ARB Protection Against Podocyte-Induced Sclerosis is Podocyte AT1-Independent

Abstract

In the present study, we tested the hypothesis that the reno-protective effect of ARB is dependent on the angiotensin II type 1 (AT1) receptor on podocytes. For this purpose, we generated podocyte-specific knockout mice for AT1 gene (Agtr1a), and crossed with NEP25, in which selective podocyte injury can be induced by immunotoxin, LMB2. Four weeks after LMB2, urinary albumin/creatinine ratio was not attenuated in Agtr1a knockout/ NEP25 (n=18), compared with that in control NEP25 mice (n=13) (8.08±2.41 in knockout vs. 4.84±0.73 in control). Both strains of mouse showed similar degrees of sclerosis (0.66±0.17 vs. 0.82±0.27 on 0 to 4 scale) and downregulation of nephrin (5.78±0.45 vs. 5.65±0.58 on 0 to 8 scale). In contrast, AT1 antagonist or an angiotensin I converting enzyme inhibitor, but not hydralazine, remarkably attenuated proteinuria and sclerosis in NEP25 mice. Moreover, continuous angiotensin II infusion induced microalbuminuria similarly in both Agtr1a knockout and wild type mice.

Thus, angiotensin inhibition can protect podocytes and prevent development of glomerulosclerosis independently of podocyte AT1. Possible mechanisms include inhibitory effects on AT1 of other cells, or through mechanisms independent of AT1.

Our study further demonstrates that measures that directly affect only non-podocyte cells can have beneficial effects even when sclerosis is triggered by podocyte-specific injury.

Introduction

Podocytes play an indispensable role as a filtration barrier for macromolecules in the glomerulus. Damage of podocytes is a key step triggering the progression of glomerulosclerosis. A large volume of evidence indicates that angiotensin (Ang) II acting on the Ang II type 1 receptor (AT1) plays important roles in this process. Blockade of Ang II synthesis with Ang I converting enzyme (ACE) inhibitors or of Ang II action with AT1 receptor blocker (ARB) is a clinically established therapeutic measure for slowing the progression of chronic kidney diseases. ACE inhibitors and ARB have been shown to attenuate podocyte damage, proteinuria and development of glomerulosclerosis in a variety of animal models, including, among others,^1–7 subtotal nephrectomy model⁸ and diabetic nephropathy models.^{9, 10}

Continuous infusion of Ang II in normal rats increased desmin expression,¹¹ suppressed nephrin and podocin mRNA.⁶ Studies conducted thus far collectively indicate that Ang II is involved in triggering, enhancing and expanding podocyte injury and in the progression of glomerular injury toward sclerosis through mechanisms beyond its effect on systemic blood pressure.

Since podocyte injury along with proteinuria ubiquitously precedes progressive development of glomerulosclerosis, and since Ang II inhibition attenuates podocyte damage and progressive glomerulosclerosis, it appears reasonable to speculate that inhibition of AT1 on podocytes is the key to the protective effect of pharmacological Ang II blockage. In fact, several lines of evidence indicate that Ang II has direct cellular effects on podocytes. Cultured mouse podocytes express mRNA and protein for AT1.^{6, 12} Podocytes in isolated glomeruli express functional AT1, and Ang II depolarizes and increases intracellular Ca²⁺.^{13, 14} In cultured podocytes, Ang II, via AT1, increases α3(IV) collagen and VEGF,^{15, 16} decreases nephrin,^{6, 17} heparan sulfate proteoglycans¹⁸ and α-actinin-4, augments ROS production, and induces redistribution of zona occludens-1 (ZO-1) and reorganization of F-actin cytoskeleton.^{19, 20} More directly, transgenic rats overexpressing AT1 receptor selectively in podocytes develop glomerulosclerosis.²¹ These data are consistent with the notion that the beneficial effect of Ang II blockade on glomerulosclerosis is attributed to its direct inhibitory effects on Ang II action on podocytes.

