Vascular Medicine Endothelial-to-Mesenchymal Transition in Pulmonary Hypertension Benoît Ranchoux, MSc*; Fabrice Antigny, PhD*; Catherine Rucker-Martin, PhD; Aurélie Hautefort, MSc; Christine Péchoux, PhD; Harm Jan Bogaard, MD, PhD; Peter Dorfmüller, MD, PhD; Séverine Remy, PhD; Florence Lecerf; Sylvie Planté; Sophie Chat; Elie Fadel, MD, PhD; Amal Houssaini, MSc; Ignacio Anegon, MD, PhD; Serge Adnot, PhD; Gerald Simonneau, MD; Marc Humbert, MD, PhD; Sylvia Cohen-Kaminsky, PhD; Frédéric Perros, PhD Background—The vascular remodeling responsible for pulmonary arterial hypertension (PAH) involves predominantly the accumulation of α-smooth muscle actin–expressing mesenchymal-like cells in obstructive pulmonary vascular lesions. Endothelial-to-mesenchymal transition (EndoMT) may be a source of those α-smooth muscle actin–expressing cells. Methods and Results—In situ evidence of EndoMT in human PAH was obtained by using confocal microscopy of multiple fluorescent stainings at the arterial level, and by using transmission electron microscopy and correlative light and electron microscopy at the ultrastructural level. Findings were confirmed by in vitro analyses of human PAH and control cultured pulmonary artery endothelial cells. In addition, the mRNA and protein signature of EndoMT was recognized at the arterial and lung level by quantitative real-time polymerase chain reaction and Western blot analyses. We confirmed our human observations in established animal models of pulmonary hypertension (monocrotaline and SuHx). After establishing the first genetically modified rat model linked to BMPR2 mutations (BMPR2Δ140Ex1/+ rats), we demonstrated that EndoMT is linked to alterations in signaling of BMPR2, a gene that is mutated in 70% of cases of familial PAH and in 10% to 40% of cases of idiopathic PAH. We identified molecular actors of this pathological transition, including twist overexpression and vimentin phosphorylation. We demonstrated that rapamycin partially reversed the protein expression patterns of EndoMT, improved experimental PAH, and decreased the migration of human pulmonary artery endothelial cells, providing the proof of concept that EndoMT is druggable. Conclusions—EndoMT is linked to alterations in BPMR2 signaling and is involved in the occlusive vascular remodeling of PAH, findings that may have therapeutic implications. (Circulation. 2015;131:1006-1018. DOI: 10.1161/ CIRCULATIONAHA.114.008750.) Key Words: bone morphogenetic protein receptors, type II ◼ cardiovascular diseases ◼ epithelial-mesenchymal transition ◼ hypertension, pulmonary ◼ models, animal ◼ neointima ◼ sirolimus ◼ TWIST1 protein, human ◼ vascular remodeling ◼ vimentin
P
ulmonary arterial hypertension (PAH) is a rare disorder, with a prevalence of 15 to 50 patients per million in the population. It is characterized by remodeling of the precapillary
Clinical Perspective on p 1018
pulmonary arteries, with endothelial cell dysfunction contributing to endothelial and pulmonary artery smooth muscle cell proliferation. This remodeling increases pulmonary vascular resistance and pulmonary arterial pressure (mean pulmonary arterial pressure ≥25 mm Hg and a pulmonary capillary wedge
Received October 16, 2013; accepted December 19, 2014. From Univ. Paris-Sud, Faculté de médecine, Kremlin-Bicêtre, France (B.R., F.A., C.R.-M., A.H., P.D., F.L., E.F., G.S., M.H., S.C.-K., F.P.); AP-HP, DHU TORINO, Centre de Référence de l’Hypertension Pulmonaire Sévère, Service de Pneumologie et Réanimation Respiratoire, Hôpital Bicêtre, Le Kremlin-Bicêtre, France (B.R., F.A., C.R.-.M., A.H., P.D., F.L., G.S., M.H., S.C.-K., F.P.); INSERM UMR-S 999, Labex LERMIT, Hypertension Artérielle Pulmonaire: Physiopathologie et Innovation Thérapeutique, Centre Chirurgical Marie Lannelongue, Le Plessis-Robinson, France (B.R., F.A., C.R.-M., A.H., P.D., F.L., E.F., G.S., M.H., S.C.-K., F.P.); INRA U1196, Génomique et Physiologie de la Lactation – Plateau de Microscopie Electronique à Transmission, Jouy-en-Josas, France (C.P., S.C.); Service de Chirurgie Thoracique, Centre Chirurgical Marie Lannelongue, Le Plessis-Robinson, France (E.F.); Service d’Anatomie Pathologique, Centre Chirurgical Marie Lannelongue, Le Plessis Robinson, France (P.D., S.P.); Department of Pulmonary Medicine, Institute for Cardiovascular Research, VU University Medical Center, Amsterdam, The Netherlands (H.J.B.); INSERM UMR 1064-Center for Research in Transplantation and Immunology-ITUN et Transgenic Rats and Immunophenomic Platform, Nantes, France (S.R., I.A.); and INSERM U955, Département de Physiologie and Service de Cardiologie, Hôpital Henri Mondor, AP-HP, Université Paris-Est Créteil (UPEC), Créteil, France (A.H., S.A.). *Drs Ranchoux and Antigny contributed equally. The online-only Data Supplement is available with this article at http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIRCULATIONAHA. 114.008750/-/DC1. Correspondence to Frédéric Perros, PhD, INSERM U999, Centre Chirurgical Marie Lannelongue, 133, Avenue de la Résistance, F-92350 Le Plessis Robinson, France. E-mail
