Glitazones are efficient insulin sensitizers that blunt the effects of angiotensin II (ANG II) in the rat. Sodium chloride is another important modulator of the systemic and renal effects of ANG II. Whether glitazones interfere with the interaction between sodium and the response to ANG II is not known. Therefore, we investigated the effects of pioglitazone on the relationship between sodium and the systemic and renal effects of ANG II in rats. Pioglitazone, or vehicle, was administered for 4 wk to 8-wk-old obese Zucker rats. Animals were fed a normal-sodium (NS) or a high-sodium (HS) diet. Intravenous glucose tolerance tests, systemic and renal hemodynamic responses to ANG II, and the renal ANG II binding and expression of ANG II type 1 (AT1) receptors were measured. The results of our study were that food intake and body weight increased, whereas blood pressure, heart rate, filtration fraction, and insulin levels decreased significantly with pioglitazone in obese rats on both diets. Pioglitazone blunted the systemic response to ANG II and abolished the increased responsiveness to ANG II induced by a HS diet. Pioglitazone modified the renal hemodynamic response to changes in salt intake while maintaining a lower filtration fraction with ANG II perfusion. These effects were associated with a decrease in the number and expression of the AT1 receptor in the kidney. In conclusion, these data demonstrate that the peroxisome proliferator-activated receptor-γ agonist pioglitazone modifies the physiological relationship between sodium chloride and the response to ANG II in insulin-resistant rats.
- peroxisome proliferator-activated receptor-γ
- blood pressure
glitazones are high-affinity agonists of peroxisome proliferator-activated receptors (PPARs) presently used in the treatment of type 2 diabetes as efficient insulin sensitizers. These agents lower blood pressure (BP) in diabetic (24) and hypertensive (12, 26) humans and in animal models of renovascular or salt-sensitive hypertension (37) and insulin resistance (15, 33). Although the glitazone-induced improvement in insulin sensitivity is often associated with a decrease in BP (12, 26), the mechanisms involved are still unclear. Several studies (8, 28, 31) demonstrate angiotensin II (ANG II)-antagonizing properties of glitazones. These effects are of particular interest because blocking the renin-angiotensin system reduces the cardiovascular risk of diabetic subjects and retards the progression of diabetic nephropathy (5).
This study further examines the interactions between pioglitazone and the renin-angiotensin system in insulin-resistant rats by focusing on the role of sodium. Sodium is an important physiological modulator of the systemic and renal response to ANG II. On a low-sodium intake, the renin-angiotensin system is activated, leading to high circulating ANG II levels and a downregulation of ANG II receptors, whereas, on a high-sodium (HS) intake, circulating ANG II levels are low and ANG II receptors are upregulated, leading to an increased sensitivity to ANG II (22). Glitazones blunt the systemic response to exogenous ANG II in rats (16) and also increase food intake (34). Thus glitazones may potentially modify the impact of salt on the vascular effects of ANG II. Whether a HS intake modulates the effects of glitazones on the activity of the renin-angiotensin system has not been investigated so far.
In this study, we examined the systemic and renal responses to ANG II in insulin-resistant Zucker rats treated with pioglitazone or vehicle on a normal-sodium (NS) and HS diet. The Zucker rat model develops obesity secondary to hyperphagia due to a missense autosomal recessive mutation of the fa gene encoding the leptin receptor (25). With time, these rats become insulin resistant, salt sensitive, and dyslipidemic, and they develop proteinuria and renal lesions.
MATERIALS AND METHODS
Animals and drug treatment.
Pioglitazone (20 mg·kg−1·day−1, n = 20), metformin (200 mg·kg−1·day−1, n = 18), or vehicle (n = 25) was administered for 4 wk to 8-wk-old obese fa/fa Zucker rats (Iffa Credo, Lyon, France). Pioglitazone (n = 9) and vehicle (n = 9) were also administered at similar doses for 4 wk to 8-wk-old lean, heterozygote (fa/?) Zucker rats. Pioglitazone was mixed with the rat chow, and metformin was added to the drinking water. These groups of animals were fed a regular sodium diet. A group of obese pioglitazone- (n = 14) and vehicle-treated rats (n = 15) were given a HS diet by having their regular drinking water replaced with 0.9% NaCl. Procedures used in this study were in accordance with principles of animal care and according to the institutional guidelines of Centre Hospitalier Universitaire Vaudois.
