA (reverse), CAGGCTGCCCGGCTCAATGG (probe). GAPDH primer sets were designed as follows: GAAGGTGAAGGTCGGAGTC (forward), GAAGATGGTGATGGGATTTC (reverse), and CAA GCTTCCCGTTCTCAGCC (probe). In addition, GAPDH primers were also purchased from Applied Biosystems.
Wilcoxon) analyses, and similar results were obtained using both approaches. Linear regression was used to relate the ACE2 expression to the level of proteinuria and fibrosis in the diabetic subjects. All statistical tests were performed using SAS (SAS Institute, Cary, NC, USA), and two-tailed P-values are provided. DISCLOSURE
All the authors declared no competing interests. ACKNOWLEDGMENTS
Andrew Herzenberg and James Scholey are supported by a Canadian Institutes of Health Research (CIHR) New Emerging Team (NET) grant (Genes, Gender and Glomerular-based Diseases NET). Andrew Herzenberg, James Scholey, and Gavin Oudit are the recipients of a Canadian Diabetes Association (CDA) grant. Heather Reich is the recipient of a KRESCENT Trainee Fellowship Award (CIHR, Kidney Foundation of Canada, and Canadian Society of Nephrology). James Scholey is the recipient of a CIHR-AMGEN Canada Research Chair in Nephrology. Gavin Oudit is a clinician-scientist supported by the Heart and Stroke Foundation of Canada, Canadian Institute for Health Research, and the TACTICS program. Josef Penninger is supported by a Eugene Heart grant. We acknowledge the expert technical assistance of Brent Steer and Dr Phil Marsden as well as Stuart Yang. REFERENCES 1.
2.
3.
4.
Immunohistochemistry
Formalin-fixed paraffin-embedded sections from the same renal biopsy were cut at 3 mm followed by heat-induced antigen retrieval. Sections were incubated with monoclonal anti-ACE2 (1:400) antibodies.14 To eliminate any potential nonspecific biotin activity, slides were stained with a secondary anti-mouse antibody using the EnVision system (Dako). Endogenous peroxidase activity was prevented by pretreating with 3% hydrogen peroxide. Image-Pro Plus (Media Cybernetics) computer image analysis software was used to analyze brown staining pixel density and quantify protein levels. Negative controls with irrelevant primary antibody and no primary antibody were performed. Calculations and statistics
The eGFR was calculated using the abbreviated Modification of Diet in Renal Disease equation.22,23 The distribution of clinical data was visualized and summary data are expressed as mean±standard error or median and range as indicated. Differences in means were analyzed using nonparametric tests when data were not normally distributed. Gene and protein expression data were analyzed using both parametric (for example, analysis of variance, with Levene’s test of the assumption of equal variance) and nonparametric (that is, Kidney International (2008) 74, 1610–1616
5.
6.
7.
8. 9.
10.
11.
12.
13.
14. 15.
Lewis EJ, Hunsicker LG, Bain RP, et al., for the Collaborative Study Group. The effect of angiotensin-converting enzyme inhibition on diabetic nephropathy. N Engl J Med 1993; 329: 1456–1462. Lewis EJ, Hunsicker LG, Clarke WR, et al., for the Collaborative Study Group. Renoprotective effect of the angiotensin-receptor antagonist irbesartan in patients with nephropathy due to type 2 diabetes. N Engl J Med 2001; 345: 851–860. Brenner BM, Cooper ME, de Zeeuw D, et al., RENAAL Study Investigators. Effects of losartan on renal and cardiovascular outcomes in patients with type 2 diabetes and nephropathy. N Engl J Med 2001; 345: 861–869. Burrell LM, Johnston CI, Tikellis C et al. ACE2, a new regulator of the reninangiotensin system. Trends Endocrinol Metab 2004; 15: 166–169. Li N, Zimpelmann J, Cheng K et al. The role of angiotensin converting enzyme 2 in the generation of angiotensin 1–7 by rat proximal tubules. Am J Physiol Renal Physiol 2005; 2: 353–362. Warner FJ, Lew RA, Smit AI et al. Angiotensin-converting enzyme 2 (ACE2), but not ACE, is preferentially localized to the apical surface of polarized kidney cells. J Biol Chem 2005; 280: 39353–39361. Ye M, Wysocki J, William J et al. Glomerular localization and expression of angiotensin-converting enzyme 2 and angiotensin converting enzyme: implications for albuminuria in diabetes. J Am Soc Nephrol 2006; 17: 3067–3075. Lely AT, Hamming I, van Goor H et al. Renal ACE2 expression in human kidney disease. J Pathol 2004; 204: 587–593. Oudit GY, Herzenberg AM, Kassiri Z et al. Loss of angiotensin-converting enzyme 2 leads to the late development of angiotensin II-dependent glomerulosclerosis. Am J Pathol 2006; 168: 1808–1820. Soler MJ, Wysocki J, Ye M et al. ACE2 inhibition worsens glomerular injury associated with increased ACE expression in streptozotocin-induced diabetic mice. Kidney Int 2007; 72: 614–623. Konoshita T, Wakahara S, Mizuno S et al. Tissue gene expression of reninangiotensin system in human type 2 diabetic nephropathy. Diabetes Care 2006; 29: 848–852. Su Z, Zimpelmann J, Burns KD. Angiotensin-(1-7) inhibits angiotensin IIstimulated phosphorylation of MAP kinases in proximal tubular cells. Kidney Int 2006; 69: 2212–2218. Wong DW, Oudit GY, Reich H et al. Loss of angiotensin converting enzyme-2 (Ace2) accelerates diabetic kidney injury. Am J Pathol 2007; 171: 438–451. Crackower MA, Sarao R, Oudit GY et al. Angiotensin-converting enzyme 2 is an essential regulator of heart function. Nature 2002; 417: 822–828. Zimpelmann J, Kumar D, Levine DZ et al. Early diabetes mellitus stimulates proximal tubule renin mRNA expression in the rat. Kidney Int 2000; 58: 2320–2330.
