Role of Vascular Smooth Muscle PPARγ in Regulating AT1 Receptor Signaling and Angiotensin II-Dependent Hypertension

Peroxisome proliferator activated receptor γ (PPARγ) has been reported to play a protective role in the vasculature; however, the underlying mechanisms involved are not entirely known. We previously showed that vascular smooth muscle-specific overexpression of a dominant negative human PPARγ mutation in mice (S-P467L) leads to enhanced myogenic tone and increased angiotensin-II-dependent vasoconstriction. S-P467L mice also exhibit increased arterial blood pressure. Here we tested the hypotheses that a) mesenteric smooth muscle cells isolated from S-P467L mice exhibit enhanced angiotensin-II AT1 receptor signaling, and b) the increased arterial pressure of S-P467L mice is angiotensin-II AT1 receptor dependent. Phosphorylation of mitogen-activated protein/extracellular signal-regulated kinase (ERK1/2) was robustly increased in mesenteric artery smooth muscle cell cultures from S-P467L in response to angiotensin-II. The increase in ERK1/2 activation by angiotensin-II was blocked by losartan, a blocker of AT1 receptors. Angiotensin-II-induced ERK1/2 activation was also blocked by Tempol, a scavenger of reactive oxygen species, and correlated with increased Nox4 protein expression. To investigate whether endogenous renin-angiotensin system activity contributes to the elevated arterial pressure in S-P467L, non-transgenic and S-P467L mice were treated with the AT1 receptor blocker, losartan (30 mg/kg per day), for 14-days and arterial pressure was assessed by radiotelemetry. At baseline S-P467L mice showed a significant increase of systolic arterial pressure (142.0±10.2 vs 129.1±3.0 mmHg, p<0.05). Treatment with losartan lowered systolic arterial pressure in S-P467L (132.2±6.9 mmHg) to a level similar to untreated non-transgenic mice. Losartan also lowered arterial pressure in non-transgenic (113.0±3.9 mmHg) mice, such that there was no difference in the losartan-induced depressor response between groups (−13.53±1.39 in S-P467L vs −16.16±3.14 mmHg in non-transgenic). Our results suggest that interference with PPARγ in smooth muscle: a) causes enhanced angiotensin-II AT1 receptor-mediated ERK1/2 activation in resistance vessels, b) and may elevate arterial pressure through both angiotensin-II AT1 receptor-dependent and -independent mechanisms.


Introduction
Peroxisome Proliferator-Activated Receptor c (PPARc) is a ligand-activated transcription factor belonging to the nuclear receptor superfamily which is well-known for its role in adipogenesis, lipid metabolism and glucose homeostasis (reviewed in [1]). Thiazolidinediones (TZD) are high affinity synthetic agonists for PPARc that improve glycemic control in type 2 diabetes, and have been reported to improve vascular function and to lower arterial pressure [2] (reviewed in [3]). Mutations in PPARc cause lipodystrophies, and in some severe cases, causes early onset insulin resistance, type 2 diabetes and hypertension [4,5]. Mice carrying the equivalent mutation in endogenous PPARc exhibit hypertension [6,7] and when bred onto the leptindeficient genetic background exhibit insulin resistance and metabolic dysfunction [8].
Genetic knock-down of PPARc in animals confirmed a protective role of PPARc in both endothelium and vascular muscle, however, the molecular mechanisms by which PPARc regulates blood pressure and improves vascular function has not been completely elucidated [9][10][11][12][13]. To address this shortcoming, we developed novel models expressing dominant negative mutations in PPARc, the same mutations that cause hypertension in human patients, selectively in endothelium and vascular smooth muscle [7,[14][15][16][17]. We validated that these mutations cause a gene expression profile opposite that of a PPARc agonist consistent with their dominant negative action [18,19].