Earlier, we established a transgenic mouse line (NEP25), which expresses human (h) CD25 (i.e., IL2 receptor) selectively on podocytes. Since hCD25 does not react with mouse IL2 ligand, it is highly unlikely that expression of hCD25 per se affects podocyte function, including Ang II signaling. By injecting a hCD25-targeted recombinant immunotoxin, anti-Tac(Fv)-PE38 (LMB2), podocyte-selective injury can be induced in NEP25 mice. LMB2 is a recombinant chimeric protein composed of PE38 (a mutant form of pseudomonas exotoxin A) and Fv domain of monoclonal anti-hCD25 antibody. LMB2 (Mw 63 kDa) can cross the glomerular basement membrane, and its half-life in the circulation is 35 minutes in mice.²² After a single injection of LMB2 (at 0.625 ng/g BW), NEP25 mice develop moderate proteinuria, which peaks one to two weeks after the injection and gradually decreases. Within two weeks, NEP25 mice have minor podocyte injury. After 3 weeks, they show progressive damage of podocytes and other glomerular cells and develop focal segmental glomerulosclerosis.²³

In order to clarify the role of Ang II on podocytes during the progression of glomerulosclerosis, we generated podocyte-specific AT1 deficient mice. Unlike humans, mice have two AT1 receptor genes, Agtr1a and Agtr1b, each encoding AT1A and AT1B receptor subtypes, respectively. In the mouse kidney, more than 99% of AT1 mRNA is derived from Agtr1a, and Agtr1a inactivation does not lead to activation of Agtr1b.²⁴ We, therefore, used podocyte-specific Agtr1a null-mutant mice.

Methods

The Animal Experimentation Committee of Tokai University approved the protocol, in accordance with the principles and procedures outlined in the National Institute of Health Guide for the Care and Use of Laboratory Animals.

Generation of podocyte-specific Agtr1a null-mutant mice

Detailed methods for generation of Agtr1a^loxP mice (C57BL/6 background) are described in the online supplements (please see http://hyper.ahajournals.org.) Nephrin-Cre mice were previously reported.²⁵ In the present study, line 10 of Nephrin-Cre mice was used. They were backcrossed with C57BL/6 strain more than three times and used for mating with Agtr1a^loxP mice. To induce podocyte-specific injury, mice carrying Agtr1a^loxP and Nephrin-Cre were further mated with NEP25 mice²³ on C57BL/6 genetic background.

Determination of genotype of podocytes

From mice carrying Agtr1a^loxP/loxP/Cre(+)/TRE-SV40T/podocin-rtTA/ROSA26^loxP or Agtr1a^loxP/+/Cre(+)/TRE-SV40T/podocin-rtTA/ ROSA26^loxP genotype, glomeruli were isolated by perfusing with Dynabeads (Dynal ASA, Oslo, Norway). ²⁶ Glomeruli were cultured on laminin-coated dishes in the presence of doxycycline (1 µg/ml) for 5 days. Sprouting cells were sparsely re-plated and cultured until they formed colonies. Cells were then fixed in 2% glutaraldehyde/PBS for 10 minutes, and stained for lacZ in a staining solution [2 mM MgCl₂, 0.02% Nonidet-P40, 0.01% Na deoxycholic acid, 5 mM K₃Fe(CN)₆, 5 mM K₄(CN)₆, 1mg/ml X-gal in PBS (pH 7.4)] at 37°C for 4 hours. After washing with PBS, isolated colonies were surrounded by O-rings and cells were lysed in [10 mM Tris-HCl, 1 mM EDTA, 1% Tween 20, 0.4 mg/ml proteinase K] at 55°C for 12 hours. The lysate containing genomic DNA was harvested, heated to inactivate proteinase K and used as templates of PCR.

The following three primers were used: AT5, ATCCCTGAATGCTAATCTTG; AT6, GGCTTTATATCCCACTGTTG; AT7, GAATGTTTGGGGGTTTTTGT (Figure 1). ES cell DNA carrying Agtr1a^−/+ or Agtr1a^loxP/+ and tail DNA carrying Agtr1a^+/+ or Agtr1a^loxP/loxP were used as controls.