1615
original article
16.
17.
18. 19.
Huang W, Gallois Y, Bouby N et al. Genetically increased angiotensin I-converting enzyme level and renal complications in the diabetic mouse. Proc Natl Acad Sci USA 2001; 98: 13330–13334. Tikellis C, Johnston CI, Forbes JM et al. Characterization of renal angiotensin-converting enzyme 2 in diabetic nephropathy. Hypertension 2003; 41: 392–397. Wysocki J, Ye M, Soler MJ et al. ACE and ACE2 activity in diabetic mice. Diabetes 2006; 55: 2132–2139. Ye M, Wysocki J, Naaz P et al. Increased ACE2 and decreased ACE protein in renal tubules from diabetic mice: a renoprotective combination? Hypertension 2004; 43: 1120–1125.
1616
HN Reich et al.: ACE2 and diabetic nephropathy
20.
21. 22.
23.
Ferrario CM, Jessup J, Gallagher PE et al. Effects of renin-angiotensin system blockade on renal angiotensin-(1–7) forming enzymers and receptors. Kidney Int 2005; 68: 2189–2196. Zimmerman D, Donnelly S, Miller J et al. Gender disparity in living renal transplant donation. Am J Kidney Dis 2000; 36: 534–540. Levey AS, Bosch JP, Lewis JB et al. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 1999; 130: 461–470. Levey AS, Coresh J, Greene T et al. Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate. Ann Intern Med 2006; 145: 247–254.
Kidney International (2008) 74, 1610–1616
http://www.kidney-international.org & 2008 International Society of Nephrology
Decreased glomerular and tubular expression of ACE2 in patients with type 2 diabetes and kidney disease Heather N. Reich1, Gavin Y. Oudit2, Josef M. Penninger3, James W. Scholey1 and Andrew M. Herzenberg4 1
Division of Nephrology, Department of Medicine, University Health Network, University of Toronto, Toronto, Ontario, Canada; 2Division of Cardiology, Department of Medicine, University Health Network, University of Toronto, Toronto, Ontario, Canada; 3Institute for Molecular Biotechnology of the Austrian Academy of Sciences, Austria and 4Department of Laboratory Medicine and Pathology, University Health Network, University of Toronto, Toronto, Ontario, Canada
Angiotensin converting enzyme (ACE) generates angiotensin II from angiotensin I, which plays a critical role in the pathophysiology of diabetic nephropathy. However, ACE2 generates angiotensin 1–7, which may protect the kidney by attenuating the effects of angiotensin II, since deletion of the Ace2 gene leads to glomerulosclerosis in mice, and pharmacologic inhibition of ACE2 exacerbates experimental diabetic nephropathy. We measured ACE2 and ACE expression in renal biopsies of patients with kidney disease due to type 2 diabetes to determine if the expression pattern is specific to diabetic nephropathy. ACE2 and ACE mRNA levels were measured by real-time PCR in laser microdissected renal biopsies from 13 diabetic and 8 control patients. ACE2 mRNA was significantly reduced by more than half in both the glomeruli and proximal tubules of the diabetic patients compared to controls, but ACE mRNA was increased in both compartments. There was a significant parallel decrease in ACE2 protein expression, determined by immunohistochemistry, in proximal tubules, a pattern not found in 12 patients with focal glomerulosclerosis or 10 patients with chronic allograft nephropathy. Our results suggest that the kidney disease of patients with type 2 diabetes is associated with a reduction in ACE2 gene and protein expression and this may contribute to the progression of renal injury. Kidney International (2008) 74, 1610–1616; doi:10.1038/ki.2008.497; published online 1 October 2008 KEYWORDS: renin-angiotensin system; diabetes mellitus; glomerulosclerosis; real-time RT-PCR; immunohistochemistry; ACE2
Correspondence: Andrew M. Herzenberg, Department of Pathology, University Health Network, 11th floor, 200 Elizabeth Street, Toronto, Ontario, Canada M5G 2C4. E-mail: [email protected] Received 6 July 2008; revised 17 July 2008; accepted 29 July 2008; published online 1 October 2008 1610
The renin-angiotensin system (RAS) is important in the pathophysiology of many progressive renal diseases including diabetic nephropathy (DN), and blockade of the RAS system with either angiotensin-converting enzyme (ACE) inhibition or Ang II receptor blockade attenuates progression of glomerular disease including DN.1–3 Angiotensin-converting enzyme-2 (ACE2) is a recently discovered homologue of ACE that regulates activity of the RAS by reducing Ang II levels and increasing the generation of Ang 1–7, and therefore expression of ACE2 in the kidney may be an important determinant of injury.4 Microdissection studies of the rat kidney have shown that ACE2 mRNA levels can be detected in almost every segment of the nephron including the glomerulus