In endothelium, interference with PPARc caused endothelial dysfunction in response to a high fat diet which was improved by scavenging reactive oxygen species (ROS) [14]. Our data were consistent with the hypothesis that high fat diet induced the production of a PPARc ligand which activates a protective anti-oxidant gene expression program in the endothelium. This protective mechanism cannot be induced in the presence of dominant negative PPARc mutants. In vascular muscle, PPARc interference causes hypertension under baseline conditions and dysfunction of both conduit and resistance blood vessels [15]. Thus, these mice faithfully phenocopy the hypertension portion of the overall phenotype exhibited by patients with identical PPARc mutations. Mechanistically, aorta from mice with smooth musclespecific expression of dominant negative PPARc (S-P467L) are resistant to endogenous and exogenous nitric oxide (NO) which is reversed by Rho kinase blockade [17]. Second order mesenteric branches from these mice exhibit increased myogenic tone which is reversed by protein kinase C (PKC) inhibition [16]. They also exhibit augmented angiotensin-II (Ang-II) mediated vasoconstriction because of decreased expression of the PPARc target gene RGS5. The present study aimed to determine whether activation of Ang-II signaling in vascular smooth muscle cells (SMC) from mesenteric arteries contributes to hypertension in S-P467L mice.

Animals
Transgenic mice were generated by expressing the protein coding sequence of human PPARc carrying the P467L mutation under the control of the smooth muscle myosin heavy chain promoter (S-P467L) as described previously [15,17]. Non-transgenic (NT) littermates were utilized as controls. All mice were continuously backcross bred to C57BL/6J mice. Mice were maintained in a 12 hour light-dark cycle (6 AM to 6 PM), fed normal rodent chow (Teklad 7013), and had access to water ad libitum. In vivo and in vitro experiments were performed using male mice that were four to five months of age. All of the experimental protocols were approved by the University of Iowa Animal Care and Use Committee and conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Primary Mesenteric SMC Culture
Vascular SMCs were isolated from second and third order mesenteric arteries of NT-littermates and S-P467L mice by the explant method [20]. Briefly, mesenteric beds were cleaned of adipose and connective tissue and placed in the culture dish and maintained with Dulbecco Modified Eagle's Medium (DMEM) containing 10% fetal bovine serum (FBS) and antibiotics and maintained in an incubator at 37uC in a humidified 5% CO 2 atmosphere. After 72h, once VSMCs migrated from the edges of the explants, the mesenteric arteries were removed and discarded. Mesenteric SMCs exhibited the typical ''hill and valley'' growth morphology and were confirmed positive (.95%) for smooth muscle a-actin (Sigma) and calponin (Santa Cruz sc-28545) immunostaining, and negative for Pecam-1 (Abcam ab28364) band CD90/Thy1 (Santa Cruz sc9163). Cells at passage 3 maintain their contractile phenotype and were used in all experiments. At 80% confluence, the culture medium was replaced with serum-free medium for 24 hours to render the cells quiescent. Cells were then treated with Ang-II (0.01 mM) for 2, 5, 10, 30 and 60 minutes for determination of ERK1/2 activation and for 24, 48 and 72 hours for analysis of proliferation (as measured by PCNA expression). In some experiments mesenteric SMC were pre-incubated with losartan (1 mM) or Tempol (5 mM) for 30 min following treatment with Ang-II.

Gene Expression
NADPH oxidase (Nox) gene expression was evaluated from a microarray dataset of mesenteric artery of S-P467L and control mice previously reported by us (NCBI Accession Number GSE36482) [16]. Total RNA was isolated from mesenteric artery samples using Trizol Reagent (Invitrogen Life Technologies Co., Burlington, ON) and RNeasy spin columns (QIAGEN), and was quantified by Nanodrop. cDNA was synthesized from 400 ng of total RNA by RT-PCR using SuperScript III (Invitrogen). Quantitative real-time PCR (qPCR) was performed using Taq-Man Fast Advanced Master Mix (Applied Biosystems), TaqMan gene expression assays (Nox4: Mn00479246_m1) and 10 ng of cDNA. DDCT was calculated using GAPDH as the reference gene to determine relative mRNA expression.

