Peroxisome proliferator-activated receptor (PPAR)-γ is reduced in pulmonary arteries (PAs) of patients with PA hypertension (PAH), and we reported that deletion of PPARγ in smooth muscle cells (SMCs) of transgenic mice results in PAH. However, the sequelae of loss of PPARγ in PA endothelial cells (ECs) are unknown. Therefore, we bred Tie2-Cre mice with PPARγflox/flox mice to induce EC loss of PPARγ (Tie2 PPARγ−/−), and we assessed PAH by right ventricular systolic pressure (RVSP), RV hypertrophy (RVH), and muscularized distal PAs in room air (RA), after chronic hypoxia (CH), and after 4 wk of recovery in RA (Rec-RA). The Tie2 PPARγ−/− mice developed spontaneous PAH in RA with increased RVSP, RVH, and muscularized PAs vs. wild type (WT); both genotypes exhibited a similar degree of PAH following chronic hypoxia, but Tie2 PPARγ−/− mice had more residual PAH compared with WT mice after Rec-RA. The Tie2 PPARγ−/− vs. WT mice in RA had increased platelet-derived growth factor receptor-β (PDGF-Rβ) expression and signaling, despite an elevation in the PPARγ target apolipoprotein E, an inhibitor of PDGF signaling. Inhibition of PDGF-Rβ signaling with imatinib, however, was sufficient to reverse the PAH observed in the Tie2 PPARγ−/− mice. Thus the disruption of PPARγ signaling in EC is sufficient to cause mild PAH and to impair recovery from CH-induced PAH. Inhibition of heightened PDGF-Rβ signaling is sufficient to reverse PAH in this genetic model.
- endothelial cells
- platelet-derived growth factor receptor-β
- pulmonary remodeling
- smooth muscle cell
- platelet-derived growth factor
the peroxisome proliferator-activated receptors (PPARs) belong to a family of three nuclear proteins [α, β(δ), and γ] that have tissue-specific overlapping functions as transcription factors (3, 6, 41, 43). PPARγ is highly expressed in normal pulmonary artery (PA) endothelial cells (ECs) but is appreciably reduced in both experimental and clinical pulmonary arterial hypertension (PAH; Refs. 1, 38). PPARγ has antiproliferative and proapoptotic effects in systemic arterial smooth muscle cells (SMCs; Refs. 9, 29), consistent with its function as a tumor suppressor gene (4, 34).
We previously created a mouse in which PPARγ was deleted in vascular SMCs by breeding a mouse expressing Cre downstream of the SM22α promoter with a mouse in which the critical exons of PPARγ were flanked by LoxP sites. The SM22α PPARγ−/− mice developed PAH as judged by elevated right ventricular systolic pressure (RVSP), right ventricular hypertrophy (RVH), and muscularization of distal vessels (10). We then showed that the mechanism is related to reduced levels of a target of PPARγ, apolipoprotein (apo)E, a molecule that inhibits PDGF-Rβ function by binding to low-density receptor-like protein-1 and by preventing the activation of the PDGF-Rβ (2, 14, 26, 47). We further established the relevance of this model to the pathology of clinical PAH by showing that bone morphogenetic protein receptor-II (BMPR-II) ligands modulate PPARγ activation and apoE production (10) and inhibit PDGF-BB-mediated pulmonary artery smooth muscle cells (PASMCs) proliferation. Moreover, dysfunction of BMPR-II, as is seen in familial and sporadic idiopathic PAH (19), causes loss of BMP-mediated inhibition of PDGF-BB-induced PASMC proliferation, a feature that can be rescued with a PPARγ agonist (10).