and the proximal tubule (but not the medullary thick ascending limb).5 ACE2 is located along the apical surface of proximal tubules,6 and in the glomerulus ACE2 is expressed in podocytes and mesangial cells.7,8 Recent studies have shown that deletion of the Ace2 gene leads to the development of de novo glomerulosclerosis in mice and pharmacologic inhibition of ACE2 worsens experimental DN,7,9,10 but there are limited data on the expression of ACE2 in the kidneys of normal human subjects and subjects with DN.8,11 In a recent report, Lely et al.8 studied ACE2 protein expression by immunohistochemistry in the kidneys of patients with a variety of glomerular and non-glomerular diseases. They observed neo-expression of ACE2 protein in glomerular and peritubular capillary endothelial cells, but DN constituted only a small minority of the study patients, and disease-specific information on ACE2 expression was not reported.8 Konoshita et al.11 measured ACE2 mRNA levels in whole-kidney biopsy cores from subjects with diabetic and nondiabetic kidney disease. They found that ACE2 mRNA levels were similar in the two groups but no control subjects were included in the study. Accordingly, the goal of the current study was to determine if DN is associated with glomerular and tubular changes in ACE2 gene and protein expression in subjects with type 2 diabetes mellitus (DM) compared to control subjects with normal kidney function. To test this hypothesis, we Kidney International (2008) 74, 1610–1616
original article
HN Reich et al.: ACE2 and diabetic nephropathy
RESULTS Patient and biopsy characteristics
The clinical parameters of study subjects are provided in Table 1. Subjects with DN (n ¼ 13) tended to be older, and there was a higher ratio of male subjects in comparison with control subjects (n ¼ 8). As expected, subjects with DN had higher hemoglobin A1C levels, and impaired renal function. Pathologic review confirmed that all of the biopsies from subjects with diabetes showed diffuse or diffuse and nodular diabetic glomerulosclerosis with marked mesangial matrix expansion, diffuse glomerular basement membrane thickening, and mild to moderate interstitial fibrosis. All of the kidney biopsies from the potential living donors were normal at the light microscopic level but five subjects had evidence of thin glomerular basement membranes by electron microscopy. Subjects did not have uncontrolled hypertension at the time of biopsy. None of the control subjects had a history of hypertension (data available for 6 of 8 subjects), and all of the diabetic subjects had a history of hypertension (data available for 7 of 13 subjects).
the glomerular compartment (Po0.05) of biopsies from patients with DN vs controls (Figure 1a). The abundance of ACE2 mRNA was generally lower in the glomerular compartment than in the tubular compartment in both groups. As illustrated in Figure 1b, the expression of ACE was twice as high in both the tubular and glomerular compartment of subjects with DN vs control (Po0.05). Effect of ACE inhibitor or angiotensin receptor blocker treatment on ACE2 expression in the diabetic kidney
Seven of the thirteen diabetic subjects were treated with ACE inhibitors or angiotensin receptor blocker at the time of the renal biopsy, and therefore kidney ACE2 expression was compared in the treated and untreated diabetic subjects (Table 2). Median values for creatinine and proteinuria were similar in the two groups of diabetic subjects and there were no differences in glomerular or tubular expression of ACE2. Fibrosis scores and gene expression
ACE2 and ACE expression were related to semiquantitative scores of segmental glomerulosclerosis, global glomerulosclerosis, and tubulointerstitial fibrosis in DM biopsies. We did not find a relationship between gene expression and these scores (data not shown). Immunohistochemistry
Immunohistochemical analysis of ACE2 protein expression was assessed in the same renal biopsies and mirrored the 1.2
ACE2 and ACE gene expression
Table 1 | Clinical parameters for the control and diabetic subjects Control (n=8) Age (years) M:F HbA1C (%)a Median SCr (mmol/l)a Median eGFR (ml/min/1.73 M2)a Median proteinuria (g/24 h)a % treated with ACEI (n)
44±3 1:7 5.1±0.1 78 (52–115) 87 (46–122) 0.125 (0.1–1) 13 (1)
2.12 (0.19–12)* 54 (7)*
ACEI, angiotensin-converting enzyme inhibitor; eGFR, estimated glomerular filtration rate based on the MDRD equation; F, female; HbA1C, hemoglobin A1C; M, male; proteinuria, 24-h urine protein excretion; SCr, serum creatinine. Data are presented as mean±s.e.m. or median and range. * Po0.05 vs the control group as indicated; #P=0.06 vs the control group. a Based on available data.