Systemic hemodynamics and intravenous glucose tolerance test.
At the end of the treatment period, all animals were anesthetized with halothane and had their right femoral arteries and veins cannulated with a PE50 catheter. The animals were then returned to their cage. On the following day, after an overnight fast, they were placed in a plastic tube for partial restriction of their movements. Intra-arterial pressure and heart rate (HR) were recorded continuously in the awake animal with a computerized data acquisition system after 1 h of rest (36). In a subgroup of animals, a BP/dose-response curve to ANG II was obtained using bolus intravenous injections of increasing doses of exogenous ANG II (Peninsula, Belmont, CA). The peak ANG II-induced change in mean BP (MBP) was monitored in obese vehicle-, obese metformin-, and obese pioglitazone-treated rats as well as in lean placebo- and lean pioglitazone-treated rats (n = 7–10/group). At the end of the ANG II dose-response curves, an intravenous glucose tolerance test was performed with a bolus injection of 0.5 g/kg glucose and the determination of whole blood glucose and plasma insulin at 0, 4, 7, 10, 15, and 30 min. Insulin resistance index at baseline was assessed by homeostasis model assessment of insulin resistance (HOMA-IR) (19).
In a subgroup of rats, the renal hemodynamic response to exogenous ANG II was examined. Hemodynamic measurements were performed using sinistrin (an analog of inulin) and PAH clearances after 1 h of rest. Sinistrin (Laevosan Gesellschaft, Zurich, Switzerland) and PAH (Clinalpha, Laufelfingen, Switzerland) were first given intravenously as a priming bolus at the concentration of 1 and 0.02 mg/min, respectively. The infusion rate was 0.9 ml·100 g−1·h−1. After a 1-h equilibration period, sinistrin and PAH clearances were measured in the plasma to determine glomerular filtration rate (GFR) and effective renal plasma flow (ERPF) according to validated clearance measurements avoiding urine sampling (9). This method has also been validated in our laboratory by comparing it with the conventional method of urine collection. Both methods used for the measurement of clearances in 100 samples were significantly correlated (r2 = 0.64, P < 0.0001). This method has the advantage of infusing PAH and sinistrin at lower rates and for longer durations, thus decreasing the risk of volume expansion. The renal hemodynamic responses to increasing perfusion doses of ANG II (0, 20, 40, and 80 ng·kg−1·min−1) were measured. At each dose modification the speed of perfusion remained the same, but the concentration of ANG II was increased. Each perfusion dose was administered for 30 min, and values of MBP obtained were averaged for the last 10 min of perfusion.
ANG II type 1 receptor binding assay on renal membranes: preparation of membrane.
Renal membranes were produced from the kidneys of several Zucker rats belonging to each treatment regimen according to a protocol already described for mice (35). The number [maximum binding capacity (Bmax)] and the dissociation constant (Kd) of ANG II type 1 (AT1) receptors in the kidneys of animals treated with pioglitazone or not was assessed using saturation curves of radiolabeled ANG II. Thus the binding of ANG II [(5-l-isoleucine)tyrosyl-125I-monoiodinated (125I-ANG II); Dupont, Boston, MA] was investigated as previously published (4). The binding was conducted in a final volume of 0.3 ml of 50 mmol/l Tris·HCl (pH 7.4) containing 5 mmol/l MgCl2 and 0.2% bovine serum albumin at 37°C in the presence of 10 μmol/l PD-123319 to block the AT2 receptors. Separation of bound labeled ANG II was achieved by centrifugation at 4°C and repeated washes with buffer. The residual radioactivity was determined by γ-counting. Nonspecific binding of 125I-ANG II to the membrane preparation was estimated in the presence of 10 μmol/l unlabeled ANG II. Specific binding was defined as total binding minus nonspecific binding. In saturation experiments, membranes (100 mg protein/assay tube) were incubated with various concentrations (final concentration 1–100 nmol/l) of 125I-ANG II for 1 h, a duration that allowed us to reach equilibrium between association and dissociation of the iodinated ANG II (18) to membranes. The experiments were done in duplicates, and the characterization of the binding saturation curves (Kd and Bmax) was obtained by fitting experimental data with one-site binding (hyperbole) equation using GraphPad Prism software, a method that is more accurate than using Scatchard plots, which distort experimental errors. Of note, it was not possible to detect any AT2 binding on the membranes when membranes were incubated with 125I-ANG II in binding buffer containing 10 μmol/l of candesartan to block the AT1 receptors (data not shown).
