|
Rat Ace allele variation determines susceptibility to AngII-induced renal damage
Abstract
Introduction: Ace b/l polymorphism in rats is associated with differential tissue angiotensin-converting enzyme (ACE) expression and activity, and susceptibility to renal damage. Same polymorphism was recently found in outbred Wistar rat strain with b allele accounting for higher renal ACE, and provided a model for studying renin-angiotensin-aldosterone system (RAAS) response behind the innate high or low ACE conditions. Methods: We investigated the reaction of these alleles on chronic angiotensin II (AngII) infusion. Wistar rats were selected to breed male homozygotes for the b (WU-B) or l allele (WU-L) (n = 12). For each allele, one group (n = 6) received AngII infusion via an osmotic minipump (435 ng/kg/min) for 3 weeks. The other group (n = 6) served as a control. Results: WU-B had higher ACE activity at baseline then WU-L. Interestingly, baseline renal ACE2 expression and activity were higher in WU-L. AngII infusion induced the same increase in blood pressure in both genotypes, no proteinuria, but caused tubulo-interstitial renal damage with increased α-SMA and monocyte/macrophage influx only in WU-B (p < 0.05). Low ACE WU-L rats did not develop renal damage. Conclusion: AngII infusion causes proteinuria-independent renal damage only in rats with genetically predetermined high ACE while rats with low ACE seemed to be protected against the detrimental effect of AngII. Differences in renal ACE2, mirroring those in ACE, might be involved.
The ACE gene encodes for the angiotensin-converting enzyme (ACE), a principal catalytic protease in the reninangiotensin-aldosterone system (RAAS). In humans, the functional ACE I/D polymorphism accounts for half of the variance in plasma and renal ACE. 1 A similar polymorphism resulting in differential tissue Ace expression has been identified in the rat. 2 This Ace polymorphism was recently reported also within the outbred Wistar rat strain. Wistar rats homozygous for the b allele (WU-B) had higher renal ACE expression when compared with Wistar rats homozygous to the l allele (WU-L), providing a model for studying the RAAS under high or low ACE conditions and against the background of similar genetic heterogeneity and allele frequencies. 3 Increased renal ACE, in both rodents and human, has been thoroughly studied in connection with the progression of renal and cardiovascular disease. [4] [5] [6] [7] [8] [9] In rats, high renal ACE has been associated with the increased susceptibility to hypertensive end organ renal damage. 10 In the model of chronic renal transplant failure, rats with high levels of intrarenal ACE mRNA and enzyme activity tended to develop more easily intraglomerular hypertension and more renal damage. 11 Furthermore, after adriamycininduced nephrosis, the differences in baseline renal ACE activity predicted the severity of damage, 12 while during nephrosis development, a positive correlation existed between renal ACE activity and proteinuria and glomerular and interstitial injury. 5 In addition, genetically predetermined high ACE has been suggested as a genetic risk marker in renal disease. 13, 14 The data on the whole are still conflicting and the exact role of predetermined high ACE in susceptibility to renal damage is still inconclusive. However, the response of the RAAS to various stimuli is probably affected by this ACE polymorphism.
Concurrently, with the expanding complexity of RAAS, ACE2 has emerged as a potential functional counterpart of ACE. The current premise is that the two enzymes stand in balance. [15] [16] [17] [18] It was shown that ACE2 is able to catalyze AngII with high efficiency, 19 hence preventing its accumulation. 20 Furthermore, increasing data provide evidence of reno-and cardio-protective properties of the ACE2 axis. [21] [22] [23] Activation of ACE2 axis was reported to lead to the decrease of oxidative stress and cell damage, 22 stimulates vasodilatation 24 and it was recently shown that supplementation of ACE2 attenuated diabetic kidney injury in mice. 23 The availability of a rat model that, with different kidney tissue ACE levels hypothetically has different RAAS symmetry, might give us an insight into the dynamics underlying the b/l Ace polymorphism. Enhanced conversion of AngI has been demonstrated in experimental setups in non-diabetic and diabetic humans, 25, 26 but considering the negative feedback within the RAAS, it is uncertain whether such acute experiments bear relevance to the susceptibility to chronic renal damage. An alternative hypothesis deserves consideration. Studying RAAS, it was shown that disease development could be followed by combined increase of renal ACE and decrease of ACE2, 27, 28 suggesting that genetically higher ACE may be associated with altered dynamics within the RAAS other than AngI conversion. If such effects downstream of ACE would be involved in the association to genetically higher ACE and susceptibility to renal damage, an increased susceptibility of the kidney to AngII-induced damage would be anticipated.