Figure 1. Verification of Agtr1a disruption in the podocytes of Agtr1a^loxP/loxP/Cre(+) mice.

(a) PCR for the genotyping of Agtr1a.

In wild type Agtr1a, the primer pair AT6 and AT7 generated a 300 bp band. The primer pair AT5 and AT7 did not amplify DNA.

In Agtr1a^loxP, the primer pair AT6 and AT7 generated a 366 bp band, while the primer pair AT5 and AT7 did not amplify DNA.

In Agtr1a⁻, the primer pair AT5 and AT7 generated 213 band. The portion recognized by AT6 is removed from the Agtr1a⁻ genome.

(b) Representative PCR result.

Lane 1: DNA size markers. Lane 2: ES cell DNA carrying Agtr1a^−/+. Lane 3: ES cell DNA carrying Agtr1a^loxP/+. Lane 4: Tail DNA carrying Agtr1a^+/+ (wild-type). Lane 5: Tail DNA carrying Agr1a^loxP/loxP. Lanes 6 and 7: LacZ-tagged podocyte clones from mice carrying Agtr1a^loxP/+/Cre(+)/TRE-SV40T/podocin-rtTA/ROSA26^loxP. Both show Agr1a^−/+ genotype. Other six clones (not shown) similarly showed Agr1a^−/+ genotype. Lane 8–12: LacZ-tagged podocyte clones from mice carrying Agtr1a^loxP/loxP/Cre(+)/TRE-SV40T/podocin-rtTA/ROSA26^loxP. All show Agr1a^−/− genotype. Thirteen out of 15 clones examined (87%) showed Agr1a^−/− genotype.

Determination of relative amount of AT1A and AT1B mRNA

Glomeruli were harvested from Agtr1a^loxP/loxP/Cre(−) and Agtr1a^loxP/loxP/Cre(+) mice without LMB2, or Agtr1a^loxP/loxP/Cre(−)/NEP25 and Agtr1a^loxP/loxP/Cre(+)/NEP25 7 days after 0.625 ng/g BW of LMB2 (each n=4). Total RNA was extracted, and cDNA was synthesized. As reported previously,²⁴ PCR was carried out using primers, GCATCATCTTTGTGGTGGG, GAAGAAAAGCACAATCGCC, which are common to both AT1A and AT1B sequences. Only the PCR product derived from AT1A mRNA has an EcoRI site. Therefore, EcoRI-cleaved and uncleaved bands represent AT1A and AT1B mRNA, respectively.

AT1A and AT1B mRNA were also quantified by real-time RT-PCR. TaqMan Primer Probe Sets for Agtr1a, Agtr1b and 18s RNA were used with Applied Biosystems 7300 Real Time PCR Systems. The amplification efficiency for AT1A and AT1B was normalized by a standard template containing equal molar of AT1A and AT1B cDNAs.

Real-time RT-PCR was also performed in RNA extracted from primary cultured podocytes after the first and second passage. Primary cultured podocytes were obtained from wild-type mice (n=3) as described previously.²⁷

Experimental protocol of immunotoxin-induced nephropathy

To study the effect of podocyte-specific AT1 inactivation, five female and eight male mice carrying Agtr1a^loxP/loxP/Cre(−)/NEP25 and 10 female and eight male mice carrying Agtr1a^loxP/loxP/Cre(+)/NEP25 (3–7 months of age) were used. 24-hour urine was collected before and 7, 14, 21, 28 days after LMB2 injection. They were sacrificed 28 days after LMB2 injection.