Analysis of AT1 mRNA expression.
Total RNA was isolated from resected kidneys using the TriPure isolation reagent (Roche Diagnostics, Rotkreuz, Switzerland). First-strand synthesis of cDNA from purified RNAs (2 μg/20 μl) was performed using oligo(dT)15 primers (10 μM; Promega, Madison, WI) and the reverse transcription Omniscript kit (Auf dem Wolf 39; Qiagen, Basel, Switzerland). Resulting cDNAs were subjected to real-time polymerase chain reaction (PCR) using the LightCycler apparatus (Roche Diagnostics) and the Lithos qPCR mastermix (Eurogentec, Seraing, Belgium). Each reaction (20 μl) contained 250 ng of kidney cDNA in 2 μl of Omniscript RT reaction buffer, 16 μl of 1× Lithos reaction mix (containing the Sybr Green I fluorescent dye and adjusted to 2.5 mM MgCl2), and 2 μl of forward and reverse primer mix (0.5 μM each). Cycling conditions consisted of an initial denaturation at 95°C for 30 s followed by 40 cycles of denaturation at 95°C for 0 s, annealing at 60°C for 5 s, and extension at 72°C for 18 s. Primers (all from Microsynth, Balgach, Switzerland) were 5′-AAGCCCATCACCATCTTCCAGGAG-3′ (forward) and 5′-AGCCCTTCCACAATGCCAAAG-3′ (reverse) for GAPDH and 5′-GTCGCACTCAAGCCTGTCTA-3′ (forward) and 5′-CCTGTCACTCCACCTCAGAA-3′ (reverse) for AT1. Amplification products were analyzed by electrophoresis on agarose gels and consisted of one band of expected size (308 bp for GAPDH and 110 bp for AT1). Results are expressed as arbitrary units based on a GAPDH cDNA expression standard curve consisting of GAPDH amplification in a series of four 50-fold dilutions of kidney cDNAs. Data are expressed as ratios of AT1/GAPDH (arbitrary units).
Plasma renin activity was determined as described previously (23). Serial whole blood glucose levels were determined with the Glucometer Elite XL (Bayer). Plasma insulin was determined by radioimmunoassay (Insulin-RIA; Pharmacia, Dübendorf, Switzerland). Creatinine was measured by the picric acid method (Cobas-Mira; Roche Diagnostics). Plasma sinistrin and PAH were determined by photometry (Autoanalyzer II-Technicon; Bran & Luebbe, Norderstedt, Germany).
Data are expressed as means ± SE. The statistical differences between the treatment groups (vehicle, metformin, and pioglitazone) were evaluated by ANOVA (Minitab) followed by Fisher's test for multiple comparisons. A level of P < 0.05 was considered statistically significant.
Body weight, food intake, and salt intake.
Pioglitazone significantly increased food intake in lean (P < 0.05) and obese rats (NS and HS: P < 0.001; Table 1). Average food intake was strongly correlated with the average final body weight (r2 = 0.79, P = 0.008). Metformin-treated rats ate significantly less than glitazone-treated rats and were slightly lighter than vehicle-treated rats. The food intake was not altered by the HS diet. The HS diet slightly increased body weight in pioglitazone (P > 0.05) and vehicle-treated rats (P < 0.05).
Intravenous glucose tolerance tests.