To test this hypothesis, we infused WU-B and WU-L rats with AngII for a period of 3 weeks. In our model we evaluated glomerular and interstitial damage as well as Ace and Ace2 expression and activity response to an external AngII stimulation. Furthermore, we analyzed systemic and renal functional parameters (blood pressure, creatinine clearance and proteinuria).
Healthy Wistar Unilever (HsdCpb:WU) rats were obtained from Harlan (Harlan Inc, Horst, The Netherlands) and genotyped for the microsatellite marker in intron 13 of the Ace gene that identifies the presence of the b or l Ace allele. In short, genomic DNA was isolated from tail tips as described previously. 29 Primers that amplify the microsatellite located at the 5' end of the intron between exons 13 and 14 were used to determine the Ace genotypes, as described previously by Hilbert et al. 30 Animals were then selected to breed 12 males homozygous for each of the b and l alleles. All animals were housed in a climate controlled space with a 12-h light/12-h dark cycle. Food and water were available ad libitum.
At the age of 9 weeks, rats of the same genotype were randomly divided into two groups. The first group received AngII infusion via an osmotic minipump (435 ng/kg/min). The second group served as a control. After 3 weeks of treatment, rats were anesthetized using an isoflurane/O 2 / N 2 O mixture, the minipump was removed and kidneys were collected. Renal tissue from the upper and lower poles was snap-frozen in liquid nitrogen to be used for immunohistochemistry, and to measure gene expression and enzyme activity. The middle section was fixed in formalin and embedded in paraffin for immunohistochemistry. At baseline and before termination, blood was taken via orbital puncture and rats were subsequently placed in metabolic cages for 24-h urine collection. Body weight was measured weekly. Blood pressure was measured with a tail cuff method (CODA™, Kent Scientific Corporation, Torrington, CT, USA) at baseline and 2 weeks into the treatment. After five pre-measurements, the mean of best three measurements was determined for every animal.
All animal experiments were approved by the local Committee for Animal Experiments of the University Medical Center Groningen.
For RNA expression, parts of kidney poles were first homogenized in the presence of lysis buffer and β-mercaptoethanol. Kidney lysate was then used for mRNA isolation with the Nucleospin RNA kit (Macherey-Nagel), according to the manufacturer's protocol. RNA concentration was measured with a Nanodrop spectrophotometer. For cDNA synthesis, 1000 ng of template mRNA in 12 μl total dilution was used according to the QuantiTect Reverse Transcription (Qiagen) protocol for cDNA isolation. Ace and Ace2 mRNA levels were determined with reverse transcriptase polymerase chain reaction (RT-PCR) on the ABI Prism 7900 HT sequence detection system (Applied Biosystems). Master mix (Eurogentec, Liege, Belgium), Ace primers and probe or Ace2 primers, and 20 ng of cDNA to a total volume of 10 μl, were used in a reaction in triplet. For the Ace gene, a custom designed primer-probe set was used, with the primers 5'-CACCGGCAAGGTCTGCTT-3', 5'-CTTGGCATAGTTTCGTGAGGAA-3', and the probe 6-FAM 5'CAACAAGACTGCCA CCTGCTGGTCC-3'TAMRA (Biolegio, Nijmegen, The Netherlands). For the Ace2 gene, a gene-specific assay was used (Rn01416295_ m1; Applied Biosystems). Ace and Ace2 mRNA levels were expressed relative to those of the hypoxanthine phosphoribosyltransferase (Hprt) housekeeping gene, which was determined to be more stable then β2 microglobulin housekeeping gene (B2m). Data were analyzed with SDS 2.1 (SABiosciences).