Recently, transgenic mice were created with a selective loss of PPARγ in ECs and some hematopoietic lineages by breeding a mouse expressing Cre under the regulation of the Tie2 promoter (17) with the PPARγflox/flox mouse (12). These Tie2 PPARγ−/− mice have systemic hypertension but only when fed a high-salt diet (27). However, the sequelae of loss of PPARγ in ECs on pulmonary hemodynamics and vascular remodeling are unknown. In view of our previous study (10) in mice in which PPARγ was deleted in vascular SMCs and because PPARγ is reduced in both ECs and SMCs of PAH patients (1), we hypothesized that Tie2 PPARγ−/− mice would also develop PAH and the associated vascular changes in room air and would have more severe PAH than wild-type (WT) mice following exposure to chronic hypoxia and after recovery in normoxia.
MATERIALS AND METHODS
Production of Tie2 PPARγ Transgenic Mice
To generate mice deficient in PPARγ in ECs (Tie2 PPARγ −/−), mice expressing Cre recombinase under the control of the Tie2 promoter (17) were bred with mice containing the critical exons of PPARγ flanked by LoxP sites (12) obtained from Jackson Laboratories (Bar Harbor, ME). All the mice were Cre gene positive. Tie2 PPARγ−/+ males and females were bred to derive Tie2 PPARγ+/+, Tie2 PPARγ−/+, and Tie2 PPARγ−/− littermates for initial analyses and for further crossbreeding. Tie2 PPARγ−/− and littermate WT mice were used (n = 10/group of each genotype 5 males and 5 females). The mice were either studied in room air at 12 wk of age, or were exposed to hypoxia (10% oxygen) for 3 wk between 8–11 wk of age, or were exposed to hypoxia and then allowed a period of room air “recovery” for 4 wk as previously described (8). All the experimental protocols used in this study were approved by the Institutional Animal Care Committee at Stanford University and adhered to the published guidelines of the National Institutes of Health and the American Physiological Society.
Hemodynamic Measurements and Assessment of Pulmonary Vascular Changes
RVSP, RV dp/dt, and heart rate were measured in unventilated mice that were under isoflurane anaesthesia (1.5–2.5%, 2 l O2/min.) using a closed chest technique, by introducing a 1.4-F Millar catheter into the jugular vein and directing it to the right ventricle (46). Systemic blood pressure was determined in conscious animals by a noninvasive computerized tail-cuff method and verified by catheterization of the left carotid artery while the animals were under isoflurane anaesthesia. Left ventricular fractional shortening and cardiac output were evaluated by echocardiography (8).
After the hemodynamic assessments were completed, heparinized blood was collected by direct cardiac puncture to analyze lipid profiles and to assess general metabolic effects of PPARγ deletion in ECs or other Tie2-expressing cells, as described in the supplemental data for this article that are available online at the Am J Physiol Lung Cell Mol Physiol website. The mice were killed by exsanguination. The heart and lungs were then removed en bloc, and RVH was later evaluated by the Fulton index, i.e., weight of right ventricle/left ventricle plus septum (RV/LV+S; Ref. 22). The pulmonary circulation was flushed with 3 ml of PBS at 37°C, and the lungs were prepared for morphometric analyses by barium gelatin injection of the pulmonary circulation and formalin inflation-fixation of the lung as described in the supplemental data. Alternatively, the right lung was quickly harvested, immediately snap-frozen in liquid nitrogen, and kept at −80°C for Western immunoblot analysis and total RNA extraction, and the left lung was prepared as described in materials and methods in the data supplement.
Morphometric analyses were performed on paraffin-embedded lung sections stained using elastic van Gieson or Movat pentachrome stains. The total number of peripheral arteries was calculated as a ratio of the number of arteries per 100 alveoli in each of 5–6 different ×20 microscopic fields per section from each lung. Muscularization was assessed in 15 higher magnification fields/per mouse (×40 magnification) by calculating the proportion of fully and partially muscularized peripheral (alveolar duct and wall) pulmonary arteries to total peripheral PAs. All morphometric analyses were performed by one observer, blinded as to genotype and condition, i.e., room air, hypoxia, and recovery.