Kidney International (2008) 74, 1610–1616
Controls DN
0.6
*
3
Controls DN 0.3
Diabetic nephropathy (n=13) 50±3 6:7 9.2±0.7* 234 (45–530)* 29 (8–145)#
*
0.9
ACE2 mRNA (AU)
Laser capture microdissection of glomeruli and tubules was performed on frozen sections. There was a 58% decrease in the expression of ACE2 in the proximal tubules of subjects with DN vs controls as quantified by real-time PCR (Po0.05), and a 56% decrease in ACE2 gene expression in
6
ACE mRNA (AU)
measured ACE2 mRNA levels with real-time PCR in lasercaptured glomeruli and proximal tubules from kidney biopsy samples of control subjects and subjects with type 2 DM and DN, and we assessed ACE2 protein expression in the kidney with immunohistochemistry and computer-assisted image analysis. To determine if the pattern of ACE2 expression is specific to diabetes, we also compared ACE2 mRNA and protein levels in microdissected kidney biopsies of subjects with nondiabetic kidney disease: primary focal segmental glomerulosclerosis (FSGS) and chronic allograft nephropathy (CAN).
* * 0
0 Proximal Glomerular tubular compartment compartment
Proximal Glomerular tubular compartment compartment
Figure 1 | ACE2 and ACE mRNA expression in human diabetic nephropathy (DN). Real-time RT-PCR from glomerular and proximal tubular compartments, following laser capture microdissection, from kidney biopsies of human subjects with DN (n ¼ 13) and controls (n ¼ 8). (a) ACE2 mRNA expression. (b) ACE mRNA expression. All real-time PCR data are expressed in arbitrary units (AUs) corrected for GAPDH expression. *Po0.05 vs nondiabetic controls. 1611
original article
HN Reich et al.: ACE2 and diabetic nephropathy
Table 2 | The effect of ACE inhibitor or angiotensin receptor blocker treatment on ACE2 expression in diabetic subjects a
Median SCr (mmol/l) Median eGFR (ml/min/1.73 M2)a Median proteinuria (g/24 h)a Glomerular ACE2, mRNA (AU) Tubular ACE2, mRNA (AU)
No ACEI (n=6)
ACEI (n=7)
220 (45–274) 30 (17–135) 2.18 (0.19–3.2) 0.070±0.040 0.52±0.22
321 (56–530) 18 (8–145) 2.17 (1.36–12) 0.075±0.060 0.32±0.37
ACEI, angiotensin-converting enzyme inhibitor; AU, arbitrary unit; eGFR, estimated glomerular filtration rate based on the MDRD equation; SCr, serum creatinine. AU corrected for GAPDH expression. Data are presented as mean±s.e.m. or median and range as indicated. There were no significant differences between the two groups of diabetic subjects. a Based on available data; proteinuria (24-h urine protein excretion).
Table 3 | Clinical parameters for the control, FSGS, and CAN subjects
Age (years) M:F Median S*Cr Median eGFR* Median proteinuria* % Rx ACEI (n)*
Control (n=10)
FSGS (n=12)
CAN (n=10)
43.6±2.6 2:8 69 (53–97) 93.5 (56–124) n/a 0
42.0±5.2 7:5 131 (45–273) 49.5 (22–132) 3.65 (0.8–10.4) 45.5 (5)
50.3±4.3 6:4 199 (122–270) 30 (18–50) o0.30 (o0.3–1) 20 (2)
CAN, chronic allograft nephropathy; eGFR, estimated glomerular filtration rate based on the MDRD equation; F, female; FSGS, focal segmental glomerulosclerosis; M, male; proteinuria, 24-h urine protein excretion; Rx ACEI, treated with an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker treated; SCr, serum creatinine. Data are presented as mean±s.e.m. or median and range as indicated. *Po0.05 across groups.
Control
DN
60 Control DN
40
*
20
Control 1
FSGS CAN
0.5 0
0 Proximal Glomerular tubular compartment compartment
Figure 2 | Immunohistochemical staining for ACE2 protein in renal biopsies in diabetic nephropathy (DN) and controls. (a) A representative study of patient with DN and a normal live human kidney donor, and (b) quantitative computer image analysis of ACE2 immunohistochemistry staining intensity of diabetic humans (n ¼ 13) and controls (n ¼ 8) (*Po0.05 vs nondiabetic controls).