To study the effect of losartan, an AT1 receptor blocker, eight female NEP25 mice with C57BL/6 genetic background (5 months of age) were treated with losartan (0.5 g/l, in drinking water, approximately 25 µg/g BW) from 5 days before LMB2 injection until the end of the experiment. Nine age-sex-matched NEP25 mice were used as controls. In both groups of mice, 0.625 ng/g BW of LMB2 was intravenously injected under diethyl ether anesthesia. Conscious systolic blood pressure was measured by tail cuff method using MK-2000 (Muromachi Kikai, Tokyo, Japan) two days before LMB2 injection. Systolic blood pressure measured by MK-2000 is reported to be well correlated with, but 8.8 mmHg lower than, that measured by cannulation into the carotid artery (Manufacture’s information). 24-hour urine was collected before and 7, 14, 21 days after LMB2 injection. Twenty-one days after the injection, all mice were sacrificed and renal injury was analyzed. A study examining the effect of ARB was repeated in 12 male Agtr1a^loxP/loxP/Cre(+)/NEP25 mice (8–11 months of age, 7 without and 5 with ARB treatment). In this study, urinary albumin/creatinine ratio and histology at 4 weeks were evaluated.

Infusion of Ang II

Eight female Agtr1a^loxP/loxP/Cre(−) and six female Agtr1a^loxP/loxP/Cre(+) mice (13 months of age) were continuously infused with Ang II (1,000 ng/kg/min) for 14 days using mini-osmotic pumps (Alza, model 2002) that were implanted subcutaneously under diethyl ether anesthesia.

Urinalysis

Twenty-four-hour urine specimens were collected using metabolic cages. Concentrations of total protein and creatinine in the urine were determined by the pyrogallol red and enzymatic methods, respectively, in an outside laboratory (SRL, Tokyo, Japan). Concentration of albumin in the urine was determined with an enzyme linked immunosorbent assay (ELISA) kit (Albuwell M; Philadelphia, PA).

Morphological analysis

Glomerulosclerosis was evaluated in PAS-stained paraffin sections (2 µm thick). Each glomerulus was graded on a 0 to 4 scale, which represent sclerotic area involving 0, 1 to 25, 26 to 50, 51 to 75, or >75% of the glomerulus. Scores for all glomeruli on a section were averaged and defined as sclerosis index for each mouse.

For evaluating podocyte injury, paraffin sections were stained for nephrin using guinea pig polyclonal antibody (GP-N2, Progen). For semi-quantification of nephrin staining, each quadrant of each glomerulus was scored as 0 (no staining), 1 (diminished) or 2 (normal), with total glomerular score range calculated from 0 (complete loss) −8 (normal). Scores for all glomeruli on a section for each mouse (>80) were averaged and defined as nephrin index.

Statistical Analysis

Results are expressed as means ± SE. Student’s t test was used to analyze the difference between two groups in blood pressure. One-way analysis of variance (ANOVA) was used to compare blood pressure among three groups. Albumin/creatinine ratio was measured repeatedly; and global test assessing between group effect was performed through multivariate analysis of variance (MANOVA) in order to prevent inflation of Type I error through multiple comparisons. When overall effect was detected with MANOVA, Student t-test was performed at individual time points. Comparison of sclerosis index and nephrin index was performed by Mann-Whitney’s U test. Values were regarded as significant at 2-sided P < 0.05.

Results

Generation of Agtr1a^loxP/loxP mice

To investigate direct effect of Ang II on podocytes, we generated podocyte-specific Agtr1a–null mutant mice. Utilizing homologous recombination in ES cells, we have established mutant mice carrying Agtr1a^loxP, in which two loxP sites were inserted at the upstream and the lower steam to the coding exon of Agtr1a (Figure S1, please see http://hyper.ahajournals.org).

Homozygous (Agtr1a^loxP/loxP) and heterozygous (Agtr1a^loxP/+) mice showed no apparent abnormal phenotype. Northern blotting analysis revealed that Agtr1a^loxP/loxP and Agtr1a^loxP/+ mice similarly expressed AT1A mRNA in the kidney compared to wild-type mice (data not shown). Systolic blood pressure was 97±13 and 103±8 mmHg in Agtr1a^loxP/loxP and Agtr1a^loxP/+ mice, respectively, similar to that in wild-type littermates (100±12 mmHg). Agtr1a^loxP/loxP showed normal renal morphology. These data confirm that the insertion of the loxP sequences did not disturb the expression and the function of AT1A mRNA.