[email protected] © 2015 American Heart Association, Inc. Circulation is available at http://circ.ahajournals.org
DOI: 10.1161/CIRCULATIONAHA.114.008750
Downloaded from http://circ.ahajournals.org/ by guest on March 16, 2015 1006
Ranchoux et al pressure ≤15 mm Hg at rest), which ultimately leads to progressive right ventricular dysfunction and death.1 The nature of the primary abnormality that triggers and perpetuates pulmonary vascular cell proliferation in PAH is unclear. PAH is associated with a spectrum of structural changes in the pulmonary arteries: increased medial thickness, eccentric and concentric intimal thickening, the obliteration and recanalization of arteries, and the appearance of plexiform and dilatation lesions.2 The plexiform lesion is a characteristic structure of the pulmonary arteriopathy in severe PAH. According to the consensus view, the plexiform lesion is a complex and disorganized pulmonary arterial proliferative lesion that consists of a network or plexus of channels lined by endothelial cells and separated by core cells.3 However, it has not been determined whether the core cells are myofibroblasts, smooth muscle cells, endothelial cells, or undifferentiated cells.3 Intimal and medial thickening are the most consistently encountered structural changes in PAH.2 All of these changes are characterized by increased numbers of cells expressing α-smooth muscle actin (α-SMA).4 It is thought that α-SMA–positive cells that accumulate in vascular lesions are derived from the expansion of resident vascular smooth muscle cells or adventitial fibroblasts. However, it is increasingly recognized that endothelial cells (ECs) can transition into mesenchymal cells expressing α-SMA and that this process contributes to the accumulation of smooth muscle–like cells in vascular pathologies.5 Endothelial-to-mesenchymal transition (EndoMT) is a biological process in which ECs progressively change their endothelial phenotype into a mesenchymal or myofibroblastic phenotype. During this process, ECs dissociate from the monolayer of tightly cohesive cells at the abluminal surface of the vessel and migrate toward the inner tissue. The migration starts with the loss of cell-cell contact mediated by membrane proteins such as vascular endothelial cadherin (VE-cadherin) and CD31/PECAM-1, and by the cytoplasmic scaffold protein p120-catenin. By losing their cell-cell junction, EndoMT-derived cells gain a migratory and invasive capacity allowing them to reach the surrounding tissues. While migrating, the cells lose specific endothelial markers such as CD31, VE-cadherin, and CD34 and progressively express mesenchymal or myofibroblastic markers like α-SMA and vimentin.6 This phenomenon occurs during certain embryonic stages of pulmonary artery development7 but also seems to be implicated in pathological fibroblast and myofibroblast accumulation in conditions such as cardiac or renal fibrosis8,9 and chronic hypertension.5 The potential role of EndoMT in PAH has been previously suggested4,5,10 based on analogies with other diseases,8,9 epithelial-to-mesenchymal transition,5 and mechanisms of embryonic vascular development,7,11 and also on the basis of in vitro experiments.6,12,13 Here, we provide the first in situ evidence of EndoMT in human and experimental pulmonary hypertension (PH). Because a reduction in peripheral arterial volume, seen as intimal thickening and arterial obliteration, is likely to be the major contributor to the onset and maintenance of PAH, we analyzed in priority the phenotype of endothelial and subendothelial cells in human intimal lesions. We also elucidated the phenotype of endothelial and core cells in plexiform
EndoMT in Pulmonary Hypertension
1007
lesions. We hypothesized that α-SMA+ cells building these lesions had an endothelial origin and resulted from EndoMT. We used explanted tissue from PAH (group 1 of the Dana Point classification)14 and from severe models of PAH induced in rats by exposure to the toxic alkaloid monocrotaline (MCT) or by the combined exposure to chronic hypoxia and vascular endothelial growth factor receptor blockade with Sugen/ SU5416.15 We analyzed the phenotype of endothelial and subendothelial cells at structural and ultrastructural levels by multiple immunofluorescence staining and correlative light and electron microscopy (CLEM). The loss of endothelial cell-cell junctions, which is essential for EndoMT, was characterized by immunostaining and by Western blot analysis (VE-cadherin, p120-catenin). We measured the expression of Twist-1, a master transcription factor for EndoMT, in both human and experimental PAH. EndoMT is stimulated by transforming growth factor β (TGF-β) signaling,16 but, at the same time, it is inhibited by BMPR2,17 a TGF-β receptor implicated in human PAH.18 After establishing the first genetically modified rat model linked to BMPR2 mutation (BMPR2Δ140Ex1/+ rats), we searched for evidence of pulmonary vascular EndoMT linked to BMPR2 signaling. Rapamycin was used in MCTinduced PAH and on cultured human pulmonary endothelial cells (PAECs) to regulate EndoMT-associated processes, like the acquisition of a migratory phenotype.