measures of mRNA expression (Figure 2a). Image analysis confirmed a 60% reduction in immunoperoxidase staining for ACE2 protein in the proximal tubules of the diabetic kidney biopsies compared with the nondiabetic control biopsies (Figure 2b). There was no significant relationship between the tubular expression of ACE2 protein and the level of proteinuria in the diabetic subjects by linear regression analysis. The ACE2 protein levels were lower than the limit of immunohistochemical detection in the glomerular compartment in both groups. ACE2 and ACE expression in FSGS and CAN
To determine if these changes in ACE2 and ACE expression are unique to DN, gene expression was also measured in the biopsies of subjects with FSGS and CAN. The clinical characteristics of these subjects are provided in Table 3. There was a significant difference in estimated glomerular filtration rate (eGFR), serum creatinine, proteinuria, and ACE inhibitor/angiotensin receptor blocker use across groups (Po0.05). The results of measurement of ACE2 and ACE mRNA expression are shown in Figure 3. There was no significant difference in glomerular or tubular ACE2, or 1612
1.5
Proximal Glomerular tubular compartment compartment
ACE mRNA (AU)
2
80 ACE2 mRNA (AU)
Immunohistochemistry score for ACE2 expression
100
10 9 8 7 6 5 4 3 2 1 0
Control FSGS CAN
Proximal Glomerular tubular compartment compartment
Figure 3 | ACE2 and ACE mRNA expression in focal segmental glomerulosclerosis (FSGS), chronic allograft nephropathy (CAN), and control biopsies. (a) ACE2 mRNA expression. (b) ACE mRNA expression. Real-time RT-PCR from proximal tubular and glomerular compartments, following laser capture microdissection from patients with FSGS (n ¼ 12), CAN (n ¼ 10), and controls (n ¼ 10). Results are expressed as arbitrary units (AUs) corrected for GAPDH. There were no significant differences across or between the three groups.
glomerular ACE expression, in FSGS or CAN compared with control biopsies (Figure 3). There was a trend toward increased tubular ACE expression in FSGS but this did not reach statistical significance. Similarly, there was no significant difference in ACE2 protein expression as measured by immunohistochemistry in either the tubular or glomerular compartment (Figure 4). When subjects with either FSGS or CAN were divided according to ACE inhibitor/ angiotensin receptor blocker use, there was an increase in the tubular expression of ACE2 noted in subjects taking these medications (2.15±1.34, n ¼ 7 users vs 0.67±0.82, n ¼ 14 nonusers, P ¼ 0.03). There was no difference in glomerular ACE2, tubular or glomerular ACE, when subjects were analyzed according to medication use. DISCUSSION
In this study we observed decreased ACE2 mRNA levels and protein expression in kidneys of human subjects with DN compared to nondiabetic healthy control subjects. The decrease in ACE2 expression occurred in both glomeruli and proximal tubule cells and the effect of DM on mRNA Kidney International (2008) 74, 1610–1616
original article
HN Reich et al.: ACE2 and diabetic nephropathy
Immunohistochemistry score for ACE2 expression
100
80
60
20
0 Control
FSGS
CAN
Control FSGS CAN
40
Proximal Glomerular tubular compartment compartment
Figure 4 | Immunohistochemical staining for ACE2 protein in renal biopsies in focal segmental glomerulosclerosis (FSGS), chronic allograft nephropathy (CAN), and controls. (a) Representative study patients with idiopathic FSGS, CAN, and a normal live kidney donor. (b) Quantitative computer image analysis of ACE2 immunohistochemistry staining intensity in biopsies of patients with FSGS (n ¼ 12), CAN (n ¼ 10), and in controls (n ¼ 10). There were no significant differences across or between the three groups.
Table 4 | Summary of available literature describing ACE2 expression in glomeruli and renal tubules in diabetic nephropathy Glomerulus
Proximal tubules/cortex
ACE2 mRNA
ACE2 protein
ACE2 mRNA
ACE2 protein
ACE2 activity
Human Reich et al. Lely et al.8 Konoshita et al.11
k NA NA
BD m NA
k NA 2
k 2 NA
NA NA NA
Animal models Tikellis et al.17 Streptozotocin rats (type 1 DM) Ye et al.7,19; Wyscocki et al.18db/db mice (type 2 DM) Wysocki et al.18 Streptozotocin mice (type 1 DM) Wong et al.13 Akita mice (type 1 DM)
m NA NA NA
m k NA NA
k 2 2 m
k m m NA
NA m m NA
k, decreased ACE2 expression; m, increased ACE2 expression; 2, no significant difference in expression; BD, below detection limit; DM, diabetes mellitus; NA, not analyzed. The results of the study reported herein and other human and rodent studies of renal ACE2 expression in diabetic nephropathy are summarized. Studies reporting mRNA expression of whole cortex renal tissue were considered to represent the proximal tubular compartment for the purposes of this summary.