Blood Pressure and Heart Rate Measurements
Blood pressure and heart rate were measured by radiotelemetry as previously described in male mice that were four to five months of age [21,22]. Under anesthesia with ketamine/xylazine (85.5 mg/kg and 12.5 mg/kg, respectively), a radiotelemetry probe (TA11PA-C10, Data Sciences International) was inserted in the left common carotid artery and the body of the transmitter was secured subcutaneously in the abdomen. Normal body temperature was maintained with a heating pad during intraoperative and postoperative care. After a 7-day recovery from transmitter implantation, baseline measurements were recorded for 10 seconds every 5 minutes over a 7-day period. Subsequently, mice were treated with losartan (30 mg/kg/day for 14 days) using ALZET osmotic mini-pumps (model 1002; Durect Corporation, Cupertino, CA) placed subcutaneously. Blood pressure and heart rate measurements were recorded over the last 5 days of losartan/ vehicle control treatment and the data from the final 2 days is presented. Control mice received saline by osmotic mini-pumps following the time-course of losartan treatment.

Statistical Analysis
Results are expressed as Mean6SEM and analyzed with 1 or 2way repeated measures ANOVA followed by Tukey's or Bonferroni multiple-comparison procedures. Student's t-test was used when appropriate. P,0.05 was considered statistically significant. SigmaStat was used for all analytical comparisons.

Results
We first characterized mesenteric SMC primary cultures from S-P467L. The cultured cells were confirmed to be vascular smooth muscle by positive staining with a-actin and calponin, typical markers of differentiated contractile vascular SMCs (Figure 1). We also established the purity of these cultures by the absence of staining for Pecam-1 and CD90/Thy1, markers for endothelial cells. The apparent decrease in calponin immunostaining in cultured cells from S-P467L mice was consistent with a decrease in calponin expression in mesenteric arteries from those mice ( Figure 2). The decrease in a classical vascular SMC marker suggested the cells may have undergone a more rapid change from a contractile phenotype to a synthetic or proliferative phenotype. Consistent with this, mesenteric SMCs from S-P467L showed increased expression of proliferating cell nuclear antigen (PCNA) protein compared to mesenteric SMC from control non-transgenic (NT) mice under basal conditions ( Figure 3A). In addition, stimulation with Ang-II (0.01 mM) for 72 hours increased PCNA protein expression by approximately 2-fold in vascular SMCs from S-P467L, but not from NT control mice ( Figure 3B).
We next tested the hypothesis that Ang-II-treated mesenteric SMC from S-P467L mice have augmented ERK1/2 activation. At baseline, there was a trend for increased activation of ERK1/2 in mesenteric SMCs from S-P467L compared to NT control cells (Figure 4). Mesenteric SMCs from S-P467L showed a greater phosphorylation of ERK1/2 than NT control cells in response to Ang-II. Ang II-induced ERK1/2 phosphorylation was maintained for 60 min in mesenteric SMCs from S-P467L, while 10 minutes was the maximum period of ERK1/2 phosphorylation in mesenteric SMCs from NT mice. There was a slight decrease in baseline phosphorylated JNK without a change in total JNK, but no increase in phosphorylated JNK in response to Ang-II (data not shown). The increase in ERK1/2 activation by Ang-II was AT 1 receptor-mediated as ERK1/2 phosphorylation could be blocked by losartan ( Figure 5A). The Ang-II-mediated activation of ERK1/2 in vascular SMCs from S-P467L was abolished in the presence of Tempol, a scavenger of superoxide suggesting this is redox mediated ( Figure 5B). Examination of microarray data of mesenteric artery RNA from S-P467L and control mice revealed no changes in the expression of Nox1 or Nox3 [16]. Expression of Nox2 was increased by 20%, whereas expression of Nox4 was decreased by 50%. Quantitative RT-PCR confirmed decreased expression of Nox4 in mesenteric artery of S-P467L (to 0.4560.08 of control levels, P,0.02). Despite a decrease in Nox4 mRNA, there was evidence of increased Nox4 protein in mesenteric vessels from S-P467L mice ( Figure 5C).