Real-Time Quantitative RT-PCR
Total RNA was isolated from frozen lungs or cultured cells using Trizol (Invitrogen, Carlsbad, CA) and RNeasy mini kit (Qiagen, Valencia, CA). Total RNA (2 μg) was reverse-transcribed using Superscript II (Invitrogen, Carlsbad, CA) per manufacturer's instructions. Gene expression levels of PPARγ were quantified using preverified Assays-on-Demand TaqMan primer/probe sets (Applied Biosystems, Foster City, CA) and normalized to 18S ribosomal RNA using the comparative Ct method.
Frozen tissue was homogenized, or cultured cells were lysed in ice-cold RIPA buffer containing protease and phosphatase inhibitors. Protein concentration was determined using protein assay reagent (Bio-Rad, Richmond, CA), and 50 μg of protein extracts were resolved on 4–12% NuPage Bis-Tris gels and electrotransferred to PVDF membranes. The PVDF membranes were incubated in blocking buffer for 1 h at room temperature and then incubated overnight at 4°C with primary antibodies for PPARγ (1:500; Santa Cruz Biotechnology, Santa Cruz, CA), apoE (1:500; Abcam, Cambridge, MA), phosphoERK1/2 (1:2,000; Cell Signaling Technology, Beverly, MA), total ERK1/2 (1:1,000; Cell Signaling Technology), caveolin-1 (1/1,000; BD Biosciences, San Jose, CA), PDGF-B, (Santa Cruz Biotechnology), phosphoPDGF-Rβ (Tyr751; 1:500; Cell Signaling Technology), and PDGF-Rβ (1:4,000; Upstate Biotechnology, Lake Placid, NY). The membranes were then washed and incubated with either sheep anti-mouse (1:1,000) or donkey anti-rabbit (1:1,000) horseradish peroxidase-conjugated secondary antibodies. Autoradiographs were developed using the ECL kit (antibodies and kit from Amersham Biosciences, GE Healthcare UK Limited, Bucks, UK). Equal loading of protein was confirmed by blotting for β-actin. Relative band strength was determined by densitometry.
Immunohistochemical staining was performed following antigen retrieval on sections from lungs that were perfused with saline to remove blood and then fixed by perfusion with 10% formalin and embedded in paraffin. The antibodies used and their dilutions were anti-PPARγ (1:200, BD Biosciences) and anti-PDGF-Rβ (1:500, Upstate Biotechnology). Labeling was detected with the use of the Vectastain ABC system (Vector Laboratories, Burlingame, CA) following the manufacturer's protocol, and all sections were counterstained with hematoxylin (Sigma-Aldrich, Mountain View, CA).
Nitric Oxide Synthase Activity and Endothelin-1 Assays
Nitric oxide synthase activity was estimated by the method of Yui et al. (44) in which the stable end-products nitrate/nitrite were estimated using the nitric oxide synthase assay kit (catalog no. 482702; Calbiochem) following the manufacturer's instructions. Plasma endothelin-1 (ET-1) was assessed by ELISA as previously described (10, 42).
Cultured human PASMCs.
The effects of recombinant apoE protein on PDGF-Rβ expression were studied in human PASMCs obtained from Lonza (Walkersville, MD) and cultured according to the manufacturer's recommendation. Growth synchronized cells (following 48-h starvation in media supplemented with 0.1% FBS) were exposed to recombinant apoE (1, 5, and 10 μg/ml) in serum-free medium for 24 h before protein extraction, and the expression of PDGF-Rβ was determined by Western immunoblot as previously described (11).
Cultured murine pulmonary ECs.