levels and protein levels was similar in both of these compartments. ACE2 is a monocarboxypeptidase that cleaves a single amino acid from the C terminus of Ang I and Ang II, which leads to the generation of Ang 1–9 and Ang 1–7, respectively. Ang 1–7 has vasodilatory and antiproliferative effects, and it antagonizes Ang-II-mediated cell signaling including transforming growth factor-b 1 expression.12 Together with the reports of gene deletion studies and pharmacologic ACE2 blockade,7,9,10,13,14 our findings on ACE2 expression in the diabetic kidney are consistent with the notion that ACE2 serves a protective role in the kidney, and support the hypothesis that reduced ACE2 expression (and activity) may contribute to the development and progression of kidney injury, including DN. However, the data do not establish a causal relationship between ACE2 expression and DN nor the mechanism(s) responsible for the decrease in ACE2. In contrast to the change in ACE2 expression, ACE expression increased in both the glomerular and proximal tubule cells of the diabetic kidneys compared to the control kidneys, in accord with previous observations by Kidney International (2008) 74, 1610–1616
Zimpelmann et al.15 in the kidneys of streptozotocin-induced diabetic rats, and the observations of Ye et al.7 in the glomeruli of diabetic db/db mice. It is possible that these relative changes might lead to a shift in angiotensin peptide processing toward generation of Ang II and away from the generation of Ang 1–7 in the diabetic kidney, an effect that is predicted to be deleterious. In support of this hypothesis, Huang et al.16 reported that there was a relationship between the development of DN and levels of ACE expression in transgenic mice. We could not measure angiotensin peptides in our kidney biopsy samples so the effect of the changes in ACE2 and ACE expression on angiotensin peptide processing remains speculative. Table 4 summarizes the current literature regarding ACE2 expression and activity in experimental and clinical DN. Tikellis et al.17 were the first to study ACE2 expression in rats with streptozotocin-induced diabetes, a model of type 1 DM. They reported that ACE2 mRNA and protein levels were reduced by 50% in the renal tubules of diabetic rats, similar to our findings, and they also observed that ACE2 expression was mainly in the tubular compartment. In contrast to our 1613
original article
results in human kidney biopsies, DM was associated with an increase in glomerular ACE2 expression. Interestingly, Wysocki et al.18 studied ACE2 expression and activity in the kidney cortex of mice with streptozotocin-induced DM and they reported that ACE2 protein expression and ACE2 activity were increased compared to nondiabetic mice. Similar findings were reported in the db/db mouse, a model of type 2 DM.18,19 Recently, the same group reported that ACE2 immunostaining was reduced in the glomeruli of the db/db mice compared to nondiabetic db/m mice.7 Taken together, it is difficult to reconcile these observations on ACE2 expression in experimental DM, and the effect of diabetes may be specific for each model and dependent on the stage of injury. There are limited but incomplete data on ACE2 expression in human DN. A study comparing ACE2 mRNA levels in the biopsies of patients with diabetic and nondiabetic kidney disease demonstrated no differences in ACE2 expression across a wide range of disease categories.11 However, diabetic subjects consisted only a small subset of the study subjects (8 of 74 subjects), gene expression was not compared with that of a normal control group, and the mRNA levels were not correlated with protein expression. Lely et al.8 reported that ACE2 protein immunostaining was uniformly increased in the glomeruli and peritubular capillaries in kidney biopsies representing a variety of kidney diseases, including some subjects with DN, however this was assessed semiquantitatively, and mRNA levels, which may be more sensitive given the relatively low abundance of the protein, were not assessed. Furthermore, the relative difference in ACE2 expression between DN and normal kidney biopsies was not a specific focus of this study.8 The effect of blockade of the RAS on ACE2 expression in the human kidney has not been well studied although Tikellis et al.17 have examined the effect of ACE inhibitor treatment on ACE2 expression in the kidneys of rats with streptozotocin-induced DM. They reported that ACE2 mRNA levels were low in the diabetic kidney and unaffected by treatment with ramipril, in accord with our observations in the kidneys of subjects with type 2 DM. They did report that ACE2 protein levels were markedly increased in association with ramipril treatment (in contrast to the effect on mRNA levels) in diabetic rats, but we did not observe any discordance between measures of ACE2 mRNA expression and protein expression in our kidney biopsy samples. Ferrario et al.20 reported studies of the effect of ACE inhibition, angiotensin receptor blockade, and combination therapy on ACE2 mRNA levels in the kidneys of normal rats. ACE2 mRNA levels were not affected by any of the treatments. To determine if the pattern of ACE2 and ACE expression that we observed in the kidneys of subjects with DN was specific for DN or might represent a generalized response to kidney injury, we studied two separate groups of subjects with kidney disease: FSGS and CAN. We chose FSGS because like DN, it is also a glomerular disease marked by proteinuria; we chose CAN as it represents a chronic progressive kidney 1614
HN Reich et al.: ACE2 and diabetic nephropathy