We previously reported that S-P467L mice exhibit a pressor response to acute Ang-II infusion that is similar to control mice [16]. We therefore tested whether augmented mesenteric SMC Ang-II signaling contributes to the chronically elevated blood pressure in S-P467L mice. As we showed previously, S-P467L mice exhibited increased baseline arterial pressure compared with littermate controls (Figure 6A, C). We treated mice with losartan to determine if endogenous RAS signaling contributes to the increased arterial pressure. Losartan significantly lowered arterial pressure in S-P467L mice ( Figure 6B-D) to the level observed in untreated littermate controls. Losartan also lowered arterial pressure in control mice, but the change in arterial pressure in response to losartan was equal in both groups (S-P467L: 2 17.2612.9 mmHg; NT: 216.464.2 mmHg), although the variability in the S-P467L group was larger. Interestingly, whereas the heart rate in control mice modestly declined in response to losartan (211.5615.6 beats/min), this did not occur in S-P467L mice (+1.769.2 beats/min).

Discussion
Studies in animals and human subjects have convincingly demonstrated that PPARc plays a crucial role in the regulation of vascular tone and control of arterial blood pressure [6,23,24]. The   strongest evidence supporting a role for PPARc in arterial pressure regulation comes from genetic studies where patients bearing rare dominant negative mutations in PPARc exhibit severe early onset hypertension [4]. This is further supported by a recent report identifying two new mutations in PPARc that are associated with increased arterial pressure due to aberrant activation of the reninangiotensin system [5,25]. The link between PPARc and the renin-angiotensin system is further supported by our studies showing that mice expressing the P467L dominant negative mutation in PPARc specifically in vascular smooth muscle (S-P467L mice) caused increased contraction of mesenteric artery to Ang-II [16]. Thus, the rationale for the current study was to explore the role of Ang-II-dependent signaling and arterial pressure regulation. The main findings of the current study are that 1) mesenteric SMC from S-P467L mice exhibit increased Ang-II ERK1/2 signaling through an AT 1 receptor-dependent and oxidant stress-dependent mechanism, and 2) the increased baseline blood pressure observed in S-P467L mice is lowered to normal levels by short-term AT 1 receptor blockade.
Mesenteric SMC from S-P467L mice exhibited increased proliferation, as measured by increased expression of PCNA, a marker of proliferation, and decreased expression of calponin, a contractile marker. These data suggest that the mesenteric SMC from S-P467L mice may be undergoing phenotypic modulation, that is, a shift from a contractile state to a proliferative state. It has been reported that vascular SMC PPARc prevents the switch from a contractile to a proliferative state during vascular injury and hypertension [26,27]. Similarly, SMC derived from mice carrying a gene targeted mutation in PPARc equivalent to the P467L mutation employed herein exhibited increased proliferation [28]. It is therefore not surprising that Ang-II further increased PCNA expression in mesenteric SMC from S-P467L. These data suggest that loss of PPARc function in vascular SMC promotes a proliferative phenotype. This is supported by studies showing that PPARc has a vascular protective effect in Ang-II-induced vascular remodeling [13,23].
Mesenteric SMC from S-P467L mice exhibited augmented responsiveness to Ang-II as measured by a rapid and potent induction of ERK1/2 phosphorylation. This effect was abolished in the presence of Tempol, suggesting it is mediated by increased oxidative stress. It is now well accepted that NADPH oxidasederived ROS is one of the main mediators of Ang-II-induced vascular dysfunction [29,30]. We evaluated the expression of NADPH oxidase isoforms in our model and Nox4 was the only one found to be significantly altered. Although Nox4 mRNA levels were decreased in mesenteric arteries from S-P467L mice, Nox4  protein was markedly increased. Interestingly, increased Nox4 was shown to correlate with decreased expression of PPARc in human pulmonary artery endothelial and smooth muscle cells, and rosiglitazone, a potent PPARc agonist blunts the increase in Nox4 expression induced by hypoxia in mouse lung [31]. Hydrogen peroxide, a product of Nox4 activity, has been demonstrated to act as a proliferative factor in VSMC [32], and decreases PPARc expression and activity in endothelial cells [33]. At face value, these results suggest that Nox4 may be a mediator of increased proliferation and AT 1 receptor signaling in SMC from S-P467L mice. However, that vascular dysfunction and increased agonist-induced contraction in aorta, and increased myogenic tone in mesenteric artery was not reversed by Tempol in vivo makes it unlikely that Nox4 mediates the vascular dysfunction or hypertension observed in S-P467L mice [16,17]. Of course, we cannot rule out that Nox4 may have a local effect in SMC that indirectly contributes over the long term to vascular dysfunction in S-P467L. Further studies would be needed to clarify its function. Future studies should also examine the importance of inflammation. There is a reciprocal relationship between PPARc and AT 1 receptors. Just as PPARc regulates AT 1 receptor expression, Ang-II also down-regulates PPARc in aorta, and this promotes inflammation [34]. We have previously reported that S-P467L mice exhibit increased expression of pro-inflammatory and proatherogenic genes particularly when bred on the ApoE-deficient genetic background [35].