To determine the effect of loss of PPARγ in the Tie2-expressing cells, the apoE expression was assessed in cultured pulmonary ECs (PECs) isolated from the Tie2 PPARγ and WT mice. Murine PECs were isolated by digesting whole lung tissue with collagenase IA (0.5 mg/ml) for 45 min at 37°C. The cell suspension was filtered through 70-μm filters and then centrifuged at 250 g for 5 min. The cell pellet was then washed and resuspended, and ECs were selectively cultured after incubation with magnetic beads (Dynabeads; Invitrogen, Carlsbad, CA) coated with anti-CD-31 antibody (11). Characterization of the cell culture after isolation was performed by labeling with Dil-conjugated Ac-LDL (Dil-Ac-LDL). Total RNA was extracted with TRIzol reagent from EC harvested from WT and Tie2 PPARγ−/− mice, cDNA was synthesized, and gene expression of PPARγ and apoE was determined by quantitative RT-PCR using TaqMan gene expression assays (Applied Biosystems). Total protein was harvested from pooled samples of WT and Tie2 PPARγ−/− mice PEC and subjected to Western immunoblot to determine the expression of PPARγ protein. In some experiments, WT murine PECs were maintained for 24 h at 1% O2 in a hypoxia chamber (Billups-Rothenberg, Del Mar, CA), permitting monitoring of O2 at 1% and CO2 in room air range, or they were kept in room air.
The number of animals used in each determination is given in Results (see Figs. 1⇓⇓⇓–5). To determine the differences between the Tie2 PPARγ−/− and WT mice, unpaired Student t-tests were performed. One-way ANOVA followed by Bonferroni's post hoc analysis was used to assess changes in a given genotype related to hypoxia and recovery vs. room air. A P value of 0.05 was considered statistically significant.
Decreased PPARγ Protein in PECs Isolated from the Tie2-PPARγ−/− Compared with WT Mice
Deletion of PPARγ in Tie2-expressing cells resulted in a loss of PPARγ mRNA and protein expressed by the PECs isolated from the Tie2-PPARγ−/− vs. WT mice by quantitative RT-PCR (Fig. 1A) and by Western immunoblot (Fig. 1B). In addition, immunohistochemistry was performed on lung sections of Tie2-PPARγ−/− and WT mice and demonstrated a similar PPARγ immunostaining in the PASMCs from both the Tie2-PPARγ−/− and WT mice but an absence of immunoreactivity in the ECs of the transgenic mice and in alveolar macrophages, cells that also express Tie2 (17, 24). Our PPARγ immunohistochemical studies showed no significant difference in the PPARγ expression in PASMCs contained in the distal pulmonary artery walls in the Tie2 PPARγ−/− mice (Fig. 1C).
Tie2 PPARγ−/− Mice Develop Mild PAH at Baseline and Impaired Recovery from Hypoxia-Induced PAH
An increase in RVSP (P < 0.001) was observed in Tie2 PPARγ−/− compared with WT mice in room air, but a similar degree of RVSP elevation was induced in both genotypes by chronic hypoxia. However, following 4 wk of recovery in room air, the Tie2-PPARγ−/− mice had higher residual RVSP than WT controls (28.5 ± 0.4 vs. 26.0 ± 0.3 mmHg, respectively; n = 10 in each group; P < 0.001; Fig. 2A). There were no significant differences between genotypes in heart rate, cardiac output, or left ventricular function (Table 1), although the Tie2 PPARγ−/− mice did have higher triglycerides and triglyceride/HDL levels than WT mice (Supplemental Table S1). No difference was observed in hematocrit values between genotypes (38.6 ± 1.4 vs. 40.5 ± 2.6, respectively; n = 5 in each group; NS). Since PPARγ suppresses asymmetric dimethylarginine and ET-1 (21, 23, 33, 37, 39), we measured nitric oxide production and ET-1 levels but no significant differences were observed between the genotypes (Supplemental Table S2).
Consistent with the RVSP values, greater RVH (RV/LV+S weight ratio) was found in the Tie2 PPARγ−/− vs. WT mice (P < 0.05) in room air. A similar degree of RVH in both genotypes after chronic hypoxia, and persistent RVH in the Tie2 PPARγ−/− mice compared with WT (P < 0.05) following 4 wk of recovery in room air (Fig. 2B). The elevation in RVSP and RVH in the Tie2 PPARγ−/− mice at baseline were also associated with an increased number of muscularized distal pulmonary arteries vs. WT mice (P < 0.001; Fig. 2, C and D). Correlating with the similar elevation in RVSP and RVH following chronic hypoxia, there was a comparable increase in the percentage of muscularized distal arteries in both genotypes. The numbers of peripheral alveolar duct and wall arteries calculated relative to 100 alveoli in the two genotypes were not different in room air, in response to chronic hypoxia, and following recovery in room air (Table 1). For these, and all other assessments, no gender-related differences were observed, and each determination is based on similar numbers of male and female mice.