disease that is not glomerular in origin, and it is characterized by prominent tubulointerstitial injury. We found that ACE2 and ACE mRNA levels were not different in the kidney biopsies of subjects with FSGS or CAN compared to controls, suggesting that decline in ACE2 expression is not a general feature of kidney injury and that there is some specificity of this response for DN. Interestingly, blockade of the RAS (with ACE inhibitor/angiotensin receptor blocker) in subjects with either FSGS or CAN was associated with an increase in the tubular expression of ACE2, and it is tempting to speculate that ACE inhibitor/angiotensin receptor blocker use may alter ACE2 levels in subjects without diabetes. The strengths of the current study are that we compared ACE2 and ACE expression in the kidney of subjects with type 2 DM with the kidneys of healthy control subjects, and we localized expression at the mRNA level by performing laser capture microscopy of the glomeruli and proximal tubule cells. We also related ACE2 mRNA levels with protein expression by immunohistochemistry and image analysis. However, this report is descriptive, and we did not correlate measures of mRNA and protein expression levels with ACE2 activity in the kidney or with measures of angiotensin peptides. In addition, gender and hypertension may be confounding variables in our analysis. Women were overrepresented in our control group, largely because of the fact that our control subjects were potential living kidney donors, which tend to be women at our center.21 A history of hypertension was prominent in our diabetic subjects but we could not correct our analysis for blood pressure levels. In conclusion, ACE2 mRNA levels and protein expression decrease in the proximal tubules of patients with DN compared with nondiabetic healthy control subjects, and a similar trend is observed in the glomeruli. ACE2 expression is relatively low in human glomeruli compared to the proximal tubule cells in both normal and diabetic kidneys, and treatment with an ACE inhibitor does not affect ACE2 expression in the diabetic kidney. Finally, the pattern of decreased ACE2 and increased ACE expression was not observed in the biopsies of subjects with FSGS or CAN. In subjects with diabetes, the combination of effects of altered ACE2 and ACE expression may contribute to the progression of DN. MATERIALS AND METHODS Renal biopsies
Renal biopsies had been performed as part of routine clinical diagnostic investigation. Informed consent was obtained from patients for use of archived human biopsy material and chart review in accord with institutional research ethics board review. The diabetes samples were obtained from patients with documented type 2 diabetes and a pathologic diagnosis of DN. Control tissues were obtained from healthy potential live transplant donors who were referred for biopsy because of microscopic hematuria, who had normal renal function and no proteinuria. Control renal biopsies were either normal or showed thin basement membranes accordKidney International (2008) 74, 1610–1616
original article
HN Reich et al.: ACE2 and diabetic nephropathy
ing to pathologist review. Routine immunofluorescence microscopy of these biopsy samples showed negative staining for immunoglobulins and complement. Laser capture microdissection and real-time PCR
Snap-frozen renal biopsies stored at 80 1C were cut at 8 mm and stained with HistoGene (Arcturus, Mountainview, CA, USA) and laser microdissected using the Pixcell IIe (Arcturus). Glomerular profiles and kidney proximal tubules were identified by an experienced renal pathologist (AMH), separately dissected, and collected for mRNA extraction. The mRNA was extracted using the PicoPure kit (Arcturus) and first-strand cDNA synthesis was performed using the Sensiscript RT Kit (Qiagen, Valencia, CA, USA). The cDNA was used undiluted for TaqMan real-time PCR on an Applied Biosystems 7900 machine. The ACE2 gene expression was expressed as a ratio of ACE2/GAPDH (glyceraldehyde-3phosphate dehydrogenase) housekeeping gene. There was no significant difference in GAPDH levels in disease samples compared to control samples, and we have found that GAPDH is a more consistent reference mRNA than 18S or bactin. Primers and probe for amplification were as follows— ACE2: CATTGGAGCAAGTGTTGGATCTT (forward), GAG CTAATGCATGCCATTCTCA (reverse), CTTGCAACACCAG TTCCCAGGCA (probe); ACE: CCGAAATACGTGGAACT CATCAA (forward), CACGAGTCCCCTGCATCTACA (reverse), CAGGCTGCCCGGCTCAATGG (probe). GAPDH primer sets were designed as follows: GAAGGTGAAGGTCGGAGTC (forward), GAAGATGGTGATGGGATTTC (reverse), and CAA GCTTCCCGTTCTCAGCC (probe). In addition, GAPDH primers were also purchased from Applied Biosystems.
Wilcoxon) analyses, and similar results were obtained using both approaches. Linear regression was used to relate the ACE2 expression to the level of proteinuria and fibrosis in the diabetic subjects. All statistical tests were performed using SAS (SAS Institute, Cary, NC, USA), and two-tailed P-values are provided. DISCLOSURE
All the authors declared no competing interests. ACKNOWLEDGMENTS
Andrew Herzenberg and James Scholey are supported by a Canadian Institutes of Health Research (CIHR) New Emerging Team (NET) grant (Genes, Gender and Glomerular-based Diseases NET). Andrew Herzenberg, James Scholey, and Gavin Oudit are the recipients of a Canadian Diabetes Association (CDA) grant. Heather Reich is the recipient of a KRESCENT Trainee Fellowship Award (CIHR, Kidney Foundation of Canada, and Canadian Society of Nephrology). James Scholey is the recipient of a CIHR-AMGEN Canada Research Chair in Nephrology. Gavin Oudit is a clinician-scientist supported by the Heart and Stroke Foundation of Canada, Canadian Institute for Health Research, and the TACTICS program. Josef Penninger is supported by a Eugene Heart grant. We acknowledge the expert technical assistance of Brent Steer and Dr Phil Marsden as well as Stuart Yang. REFERENCES 1.
2.
3.
4.
Immunohistochemistry
Formalin-fixed paraffin-embedded sections from the same renal biopsy were cut at 3 mm followed by heat-induced antigen retrieval. Sections were incubated with monoclonal anti-ACE2 (1:400) antibodies.14 To eliminate any potential nonspecific biotin activity, slides were stained with a secondary anti-mouse antibody using the EnVision system (Dako). Endogenous peroxidase activity was prevented by pretreating with 3% hydrogen peroxide. Image-Pro Plus (Media Cybernetics) computer image analysis software was used to analyze brown staining pixel density and quantify protein levels. Negative controls with irrelevant primary antibody and no primary antibody were performed. Calculations and statistics
The eGFR was calculated using the abbreviated Modification of Diet in Renal Disease equation.22,23 The distribution of clinical data was visualized and summary data are expressed as mean±standard error or median and range as indicated. Differences in means were analyzed using nonparametric tests when data were not normally distributed. Gene and protein expression data were analyzed using both parametric (for example, analysis of variance, with Levene’s test of the assumption of equal variance) and nonparametric (that is, Kidney International (2008) 74, 1610–1616
5.