Confirmation of Agtr1a disruption in podocytes of Agtr1a^loxP/loxP/Cre(+) mice

The Agtr1a^loxP/loxP line was mated with a Nephrin-Cre line, which expresses Cre recombinase selectively in podocytes. Previously, we tested efficiency of Cre-mediated recombination in podocytes by mating Nephrin-Cre mice with ROSA26^loxP, a tester strain. 100% of podocytes were lacZ positive in Nephrin-Cre/ROSA26^loxP. We next tested whether recombination occurs in similar efficiency in podocytes of Agtr1a^loxP/loxP/Nephrin-Cre (Agtr1a^loxP/loxP/Cre(+)) mice. Reliable anti-mouse AT1 antibodies suitable for immunohistochemical study were not available to us, and we therefore cultured podocytes, and then cloned and determined the Agtr1a genotype by PCR. For this purpose, Agtr1a^loxP/loxP/Cre(+) mice were mated with TRE-SV40T/podocin-rtTA mice, which express SV40 T antigen in podocytes in the presence of doxycycline. To mark the podocyte-lineage with lacZ, the mice were further mated with ROSA26^loxP line.

Glomeruli obtained from Agtr1a^loxP/loxP/Cre(+)/TRE-SV40T/podocin-rtTA/ ROSA26^loxP mice were cultured in the presence of doxycycline. Colonies, each stems from a single cell, were stained for lacZ. PCR analysis revealed that 13 out of 15 (87%) lacZ-positive clones examined showed only deleted allele (Agtr1a⁻)(Figure 1). Two lacZ positive colonies showed both Agtr1a^loxP and Agtr1a⁻, indicating that they were heterozygote (Agtr1a^loxP/−). LacZ-negative cobblestone-like cells often grew even without doxycycline. PCR analysis revealed that all three such colonies examined had Agtr1a^loxP/loxP genotype. Similar analysis in eight lacZ positive colonies from Agtr1a^loxP/+/Cre(+)/TRE-SV40T/podocin-rtTA/ROSA26^loxP showed that all lacZ-positive clones had Agtr1a^−/+ genotype. These confirmed that Cre-mediated recombination of Agtr1a^loxP occurs efficiently in podocytes and most podocytes in Agtr1a^loxP/loxP/Cre(+) mice were indeed null-mutated for Agtr1a.

Basal phenotype of Agtr1a^loxP/loxP/Cre(+) mice

Podocyte-selective AT1 knockout mice, Agtr1a^loxP/loxP/Cre(+), showed no apparent abnormal phenotype in a basal condition. Thus, at 4 months of age, urinary albumin/creatinine ratio in Agtr1a^loxP/loxP/Cre(+) mice was not different from that in control Agtr1a^loxP/loxP/Cre(−) mice, male (0.17±0.02, n=8 vs. 0.14±0.02, n=10) or female (0.05±0.01, n=8 vs. 0.05±0.02, n=10). Systolic blood pressure measured at 4 months of age in Agtr1a^loxP/loxP/Cre(+) was not different from that in Agtr1a^loxP/loxP/Cre(−) mice (118.0±3.2 vs 113.4±5.0 mmHg, each n=6). Renal histology and nephrin staining in Agtr1a^loxP/loxP/Cre(+) mice were normal over a range from age 1 to 12 months (data not shown).

To determine relative amount of AT1A and AT1B mRNA, RT-PCR was performed in glomerular RNA from Agtr1a^loxP/loxP/Cre(+) and Agtr1a^loxP/loxP/Cre(−) mice using primers common to AT1A and AT1B sequences following digestion with EcoRI, which is specific to AT1A. AT1B mRNA was undetectable in glomeruli of either type of mice with this method. Real-time RT-PCR revealed that AT1B mRNA was detectable in the glomerulus, but the quantity was less than 0.5% (0.0 – 0.5 %, n=4) of that of AT1A in either Agtr1a^loxP/loxP/Cre(+) or Agtr1a^loxP/loxP/Cre(−) mice. Induction of podocyte injury by LMB2, which is shown below, did not enhance AT1B mRNA. In addition, AT1A mRNA was detectable in primary cultured podocytes by RT-PCR, but AT1B mRNA was undetectable.