Blood glucose increased similarly in the obese NS rats treated with metformin and vehicle (Fig. 1). Blood glucose was significantly lower with pioglitazone than with vehicle or metformin in the obese NS rats but not in the lean rats.
The increase in insulin levels in the obese rats was blunted with pioglitazone and metformin, and differences were significantly different from vehicle. The insulin resistance index at baseline assessed by HOMA-IR was significantly lower in the metformin (P < 0.02) and pioglitazone (P < 0.001) rats than in the vehicle-treated rats (3.15 ± 0.34, 0.36 ± 0.03, 4.6 ± 0.44, respectively; means ± SE). Pioglitazone-treated rats had lower insulin levels than metformin-treated rats at all times. Again, pioglitazone did not change insulin levels in the lean rats when compared with vehicle-treated rats. Overall, pioglitazone restored insulin and glucose curves to normal lean levels in the obese rats. With the HS diet, insulin levels were significantly lower from 4 to 15 min in the vehicle-treated rats but not different in the pioglitazone-treated rats. Blood glucose levels were similar at all times with both diets. HOMA-IR did not change with the HS diet in vehicle and pioglitazone-treated rats.
Pioglitazone decreased MBP significantly in the obese (P < 0.001) but not in the lean rats (Table 2). BP values achieved with pioglitazone in the obese rats were significantly lower than those measured in the lean control rats (P < 0.01), and the hypotensive effects of pioglitazone were even more pronounced in rats fed a HS diet than in NS rats (difference: P = 0.003). HR decreased significantly with pioglitazone in the obese (P < 0.01) and lean rats (P < 0.05). Again, the decrease in HR was more pronounced in HS than in NS rats (difference: P = 0.001). BP and HR were comparable in metformin- and vehicle-treated rats.
Plasma renin activity and ANG II response.
In control animals, plasma renin activity was significantly lower in obese than in lean rats (Table 2). Plasma renin activity was slightly but not significantly higher in pioglitazone- and metformin-treated than in vehicle-treated obese rats and decreased with the HS diet. The fact that the HS diet was unable to decrease plasma renin activity may be due to the low renin state of the obese Zucker rats and the fact that the sodium provided by the HS diet was only three times higher than the NS diet compared with other studies where it was 20 times higher (21).
The BP response to bolus doses of ANG II was comparable in metformin- and vehicle-treated obese rats but was significantly blunted in pioglitazone-treated obese rats (Fig. 2). In contrast, pioglitazone did not influence the BP response to ANG II in lean rats.
Pioglitazone also blunted the BP response to 30-min infusions of increasing doses of ANG II (Fig. 3). This effect was more pronounced with the HS than the NS diet. As expected, the BP response to exogenous ANG II was amplified in HS control obese rats. However, the salt-induced increase in vascular responsiveness to ANG II was abolished by pioglitazone.
Renal function tests.
Filtration fraction was lower in pioglitazone-treated rats. The HS diet significantly increased GFR and ERPF and reduced filtration fraction in pioglitazone-treated rats but not in vehicle-treated rats (Table 3). The effects of ANG II on renal hemodynamics were studied in five vehicle-treated NS rats and five pioglitazone-treated NS rats. Means ± SE values for 0, 20, 40, and 80 ng·kg−1·min−1 ANG II infusion, respectively, were GFR (ml·100 g−1·min−1) of vehicle treated rats: 0.77 ± 0.12, 0.59 ± 0.05, 0.57 ± 0.05, and 0.64 ± 0.07; pioglitazone-treated rats: 0.37 ± 0.02, 0.35 ± 0.02, 0.33 ± 0.01, and 0.4 ± 0.03; ERPF (ml·100 g−1·min−1) of vehicle-treated rats: 2.9 ± 0.5, 1.6 ± 0.2, 1.5 ± 0.2, and 2.0 ± 0.4; pioglitazone-treated rats: 1.8 ± 0.1, 1.1 ± 0.1, 1.0 ± 0.1, and 2.2 ± 0.5. The effects of ANG II on renal hemodynamics were studied in six vehicle-treated HS rats and five pioglitazone-treated HS rats. Means ± SE values for 0, 20, 40, and 80 ng·kg−1·min−1 ANG II infusion, respectively, were GFR (ml·100 g−1·min−1) of vehicle-treated rats: 0.92 ± 0.06, 0.79 ± 0.05, 0.71 ± 0.05, and 0.96 ± 0.08; pioglitazone-treated rats: 1.06 ± 0.23, 0.61 ± 0.03, 0.58 ± 0.02, and 0.67 ± 0.04; ERPF (ml·100 g−1·min−1) of vehicle-treated rats: 3.8 ± 0.4, 2.5 ± 0.2, 2.1 ± 0.2, and 4.8 ± 0.7; pioglitazone-treated rats: 9.1 ± 2.9, 2.2 ± 0.4, 1.8 ± 0.1, and 3.6 ± 0.5. Pioglitazone blunted the changes in filtration fraction at all doses of ANG II in the NS-treated obese rats (Fig. 4A) but only at the medium dose in HS-treated rats (Fig. 4B).