For ACE and ACE2 activity, kidneys were homogenized in the RIPA lysis buffer (pH 8) with addition of 4% (v/v) EDTA-free protease inhibitor cocktail (Roche Diagnostics Nederland BV) and 0.2% (v/v) benzonase nuclease (Novagen, Merck KGaA, Darmstadt, Germany). Every sample was incubated in Assay Buffer (Tris-HCl, pH 7.4), with and without an inhibitor (ACE inhibitor: Captopril; ACE2 inhibitor: MLN-4670), for 30 min at 37°C. The fluorescent substrate was added to the reaction (ACE: Abz-FRK(Dnp)-P;
ACE2:
Mca-APK(Dnp) (Biomol International LP)). Abz or Mca fluorescence is quenched by the DnP group, until ACE or ACE2 cleaves and releases it. The fluorescence was measured using a Wallac Victor 1420 Multilabel counter (ACE: excitation 320 nm; emission 420 nm; ACE2: excitation 340 nm; emission 405 nm) at 10 minute intervals for 400 minutes. Every sample was measured in triplicate in the presence and absence of the appropriate inhibitor, under conditions in which the initial reaction rate was linear, to determine ACE-and ACE2specific activity. Activity was expressed in units of pmol product/min/μg protein.
Deparaffinized sections (4 mm) were stained by periodic acid-Schiff (PAS) on a DAKO automatic slide stainer (DAKO Group, Denmark) to evaluate renal morphology. To assess renal damage and cellular infiltration, a myofibroblast transformation marker α-smooth muscle actin (α-SMA) and a cellular marker specific for activated monocytes and macrophages (ED1) were detected. Mouse monoclonal antibodies were used (α-SMA: clone 1A4, Sigma, St. Louis, Mo, USA, dilution 1:10,000; ED1/CD68: Serotec Ltd., Oxford, UK, dilution 1:750). α-SMA staining was performed by an automated staining system (Dako Autostainer Universal Staining System, Glostrup, Denmark). Deparaffinized sections were incubated overnight in 0.1 M Tris-HCl buffer (pH 9.0) at 80°C for antigen retrieval. Endogenous peroxidase was blocked with 0.3% H 2 O 2 in phosphate buffered saline (PBS; pH 7.4) for 30 minutes followed by a 60-minute incubation of primary antibodies at room temperature. Antibody binding was detected by sequential incubation of peroxidase-labeled secondary and tertiary antibodies (Dakopatts, Glostrup, Denmark, dilution 1:100) for 30 minutes at room temperature. Peroxidase activity was developed by 3,3'-diaminobenzidine tetrahydrochloride (DAB) with 0.03% H 2 O 2 for 10 minutes. All antibody dilutions were made in 1% bovine serum albumin (BSA) in PBS with addition of 1% normal rat serum (NRS) to secondary and tertiary antibodies. Counterstaining was performed using Mayer's hematoxylin.
The degree of tubulo-interstitial fibrosis (TIF) was determined in PAS-stained sections in a blinded fashion. TIF was scored positive when tubular atrophy and fibrosis of the interstitium were present simultaneously. Semiquantitatively, damage scores of 0-5 were assigned to 20 cortical fields per slide: a score of 0 indicated no interstitial fibrosis, score of 1 indicated 0-10% involvement, score of 2 indicated 10-25%, score of 3 indicated 25-50%, a score of 4 indicated 50-70% and a score of 5 indicated 75-100% involvement of the cortical field. The final damage score per slide was calculated by multiplying the degree of damage by the number of positive fields/score ratio and adding these scores.
Interstitial ED1 positive cells were counted in 40 interstitial fields per slide excluding glomeruli and large arteries.