Effect of podocyte-specific inactivation of AT1 on progression of glomerulosclerosis

To examine whether podocyte-specific deletion of AT1 receptor can slow the progression of glomerulosclerosis, we mated Agtr1a^loxP/loxP/Cre(+) mice with NEP25 line, generating Agtr1a^loxP/loxP/Cre(+)/NEP25 and Agtr1a^loxP/loxP/Cre(−)/NEP25 mice. Without LMB2, both types of mice showed no proteinuria (Figure 2a, before LMB2), and renal morphology and nephrin staining were normal (Figure 3a and d) with sclerosis index 0, and nephrin score 8.

Figure 2. Effect of podocyte-specific AT1 deletion on glomerular injury.

Agtr1a^loxP/loxP/Cre(−)/NEP25 (open columns) and Agtr1a^loxP/loxP/Cre(+)/NEP25 (closed columns) mice were injected with LMB2 (0.625 ng/g BW). There was no significant difference in urinary albumin/creatinine ratio (ACR)(a), sclerosis index (b) or nephrin index (c) between the two types of mice. Without LMB2, both types of mice showed no sclerosis or podocyte damage with sclerosis index 0, and nephrin index 8.

Figure 3. Representative pictures of Agtr1a^loxP/loxP/Cre(−)/NEP25 and Agtr1a^loxP/loxP/Cre(+)/NEP25 mice.

Without LMB2, Cre (+) mice showed normal glomerular morphology (a) and normal nephrin staining pattern (d). After LMB2 injection, both Cre (−) and Cre (+) mice showed similar degrees of glomerulosclerosis (b and c) and similar degree of downregulation of nephrin (e and f, arrows).

(a–c, PAS, X 200; d–f, anti-nephrin, X400)

After injection of 0.625 ng/g BW of LMB2, both types of mice showed moderate proteinuria, which peaked two weeks after the injection, with no statistical difference at any time point (Figure 2a).

The degree of focal segmental glomerulosclerosis was similar in both mouse groups at 4 weeks. Agtr1a^loxP/loxP/Cre(+)/NEP25 mice had sclerosis index of 0.65±0.16, which was not statistically different from that of Agtr1a^loxP/loxP/Cre(−)/NEP25 mice, 0.82±0.27 (Figure 2b and Figure 3). Both types of mice had glomeruli with similarly diminished nephrin staining. The nephrin index in Agtr1a^loxP/loxP/Cre(+)/NEP25 mice was, on average, 5.78±0.45, which was not statistically different from that of Agtr1a^loxP/loxP/Cre(−)/NEP25 mice, 5.63±0.58 (Figure 2c and Figure 3). Thus, podocyte-specific inactivation of AT1 showed no impact on the progression of glomerulosclerosis triggered by podocyte injury.

ARB attenuates progression of glomerulosclerosis in NEP25 model

We next examined whether ARB can protect against the glomerular injury in NEP25 transgenic mice which progressively develops after the injection of LMB2 (0.625 ng/g BW). The ARB-treated mice showed significantly lower systolic blood pressure than the control mice without treatment (66±6 vs. 101±5 mmHg). NEP25 mice without ARB showed moderate proteinuria with marked decrease in ARB NEP25 (urinary protein/creatinine ratio, 66.6±20.7 vs. 22.8±10.3 mg/mg) seven days after LMB2 injection (Figure 4a). Twenty-one days after injection, NEP25 mice showed focal segmental sclerosis with sclerosis index averaging 0.83±0.36, demonstrating marked protection in ARB NEP25 mice (0.01±0.01) (Figure 4b and Figure 5).