Renal AT1 receptor binding and expression.
The number (Bmax), Kd, and mRNA expression of AT1 receptors were measured in renal tissue membranes. Pioglitazone increased the Kd in NS rats but not in HS rats. The number of receptors were significantly decreased with pioglitazone compared with vehicle-treated obese rats (P < 0.01; Fig. 5). Accordingly, the renal mRNA expression was reduced more than twofold in pioglitazone-treated rats.
This study shows that the PPARγ agonist pioglitazone favors hyperphagia and weight gain and decreases insulin and serum glucose levels in obese Zucker rats. Despite a higher food and sodium intake and a net gain in weight, basal BP and its response to exogenous ANG II is significantly reduced in obese rats treated with pioglitazone. However, these ANG II antagonistic effects are not found in the lean Zucker rat. Although metformin also improves insulin resistance in the obese model, no ANG II antagonistic effect was found. As expected, the BP response to exogenous ANG II increases in the obese Zucker rats receiving a HS diet. Pioglitazone totally abolishes the salt-induced, enhanced responsiveness independently of any change in insulin sensitivity. Moreover, in the obese rat, pioglitazone modifies the renal hemodynamic response to changes in salt intake while maintaining a lower filtration fraction at baseline and during ANG II perfusion. These effects are associated with a lower number and expression of the ANG II AT1 receptor in the kidneys of pioglitazone-treated rats. Taken together, these results confirm the ANG II antagonistic effects of pioglitazone and demonstrate for the first time that pioglitazone modifies the physiological relation between sodium and ANG II both at the vascular and renal level in obese insulin-resistant rats.
The hypotensive effects of glitazones have already been observed in the obese Zucker rat model (33) and, likewise, the absence of effect in the lean Zucker rat (11, 32). In these studies, the glitazone restored BP to levels found in lean rats. In contrast, in our study, where intra-arterial BP was measured in the awake state, pioglitazone decreases BP to levels significantly lower than in lean rats, suggesting that pioglitazone blunts one or several overactive hypertensive mechanisms characteristic of the obese state. Surprisingly, BP and HR measured under pioglitazone in HS obese rats were significantly lower than in NS rats despite similar improvements in glucose tolerance. An effect of pioglitazone on the sympathetic nervous system could also contribute to the observed BP reduction. Indeed, HR and insulin levels are significantly reduced in pioglitazone-treated rats. Because hyperinsulinemia is a known stimulus of the sympathetic nervous system, lower insulin levels with pioglitazone may be associated with a lower sympathetic nervous system activity (2, 27). Today, several other mechanisms have been proposed to explain the hypotensive properties of glitazones. These include a decrease in endothelin-1 expression (14, 29), an improved endothelial function (10), and blockade of calcium uptake by vascular smooth muscle cells (3). In addition, the ability of glitazones to interfere with the activity of the renin-ANG II system may play a very important role. In vitro, troglitazone reduces the ANG II AT1 receptor expression and the calcium response to ANG II in vascular smooth muscle cells (31). Other in vivo studies demonstrate ANG II antagonizing effects of glitazones (8, 16, 28). Our experiments extend these initial observations and demonstrate that, in addition to the ANG II-antagonizing properties, pioglitazone interferes with the physiological relationship between sodium and ANG II at the vascular level.