We considered two possible mechanisms accounting for the increase in AT 1 receptor signaling. The first is that PPARc interference caused increased expression of AT 1 receptor mRNA. In support of this, PPARc activation decreases expression of the Ang-II AT 1 receptor in smooth muscle [36] and cardiac fibroblasts [37], and PPARc mutations increase expression of genes in the renin-angiotensin system including renin, angiotensinogen and AT 1 receptor [5]. However, we reported that AT 1 receptor mRNA was unchanged in mesenteric artery of S-P467L mice [16]. Alternatively, PPARc-interference can increase activity of the AT 1 receptor signaling complex. PPARc activation represses Ang-II signaling in IgA nephropathy [38] and attenuates Ang-II-induced inflammation in vascular smooth muscle cells [39]. We previously showed that PPARc-interference results in decreased expression of regulator of G protein signaling 5 (RGS5) mRNA in mesenteric artery from S-P467L mice [16]. RGS5 is a novel PPARc target gene, and it's decreased expression was responsible for augmented myogenic tone and Ang-II-mediated contraction of mesenteric artery. Consequently, the available evidence suggests that interference with PPARc increases AT 1 receptor signaling as a result of increased post-receptor activation and not due to increased expression of AT 1 receptor mRNA.
PPARc activation lowers blood pressure in models of Ang-II infusion [23], and in Ang-II-dependent hypertensive transgenic mice [24]. To evaluate whether augmented Ang-II signaling in vascular SMC contributes to the hypertensive phenotype found in S-P467L mice, we treated S-P467L mice with the AT 1 receptor antagonist losartan. Losartan lowered blood pressure in S-P467L to the level observed in untreated NT control mice. The fact that blood pressure was lowered to control levels by losartan would be consistent with the observation that AT 1 receptor blockers were said to be effective antihypertensives in patients with PPARc mutations leading to activation of the renin-angiotensin system [5]. These patients carry PPARc mutations in all cells whereas in S-P467L, the mutation is restricted to vascular smooth muscle. This suggests that the loss of PPARc signaling in smooth muscle may be sufficient to phenocopy the hypertension that results from the presence of the mutation in all cells in human patients. That losartan normalizes blood pressure in S-P467L mice would also be consistent with our observation that there is a large increase in contraction of mesenteric arteries to Ang-II [16]. However, it is important to recognize that the losartan-induced depressor response was similar in S-P467L and in control mice. This observation would be consistent with other data showing that acute Ang-II infusion equivalently increased blood pressure in S-P467L and control mice [16]. We cannot rule out that the precise mechanisms of the losartan-induced depressor response is different in S-P467L mice, and that S-P467L and NT mice may exhibit subtly different Ang-II-dependent mechanisms regulating arterial pressure. This may be evidenced by the differential effects of losartan on heart rate. NT mice exhibited a modest bradycardia whereas there was no change in heart rate in S-P467L. We have recently reported that S-P467L exhibit autonomic dysfunction which as we show here may be preserved after AT 1 receptor blockade [40]. We also cannot rule out that other mechanisms serve to chronically regulate blood pressure in the S-P467L model. Indeed, we previously reported that unlike mesenteric arteries, contraction of the aorta is augmented in response to RhoA/Rho kinase agonists and that arterial blood pressure in S-P467L mice is markedly decreased after infusion of a Rho kinase inhibitor [17].