The percentage of α-SMA-positive staining was quantified by computerized morphometry. Glomerular α-SMA was determined on 50 glomeruli, interstitial α-SMA on 30 interstitial fields per kidney, excluding glomeruli and blood vessels. The surface area found positive was divided by the total area of the glomerulus/interstitial fields measured, providing a percentage of α-SMA-positive tissue. The average percentage of all glomeruli/interstitial field was then calculated for every group.
Urinary protein excretion was measured in 24-h urine with the Behring Nephelometer Analyzer II (BNII, Dade Behring Marburg GmbH, Marburg, Germany) by using a 20% trichloroacetic acid (TCA) solution as the reagent and tuberculin protein prepared by means of the ultrafiltration method (TPU) as a control.
Urinary and plasma creatinine were determined colorimetrically using a multianalyzer (Modular Analytics System, modules: ISE900, P800 en E170; Roche Diagnostics GmbH, Mannheim, Germany). Creatinine clearance (CrCl) was calculated per 100 g of body weight for every animal.
Data are expressed as median and inter-quartile range unless stated otherwise, and the distribution was assumed to be non-parametric. Data were analyzed using the statistical program Graph Pad, Prism, Version 5.00. Statistical differences were determined using the Mann-Whitney U-test. Statistical significance was set at p < 0.05.
Body weight is presented in Table 1 . In the course of the experiment all rats significantly gained weight except the WU-B AngII-infused group. Baseline weight in the WU-L group (before AngII infusion), was somewhat higher than in the same WU-B group. Kidney weight at termination (3 weeks) did not differ between groups.
Characterization of the model. As expected, the activity of renal ACE was higher in the WU-B than in the WU-L group. In the current study the difference in Ace mRNA expression level was smaller than shown previously in this strain, 3 but nevertheless the effect of the Ace allele difference was confirmed with the presence of two different ACE phenotypes. Interestingly, the two phenotypes also differed in baseline renal ACE2 levels. Both mRNA expression and activity of ACE2 was consistently significantly higher in the WU-L group compared with the WU-B group (Figure 1) .
AngII infusion. AngII lead to an increase of Ace mRNA in WU-L but not in WU-B when compared with their respective controls (Figure 1) . Also, the difference between genotypes in Ace expression was statistically significant (p < 0.05). On the other hand, in both genotypes excess of AngII led to an increase of ACE activity. However, the ACE activity tendency between genotypes remained the same, with higher activity in WU-B compared with WU-L (p < 0.05). As for Ace2, WU-L reacted by increasing mRNA, which was significantly higher than in WU-B (p < 0.05). Conversely, ACE2 activity was increased only in WU-B rats after AngII infusion where ACE2 activity levels reached the levels detected in WU-L rats, which showed no difference between treated and untreated groups ( Figure 1 ).
Characterization of the model. At the age of 9 weeks normal renal structural damage scores were present in both genotypes. Interstitial fibrosis score was negligible and incidence of pro-fibrotic marker α-SMA was below 5% in both genotypes. The percentage of ED1 positive cells was similar between genotypes (8.15% (7.32-9.55) in WU-B and 8.23% (6.9-10.06) in WU-L; see Figure 2 ).
AngII infusion. Three weeks of AngII infusion caused glomerular and interstitial renal damage only in WU-B rats (p < 0.05) (Figure 2) . TIF was significantly increased as was glomerular and interstitial pro-fibrotic marker α-SMA when compared with their respective controls. The number of cells expressing the monocyte/ macrophage marker ED1 was also significantly increased only in WU-B (p < 0.05). The consistent increase of these damage markers in the high ACE group indicates a pro-fibrotic and pro-inflammatory response of this genotype to high AngII. On the other hand, in the WU-L group, AngII infusion did not result in renal damage. TIF, α-SMA and monocyte/macrophage influx were not increased in WU-L after AngII infusion, but remained at the control levels.