Several studies support the hypothesis that hypoxia induces changes in PPARγ gene expression in different cell types. Recent findings by Nisbet et al. (28) show that in vitro hypoxia exposure (1% O2 for 72 h) significantly reduced PPARγ protein levels in both isolated human pulmonary artery ECs and smooth muscle cells. In addition, Li et al. (20) reported that compared with human proximal renal tubular epithelial cells under normoxic condition, amounts of PPARγ mRNA were significantly decreased by 57% at 24 h and by 80% at 48 h in human proximal renal tubular epithelial cells under hypoxia (1% O2). Similarly Yun et al. (45) reported in 3T3-L1 preadipocytes that hypoxia inhibits the induction of PPARγ2 mRNA by Northern blot (45). Therefore, to investigate whether the similar degree of PAH observed between both genotypes following chronic hypoxia could be explained by reduced endothelial PPARγ expression in WT mice, we performed additional experiments. First, we used Western immunoblots to assess the PPARγ protein levels in lung homogenates from normoxia and chronic hypoxia exposed WT mice but found no significant differences (n = 3 in each group; P = NS; see Supplemental Fig. S1). Then, we assessed the effect of hypoxia (1% O2 for 24 h) vs. normoxia on the PPARγ protein level in murine PECs isolated from WT mice. Consistent with our in vivo results, we found no difference in the PPARγ protein level in treated or untreated cultured PECs (n = 3 in each group; P = NS; see Supplemental Fig. S2).
Tie2 PPARγ−/− Mice Have Increased Expression and Activation of PDGF-Rβ
Exaggerated PDGF-Rβ-mediated signaling appears to be a key factor in the PASMC proliferation associated with PAH, including that observed in the SM22 PPARγ−/− mice (10, 30, 35). We therefore investigated whether the increased muscularity developing in the Tie2 PPARγ−/− mice was the result of increased PDGF signaling by measuring lung proteins levels of PDGF-B and the PDGF-Rβ in the Tie2 PPARγ−/− and WT mice. We found that while the level of PDGF-B was similar in both genotypes, the expression of the receptor PDGF-Rβ was increased by approximately twofold in the lungs of the Tie2 PPARγ−/− vs. WT mice in room air. In keeping with the hemodynamic results, similar increases in PDGF-Rβ were seen in both groups after chronic hypoxia and after recovery in room air (Fig. 3, A and B). Consistent with this observation, more intense immunoreactivity was noted for PDGF-Rβ in the distal pulmonary artery walls from Tie2 PPARγ−/− vs. WT mice by immunohistochemistry (Fig. 3C).
Tie2 PPARγ−/− Mice Have Increased Expression of PDGF-Rβ Despite Increased Expression of ApoE
In looking for a mechanism by which disruption of PPARγ in Tie2-expressing cells could result in the increased PDGF-Rβ expression in the PASMCs, we measured levels of the PPARγ target apoE. ApoE is a transcriptional target of PPARγ (10), and it can inhibit PDGF-Rβ signaling (2) and PASMC proliferation (10). We first investigated the ability of apoE to regulate PDGF-Rβ expression in cultured human PASMCs. We found a dose-dependent inverse relationship between apoE and PDGF-Rβ expression, with 10 μg/ml of recombinant apoE significantly decreasing PDGF-Rβ expression by ∼35% (P < 0.01; Fig. 4A). However, when we determined whether the loss of PPARγ reduced apoE expression in the transgenic mice, we found that the apoE mRNA level was similar in whole lung homogenates from both genotypes (data not shown) and that apoE protein was increased by approximately twofold in the PECs isolated from the Tie2 PPARγ−/− vs. WT mice (Fig. 4B).