6.
7.
8. 9.
10.
11.
12.
13.
14. 15.
Lewis EJ, Hunsicker LG, Bain RP, et al., for the Collaborative Study Group. The effect of angiotensin-converting enzyme inhibition on diabetic nephropathy. N Engl J Med 1993; 329: 1456–1462. Lewis EJ, Hunsicker LG, Clarke WR, et al., for the Collaborative Study Group. Renoprotective effect of the angiotensin-receptor antagonist irbesartan in patients with nephropathy due to type 2 diabetes. N Engl J Med 2001; 345: 851–860. Brenner BM, Cooper ME, de Zeeuw D, et al., RENAAL Study Investigators. Effects of losartan on renal and cardiovascular outcomes in patients with type 2 diabetes and nephropathy. N Engl J Med 2001; 345: 861–869. Burrell LM, Johnston CI, Tikellis C et al. ACE2, a new regulator of the reninangiotensin system. Trends Endocrinol Metab 2004; 15: 166–169. Li N, Zimpelmann J, Cheng K et al. The role of angiotensin converting enzyme 2 in the generation of angiotensin 1–7 by rat proximal tubules. Am J Physiol Renal Physiol 2005; 2: 353–362. Warner FJ, Lew RA, Smit AI et al. Angiotensin-converting enzyme 2 (ACE2), but not ACE, is preferentially localized to the apical surface of polarized kidney cells. J Biol Chem 2005; 280: 39353–39361. Ye M, Wysocki J, William J et al. Glomerular localization and expression of angiotensin-converting enzyme 2 and angiotensin converting enzyme: implications for albuminuria in diabetes. J Am Soc Nephrol 2006; 17: 3067–3075. Lely AT, Hamming I, van Goor H et al. Renal ACE2 expression in human kidney disease. J Pathol 2004; 204: 587–593. Oudit GY, Herzenberg AM, Kassiri Z et al. Loss of angiotensin-converting enzyme 2 leads to the late development of angiotensin II-dependent glomerulosclerosis. Am J Pathol 2006; 168: 1808–1820. Soler MJ, Wysocki J, Ye M et al. ACE2 inhibition worsens glomerular injury associated with increased ACE expression in streptozotocin-induced diabetic mice. Kidney Int 2007; 72: 614–623. Konoshita T, Wakahara S, Mizuno S et al. Tissue gene expression of reninangiotensin system in human type 2 diabetic nephropathy. Diabetes Care 2006; 29: 848–852. Su Z, Zimpelmann J, Burns KD. Angiotensin-(1-7) inhibits angiotensin IIstimulated phosphorylation of MAP kinases in proximal tubular cells. Kidney Int 2006; 69: 2212–2218. Wong DW, Oudit GY, Reich H et al. Loss of angiotensin converting enzyme-2 (Ace2) accelerates diabetic kidney injury. Am J Pathol 2007; 171: 438–451. Crackower MA, Sarao R, Oudit GY et al. Angiotensin-converting enzyme 2 is an essential regulator of heart function. Nature 2002; 417: 822–828. Zimpelmann J, Kumar D, Levine DZ et al. Early diabetes mellitus stimulates proximal tubule renin mRNA expression in the rat. Kidney Int 2000; 58: 2320–2330.
1615
original article
16.
17.
18. 19.
Huang W, Gallois Y, Bouby N et al. Genetically increased angiotensin I-converting enzyme level and renal complications in the diabetic mouse. Proc Natl Acad Sci USA 2001; 98: 13330–13334. Tikellis C, Johnston CI, Forbes JM et al. Characterization of renal angiotensin-converting enzyme 2 in diabetic nephropathy. Hypertension 2003; 41: 392–397. Wysocki J, Ye M, Soler MJ et al. ACE and ACE2 activity in diabetic mice. Diabetes 2006; 55: 2132–2139. Ye M, Wysocki J, Naaz P et al. Increased ACE2 and decreased ACE protein in renal tubules from diabetic mice: a renoprotective combination? Hypertension 2004; 43: 1120–1125.
1616
HN Reich et al.: ACE2 and diabetic nephropathy
20.
21. 22.
23.
Ferrario CM, Jessup J, Gallagher PE et al. Effects of renin-angiotensin system blockade on renal angiotensin-(1–7) forming enzymers and receptors. Kidney Int 2005; 68: 2189–2196. Zimmerman D, Donnelly S, Miller J et al. Gender disparity in living renal transplant donation. Am J Kidney Dis 2000; 36: 534–540. Levey AS, Bosch JP, Lewis JB et al. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 1999; 130: 461–470. Levey AS, Coresh J, Greene T et al. Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate. Ann Intern Med 2006; 145: 247–254.
Kidney International (2008) 74, 1610–1616