Baseline blood pressure in all animals was 143 ± 15 mm Hg. AngII infusion increased blood pressure significantly in both genotypes (p < 0.05) (Figure 3 ). Although the WU-L control rats showed a slight increase in blood pressure compared with WU-B control rats, the effect of AngII infusion on blood pressure in both groups proved to be statistically significant. Baseline proteinuria was not different between groups with an average of 27 ± 18 mg/24 h. Three weeks of AngII infusion did not cause a significant Table 1 . Body weight at baseline and termination, and kidney weight at termination, in Control and Angiotensin II infusion (AngII) groups. In the 3-week course of the experiment all rats significantly gained weight except the WU-B group receiving AngII infusion. No difference was found in kidney weight between groups at termination. Data are expressed as median (interquartile range).
WU-L Control (n = 6) AngII (n = 6) Control (n = 6) AngII (n = 6) increase in urinary protein excretion either in WU-B or in WU-L. As shown in a separate box, one rat in the WU-B group exhibited more sensitivity to AngII, while proteinuria levels of the five other rats in the WU-B group tended to cluster more closely. This accounted for the ascending proteinuria line for WU-B, but also for the lack of statistical significance when compared with the other groups ( Figure 4(a) ). Baseline CrCl was not different between groups with an average of 1.3 ± 0.3 ml/min/100 g body weight. Three weeks of AngII infusion lead to a drop of CrCl only in the WU-B group (p < 0.05 versus baseline). At termination, in the WU-B control group CrCl was significantly higher than at baseline (p < 0.05), while in the WU-L this was lower than at baseline (p < 0.05) (Figure 4(b) ).
Three weeks of AngII infusion induces proteinuriaindependent, pro-inflammatory and pro-fibrotic renal response in rats with genetically determined high Ace expression. In our study, these WU-B rats developed glomerular and tubulo-interstitial damage with increased monocyte/macrophage influx. CrCl and weight gain were reduced in comparison with controls, indicating a decrease in renal function and welfare in WU-B. On the other hand, rats with intrinsically low Ace expression (WU-L) did not develop renal damage, hence they seemed to be protected against increased AngII levels. As expected, AngII infusion lead to a significant increase in blood pressure in both (c) Glomerular and interstitial α-SMA score and staining in the kidney. Arrows are indicating the difference in glomerular α-SMA between genotypes. Renal damage after AngII infusion was consistently present only in rats with WU-B genotype. The WU-L genotype seemed to be protected against AngII. # p < 0.05, Control versus AngII infusion, same genotype; *p < 0.05, WU-B versus WU-L, same timepoint.
genotypes but interestingly without an increase in proteinuria. An excess of AngII is involved in the development of hypertensive nephropathy and nephrosclerosis 31 that is followed by an increase of proteinuria. 32 In other studies a significant increase in protein excretion was reported only after prolonged (6 weeks) AngII infusion. 33 Conversely, the absence of proteinuria in our study allowed us to investigate glomerular and tubulo-interstitial damage without the detrimental effects of high urinary protein levels.
The (patho)physiological effects of high or low ACE levels are still inconclusive. In humans, the high ACE activity determined by the D ACE allele has been extensively studied in connection with the development and progression of renal and cardiovascular disease. While many studies show a connection, [34] [35] [36] [37] [38] others have failed to establish it in the risk and disease prediction. 39, 40 The genetic variability as well as the multifactorial nature of cardio-renal diseases might account for these discrepancies. In rats, the significant effect of background ACE genotype on baseline renal hemodynamics was shown in a study by Liu et al. 10 where renal plasma flow and GFR were lower in high ACE Ren2.F than in low ACE Ren2.L rats. They also identified the Ace gene as a modifier of hypertensive end organ damage. 10 On the other hand, a genetic linkage has been found between the Ace gene and plasma ACE activity, but not between the Ace gene and blood pressure per se, 41 so it could be argued that possible detrimental effects of high ACE could be mediated by a more tissue-oriented pathway. Physiologically, ACE can regulate vascular reactivity and response to hypertension injury by altering the conversion of AngI and hence the concentration of AngII, which is known to have The ascending proteinuria line for WU-B was due to one rat with possible high AngII sensitivity (as shown in a separate box). The difference to other groups was not statistically significant. (b) At baseline, CrCl was not different between groups. At termination, CrCl has decreased in WU-B AngII infused group and has increased in WU-B Control group, making the difference between groups statistically significant. In WU-L Control group CrCl has also decreased, while in WU-L AngII infused group has not changed. Data are presented as median and interquartile range. *p < 0.05, baseline versus termination, same genotype; # p < 0.05, Control versus AngII infusion group, same timepoint.