Blocking PDGF-Rβ Signaling Is Sufficient to Prevent the PAH that Develops in the Tie2 PPARγ−/− Mice Under Room Air
Although we could not ascribe the increased PDGF-Rβ expression in the transgenic mice to a reduction in apoE, we next determined whether blocking PDGF signaling would reverse the PAH observed in these animals at baseline. We administered over the course of 2 wk, a daily intraperitoneal injection of imatinib, a competitive inhibitor of PDGF-Rβ tyrosine kinase activity (35). We demonstrate that treatment with imatinib effectively blocked PDGF-Rβ signaling in the Tie2 PPARγ−/− mice as evidenced by decrease in its phosphorylation (tyrosine 751) and its activity as reflected by phosphorylation of ERK1/2 (Fig. 5, A and B). Furthermore, this treatment reversed the PAH in the Tie2 PPARγ−/− mice, decreasing RVSP, RVH, and muscularization of peripheral arteries to similar levels observed in WT controls (Fig. 5, A–E).
In our previous studies (10), we showed that PAH develops in mice in which PPARγ is decreased in SMC, and that PPARγ acts downstream of BMPR-II signaling. Since PPARγ is reduced in both the ECs and SMCs of PAH patients (1), we hypothesized that the loss of PPARγ in ECs might also lead to PAH. Consistent with the fact that the Tie2 promoter drives expression in both EC and bone marrow-derived hematopoietic cells (17, 40), we found a loss PPARγ in the ECs of the Tie2 PPARγ−/− mice. Although the severity of the PAH under room air conditions was mild, there was associated significant RVH and muscularization of distal arteries. Given that the hemodynamic and structural response to hypoxia was similar in both genotypes, it could be that the mechanism responsible for the baseline PAH observed in the Tie2 PPARγ−/− mice in room air may be similar to the one that produces PAH in the WT mice during chronic hypoxia. Indeed, elevated PDGF-Rβ expression and activation were present in the Tie2 PPARγ−/− mice in room air and in both genotypes following chronic hypoxia. Inhibition of this abnormal PDGF-Rβ activity in the Tie2 PPARγ−/− mice in room air with imatinib normalized the hemodynamic and pulmonary vascular remodeling to levels close to those found in WT mice, implicating the increased PDGF-Rβ activity as one mechanism necessary for the evolution of the PAH in our model. However, the expression of PDGF-Rβ in the two groups was similar following 4 wk of recovery in room air despite that a persistent increase in PAH severity was observed in the Tie2 PPARγ−/− mice upon return to normoxia, suggesting that an additional mechanism is likely responsible for the impaired recovery.
We (10) and others (30, 35) have shown that PDGF-Rβ-mediated signals drive the PASMC proliferation associated with PAH. PDGF expression increased in lung biopsies obtained from patients with severe PAH (30, 35). In addition, treatment of rats with either monocrotaline- or chronic hypoxia-induced PAH with the PDGF-Rβ inhibitor imatinib produced a dose-dependent significant improvement of pulmonary hemodynamics and vascular remodeling within 2 wk (35). Given that the PPARγ expression is intact in the pulmonary vascular SMCs of the Tie2 PPARγ−/− mice, the increased PDGF-Rβ expression and signaling observed in the walls of the distal PAs are likely due to paracrine effects. Endothelial abnormalities and dysfunction in the intercellular communications between ECs and SMCs have been implicated in the pathogenesis of PAH by our group (31) and by others (5, 7, 13, 15, 32). In support of this, a recent study from Hong et al. (13) demonstrated that conditional heterozygous or homozygous BMPR-II deletion in PEC predisposes mice toward the development of PAH and induces a higher proliferation index of both PECs and pulmonary SMCs. As PPARγ acts downstream of BMPR-II signaling (10), we cannot exclude the possibility that mechanisms causing the spontaneous PAH in our model are similar to those leading to the PAH that develops in mice following ablation of the BMPRII gene in ECs. In addition, similar to BMPR-II, activation of PPARγ protects the vascular