Figure 4. Effect of ARB treatment on NEP25 mice.

NEP25 mice were injected with LMB2 (0.625 ng/g BW). Mice with ARB treatment that started before LMB2 injection (closed columns) showed significantly attenuated urinary total protein/creatinine ratio (U-Pro/Cr) (a), glomerulosclerosis (b) and downregulation of nephrin (c) when compared to those in control mice without ARB (open columns). Without LMB2, NEP25 mice showed no sclerosis or podocyte damage with sclerosis index 0, and nephrin index 8. * represents p<0.05.

Figure 5. Representative pictures of NEP25 mice with or without ARB.

NEP25 mice 21 days after LMB2 without ARB treatment show severe glomerulosclerosis (a) and remarkable decrease in nephrin staining (c, arrows). In ARB-treated NEP25 mice 21 days after LMB2, glomerular injury (b) and nephrin staining (d) are remarkably improved.

(a, b and c, PAS, X 200; d, e and f, anti-nephrin, X400.)

Podocyte injury, assessed by nephrin staining, was significantly attenuated in NEP25 vs. ARB NEP25 (nephrin staining index, 5.61±0.76 vs. 7.96±0.01, 0–8 scale)(Figure 4c and 5).

Similar protection was observed in Agtr1a^loxP/loxP/Cre(+)/NEP25 mice treated with ARB(Figure 6), Agtr1a^loxP/loxP/Cre(−)/NEP25 mice treated with ARB starting after LMB2 injection (Figure S2), and NEP25 mice with captopril (Figure S3), but not with hydralazine (Figure S4) (please see http://hyper.ahajournals.org.).

Figure 6. Effect of ARB treatment on Agtr1a^loxP/loxP/Cre(+)/NEP25 mice.

Agtr1a^loxP/loxP/Cre(+)/NEP25 mice were injected with LMB2 (0.625 ng/g BW). Mice with ARB treatment (closed columns) showed significantly attenuated urinary albumin/creatinine ratio (ACR)(a), glomerulosclerosis (b) and downregulation of nephrin (c) when compared to those in control mice without ARB (open columns). * p<0.05.

Effect of podocyte-specific inactivation of AT1 on Ang II-induced microalbuminuria

We next examined the effect of podocyte-specific inactivation of AT1 on microalbuminuria induced by Ang II infusion. Pressor dose of Ang II (1,000 ng/kg/min) was continuously infused for 14 days using mini-osmotic pumps in Agtr1a^loxP/loxP/Cre(−) and Agtr1a^loxP/loxP/Cre(+) mice. Systolic blood pressure was similarly elevated in both groups (149.3±11.2 vs. 143.5±14.3, respectively). As shown in Figure 7, both strains of mice showed similar degree of microalbuminuria at all time points examined. Under this experimental condition, no mouse in these groups showed glomerular sclerosis or downregulation of nephrin staining.

Figure 7. Microalbuminuria induced by Ang II infusion.

Agtr1a^loxP/loxP/Cre(−) (open columns) and Agtr1a^loxP/loxP/Cre(+) (closed columns) mice were continuously infused with Ang II (1000 ng/Kg/min) for 14 days. There was no significant difference in urinary albumin/creatinine ratio (ACR) at any time point between the two types of mice.

Discussion

In the present study, blockage of Ang II, by either ARB or ACE inhibitor, attenuated proteinuria, podocyte injury and glomerulosclerosis in the NEP25 model in a fashion similar to those of other models for podocyte injury. Treatment with hydralazine showed no renal protective effect although it effectively decreased blood pressure, indicating that the protective effect of Ang II blockage is independent of its systemic blood pressure lowering effect. Since ACE inhibitor was also effective, the beneficial effect of ARB is ascribed to inhibition of AT1, not to an activation of non AT1 Ang II receptor(s).