pro-inflammatory and pro-fibrotic properties. 15, 31, 42, 43 However, negative feedback can be expected to offset these effects. Given that ACE holds one of the central roles in the RAAS, it can be hypothesized that with this basal high or low ACE levels goes also a baseline variation in RAAS system equilibrium. Consequently, the activation of a RAAS cascade in pathological situations 44 could lead to a different response to damage as found in our study.
Further, the renal ACE and its homologue ACE2 are reciprocally expressed between the two genotypes. At baseline, higher ACE2 enzyme and activity seemed to have been combined with lower ACE activity in the kidney, and vice versa. It has been argued that the two enzymes may be counter-balancing each other in vivo. 17, 18 The ACE2 enzyme has a high substrate specificity for AngII by converting it into Ang(1-7). 19, 45 Endogenous ACE2 was suggested to have renoprotective properties in chronic kidney disease 24 while infusion of Ang(1-7) was recently shown to counteract AngII and reduces glomerulosclerosis in rats. 46 The reciprocal nature of these two enzymes found between WU-B and WU-L, leads us to question whether the baseline predominance of one could lead to different susceptibility to AngII-induced renal damage. Several factors could contribute to the elevated baseline ACE2 level in the WU-L group. It could be that the innate RAAS balance with high ACE (as in WU-B) favors the ACE/ AngII/AT1 axis while low innate ACE in the WU-L favors the ACE2/Ang(1-7)/Mas receptor axis in the kidney. In renal cortex of diabetic mice, it has been shown that there is a mismatch between ACE2 mRNA and protein levels, and proposed that ACE2 might be regulated at the posttranscriptional level, 47 while in renal epithelial cells AngII increased vascular epithelial growth factor protein expression without altering the gene expression levels. 48 Even though ACE2 activity was not increased in the WU-L after AngII infusion, the initial higher ACE2 baseline values might have been one of the beneficial factors in counteracting the renal effects of AngII.
AngII infusion lead to a significant increase in blood pressure in both lines. Original studies describing b and l Ace alleles have consecutively shown no difference in blood pressure levels. 2, 49 Furthermore, Wistar rats do not spontaneously develop hypertension. Nevertheless, during the current experiment we did measure higher blood pressure in WU-L at the age of 11 weeks. We appreciate that the blood pressure values we measured do not correspond to the previously reported studies, and we do not have a specific explanation for this discrepancy. Also, although there was a difference in main baseline body weight in the WU-L group that would receive AngII infusion, the animals from the baseline WU-L control group did not have significantly different body weight from the WU-B control group. During the experiment all animals were placed in the same environment and had same access to food and water. In addition, when we first characterized the WU-B and WU-L lines we observed no difference in body weights between the two lines at the same age. We believe that these differences are not a phenotypic trait characteristic for these two lines.
We did not measure AngII or Ang(1-7) levels during the experiment, thus it is hard to speculate about the exact regulatory pathway in the presented different susceptibility to AngII infusion.
In conclusion, rats with genetically predetermined high Ace expression are more susceptible to develop AngIIinduced renal damage. Behind this genetic Ace variability might exist a different RAAS symmetry. The full understanding of this polarity is yet to be determined, as is the possible value in disease prediction, progression and therapy.
|