endothelium from inflammation, by decreasing the nitric oxide synthase inhibitor asymmetric dimethylarginine (23, 37, 39), ET-1 (21, 33), and cell surface and secreted molecules related to inflammation such as vascular cell adhesion molecule and fractalkine (25, 36). While inflammatory mediators known to play a role in PAH may be increased and could be an additional mechanism contributing to the pulmonary vascular changes observed, no significant changes in plasma ET-1 levels and nitrate/nitrite productions were seen in the lungs from the Tie2 PPARγ mice compared with WT mice. Using a similar Tie2-mediated PPARγ knockout model, Kanda et al. (16) demonstrated that these transgenic mice vs. controls had impaired arterial relaxation only after feeding with a high-fat diet. In contrast, Kleinhenz et al. (18) demonstrated impaired endothelium-dependent vasodilatation in aortic rings following feeding with standard chow diets and found that intact aortic ring segments from Tie2 PPARγ−/− mice released less nitric oxide than WT mice. These studies indicate that different vascular beds may have different sensitivities to the loss of EC PPARγ. While reduced PPARγ has been demonstrated in human PAECs after 72 h of hypoxia (28), we could not document a similar decrease after 24 h and our cells did not tolerate longer exposures.
In view of our previous report (10) that the spontaneous PAH developing in the SM22 PPARγ−/− mice was related to reduced levels of a target of PPARγ, apoE, we measured apoE levels in the Tie2 PPARγ−/− mice. ApoE is a molecule that inhibits PDGF-Rβ function by binding to low-density receptor-like protein-1 and by activation of the PDGF-Rβ (2, 14, 26, 47). We demonstrated that the incubation of primary human PASMCs with recombinant apoE resulted in a dose-dependent inverse relationship between apoE and PDGF-Rβ protein expression. However, we found that the apoE levels were elevated rather than decreased in two different Tie2-expressing cells in the Tie2 PPARγ−/− mice, PECs, and alveolar macrophages (data not shown). Although the precise mechanisms causing the increased expression and activation of PDGF-Rβ in the Tie2 PPARγ−/− mice in room air are not clear, these data suggest that heightened apoE expression, by preventing a further increase in PDGF-Rβ, might represent a potent protective mechanism against the development of more severe PAH in these mice and might explain the similar severity of PAH observed in chronic hypoxia in the Tie2 PPARγ−/− and the WT mice. In our previous studies (10) in which PPARγ is decreased in SMCs we identified a BMP-mediated increase in apoE that was PPARγ independent, so it is possible that this pathway compensates for the loss of PPARγ-mediated apoE.
In summary, although the precise pathophysiological mechanisms explaining the increased expression and activation of PDGF-Rβ in the Tie2 PPARγ−/− mice are not clear, our data demonstrate that loss of endothelial PPARγ results in PA muscularization and development of PAH. This model of disease supports the hypothesis that endothelial dysfunction plays a central role in the pathogenesis of the structural alterations of the pulmonary vasculature accompanying PAH.
This work was supported by Postdoctoral Fellowships from the Délégation à la Recherche Clinique de l'AP-HP (to C. Guignabert); the Sigrid Juselius Foundation, Instrumentarium Foundation, the Finnish Foundation for Cardiovascular Research, and the Academy of Finland (to T. Alastalo); the Department of Pediatrics of Mie University Graduate School of Medicine, Japan (to H. Sawada); the American Heart Association/Pulmonary Hypertension Association (to N. El-Bizri and G. Hansmann); and National Heart, Lung, and Blood Institute Grants R01-HL074186 and R01-HL087118 and the Dwight and Vera Dunlevie Endowed Professorship (to M. Rabinovitch).
None of the authors have any disclosures related to conflict of interest.
Present address of G. Hansmann: Dept. of Cardiology, Children's Hospital Boston, Harvard Medical School, Boston, MA.
- Copyright © 2009 the American Physiological Society