Although the definite cellular or molecular mechanisms for the ANG II-antagonizing properties of PPARγ agonists are not known, our data provide some additional clues. This effect is not exclusively linked to the improvement in insulin sensitivity because it was not observed in rats treated with metformin, although one has to acknowledge that the changes in plasma insulin were less with metformin than with pioglitazone. Our data show that pioglitazone interferes with the renin-ANG system at the AT1 receptor level by reducing the expression of the receptor and, hence, its density at the cell surface. Because the whole kidney was used, we could not differentiate between vascular and nonvascular AT1 expression. The primers used in this study could not distinguish the AT1A from the AT1B receptor, and thus we could not examine a specific contribution of pioglitazone on the expression of each of these receptors. Binding studies were performed on the ANG AT2 receptor. However, because the expression was too low, a specific contribution of pioglitazone on the AT2 receptor could not be demonstrated. A direct effect of pioglitazone on the expression of the AT1 receptor had already been demonstrated in vitro, independently of the metabolic effects of pioglitazone (31). Of note, in our study, pioglitazone did not blunt the ANG II effects in lean rats, although it did decrease the expression of the AT1 receptor in kidney membranes obtained from lean rats receiving pioglitazone (data not shown). Thus the effects of pioglitazone on the acute BP response to ANG II in vivo are not exclusively related to a decrease in the AT1 receptor expression but are linked to the particular phenotype of the insulin-resistant obese rat. The obese Zucker rat is known to be a low-renin model with a higher sensitivity to ANG II than the control lean Zucker rat (1).
So far, no study has evaluated the impact of salt intake on the effects of glitazones. Moreover, the impact of glitazones on the renal response to ANG II has never been assessed. The original finding of our experiments is that pioglitazone appears to dissociate the effect of salt on the response to exogenous ANG II at the vascular level. A high-salt diet is known to increase the expression of the AT1 receptors (22). The molecular pathways involved in these effects are not known. Presumably, the increased expression of the AT1 receptor is a compensatory mechanism secondary to the low renin state induced by a high-sodium diet. A moderate increase in serum sodium levels could also contribute to a direct effect on the AT1 receptor expression (7). In our experiments, pioglitazone completely abolished the salt-induced increase in vascular responsiveness to exogenous ANG II independently of any change in plasma Na levels, as they remained the same in pioglitazone and vehicle-treated rats. At the present time, the precise mechanism of this dissociation of the physiological relationship between ANG II and salt is not fully understood, although the direct influence of pioglitazone on ANG II receptor expression may be an important phenomenon (31). At the renal level, we also found that the renal hemodynamic response to exogenous ANG II is modified by pioglitazone, leading to a lower filtration fraction, mainly when rats are fed a high-sodium diet. When exogenous ANG II was infused, an increase in filtration fraction due to the predominant effect of ANG II on the efferent arteriole was found. In glitazone-treated rats, the response to ANG II was blunted, particularly with a normal salt intake. Indeed, these results need to be confirmed by more precise studies of the renal actions of pioglitazone on the ANG II-induced modifications of renal hemodynamics.
These results may have important clinical implications, as large clinical trials show that the inhibition of the renin-angiotensin system either with an angiotensin-converting enzyme or an ANG II antagonist decreases the cardiovascular and renal risk in diabetic subjects (5, 6). Glitazones have already been shown (13, 17) to improve microalbuminuria in animal models of diabetes and in humans to slow the progression of mesangial cell expansion and glomerulosclerosis in the Zucker rat (20, 30) and to prevent the deterioration of renal function in nondiabetic models. The questions of whether the beneficial effects of glitazones are mediated by their ANG II antagonistic properties and whether these are additive to the effects of angiotensin-converting enzyme inhibitors or ANG II receptor antagonist should now be investigated.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 by American Physiological Society