As discussed earlier, podocytes in vitro, as well as in vivo, have been shown to express functional AT1.^{13, 14} Moreover, transgenic rats overexpressing AT1 receptor selectively in podocytes showed progressive increase in albuminuria and pseudocyst formation in podocytes, which were followed by development of glomerulosclerosis.²¹ The present study failed to show that podocyte-specific AT1 inactivation has any impact on baseline or Ang II-induced proteinuria. One might consider the possibility that this apparent discrepancy is due to a difference in the duration and/or magnitude of AT1 stimulation. In this regard, our study on podocyte-specific AT1 knock out mice indicates that the microalbuminuria induced by short-term (2 weeks) Ang II infusion does not reflect its local effect on podocytes. This, in turn, points to the notion that an event occurring in non-podocyte cells lead to alteration in the sieving function of the glomerulus.

Of importance, the present study convincingly demonstrated that AT1 blocker protects podocytes primarily by a mechanism independent of its inhibitory effect on the AT1 on podocytes. Ang II can increase glomerular capillary ultrafiltration pressure by increasing efferent arteriolar resistance and/or by lowering afferent arteriolar resistance.^{28, 29} This leads to an enhancement of leakage of macromolecules through the glomerular capillary wall.³⁰ A variety of studies have demonstrated association between the glomerular capillary pressure and progression of glomerulosclerosis. Although molecular mechanism for this connection is yet to be established, the present study is consistent with the view that the capacity of ARB to decrease glomerular capillary pressure contributes to the protective effect of ARB in this NEP25 model. Theoretically, this notion can be verified by a study on efferent arteriole-specific AT1 knockout mice. However, no promoter segment is currently known that can drive efferent arteriole-specific expression of the Cre gene.

LMB2 inhibits protein synthesis by inactivating elongation factor 2 in targeted podocytes. One may, therefore, consider the possibility that the lack of beneficial effect of podocyte-specific AT1 inactivation is ascribed to suppression of AT1 protein in podocytes of the control NEP25 mice (Agtr1a^loxP/loxP/Cre (−)/NEP25). Although we did not quantify the amount of AT1 protein in podocytes, the dose of LMB2 in this study did not decrease any of the other proteins examined thus far, including nephrin, Wilms tumor 1 (WT1), VEGF and synaptopodin, at a week after injection in NEP25 mice. Further, ARB was also equally effective on control NEP25 mice given LMB2.

Unlike human, mice have AT1B, another subtype AT1 receptor. The content of AT1B mRNA was less than 0.5% of AT1A mRNA in whole kidney.^{24, 31} Recently, Crowley et al reported that AT1B mRNA is concentrated in podocytes, and that AT1B receptor may be stimulated by increased Ang II ligand in whole-body AT1A knockout mice.³² Our results of lack of protective effect of podocyte-specific AT1A inactivation are not attributed to compensation by AT1B. Previously, no Ang II binding in the presence of AT2 antagonist was detected in the kidney of AT1A knockout mice by binding autoradiography.³³ In the present study, we found that AT1B mRNA was less than 0.5% of AT1B mRNA in the glomerulus, and not increased by podocyte-AT1A knockout or by LMB2 injection. In primary cultured-podocytes, AT1A mRNA, but not AT1B mRNA, was detectable by RT-PCR. Moreover, podocyte specific AT1A knockout mice had normal blood pressure, therefore it is unlikely that significant upregulation occurred in renin or Ang II ligand.

Remaining possibilities for the beneficial effects of Ang II blockage include inhibition of AT1 on non-podocyte cells within and outside the glomerulus. The latter includes zona glomerulosa of the adrenal gland, i.e., via inhibition of aldosterone synthesis and release. Some actions of ARB unrelated to Ang II, such as antioxidant effect of ARBs and ACE inhibitors may also be contributory.^34–36

Perspectives

The unique design of the present study reveals a novel concept, namely, therapeutic measures even when targeting only non-podocyte cells directly can profoundly affect the process of glomerulosclerosis, including ones initially triggered by selective podocyte injury. This notion will broaden the options in designing therapeutic measures to disrupt the process of glomerulosclerosis.