Methods Patients Human lung specimens were obtained during lung transplantation from patients with PAH and during lobectomy or pneumonectomy for localized lung cancer from control subjects (n numbers are indicated in every legend of the figures). In the lung specimens from control subjects, pulmonary arteries were studied at a distance from tumor areas. Transthoracic echocardiography was performed preoperatively in the control subjects to rule out PH. Patients studied were part of the French Network on Pulmonary Hypertension, a program approved by our institutional Ethics Committee, and had given written informed consent (Protocol N8CO-08- 003, ID RCB: 2008-A00485-50, approved on June 18, 2008).
Immunofluorescence Staining The full description of the immunofluorescence staining procedures is available in the online-only Data Supplement (Tables I and II in the online-only Data Supplement).
Immunohistochemical Detection of Twist-1 in Paraffin-Embedded Lung Tissues After classical dewaxing and heat antigen retrieval at pH 6, immunohistochemistry was performed with a rabbit anti–Twist-1 (Ref ab50581, Abcam, UK), detected by a biotinylated goat anti-rabbit and streptavidin peroxidase (Thermo-Scientific, France) and permanent AEC kit (Ref ZUC054-200, Zytomed, Germany). Slides were counterstained with hematein.
Transmission Electron Microscopy and CLEM The full description of the transmission electron microscopy and CLEM procedures are available in the online-only Data Supplement. To quantify α-SMA and phospho-vimentin labeling, gold particles localized in ECs or SECs of lung control artery from intimal or plexiform lesion of PAH lung were counted on 6 to 40 micrographs randomly taken, at the same magnification of 12 000. The particle density (number of gold particles per 10 μm2 of cytoplasm area) was
Downloaded from http://circ.ahajournals.org/ by guest on March 16, 2015
1008
Circulation
March 17, 2015
calculated as previously described.19 Area was measured with iTEM software.
Rat Specimens and In Vivo Study Design PH and control age/sex-matched pulmonary rat tissues came from previously published studies.20–22 PH was induced in rats by MCT (60 mg/kg)20,21 or by the combined exposure to chronic hypoxia and vascular endothelial growth factor receptor blockage with Sugen (SuHx model).22 We had access to the tissues from different time points of PH development (kinetic of MCT-induced PH development)20 and from rats exposed to MCT and rapamycin (5 mg/kg from day 21 to 28).21 Hemodynamic data (mean pulmonary artery pressure) and right ventricular morphology (Fulton Index) were available from animals of the kinetic study (unpublished data that served in setting up the experiments). The SuHx model is a severe angio-obliterative PH model that reproduces multiple salient histological features of human PAH23 (see the online-only Data Supplement).
Generation of BMPR2-Deficient Rats BMPR2-deficient rats were generated by using zinc-finger nucleases (Sigma, St. Louis, MO) as previously described.24,25 In brief, we microinjected into the cytoplasm of Sprague-Dawley zygotes mRNA at 5 ng/μL encoding a pair of zinc-finger nucleases recognizing rat BMPR2 sequences. Newborn animals were genotyped by a T7 endonuclease assay and sequencing of polymerase chain reaction products of the targeted sequence. Founders displaying a shift in the coding reading frame with a premature stop codon were used to derive animal lines. A rat line with a heterozygous 140 base pairs deletion in the first exon (BMPR2Δ140Ex1/+ rats) was chosen for this study because it displayed an intense pulmonary vascular remodeling at 3 months of life that was absent in the wild-type littermates.
Migration and Proliferation Assays on Cultured Human PAECs Human PAECs from control and PAH patients were cultured as previously described26 and were used between passages 3 and 5. Proliferation was measured with the DELFIA Cell Proliferation Kit (PerkinElmer) and migration with the CytoSelect 96-Well Cell Migration Assay, 8 μm (cell Biolabs), following manufacturer instructions. Cells were pretreated 24 hours before starting the experiment and treated during the time of the experiment with rapamycin (50 ng/mL) or by the same volume of its solvent (dimethyl sulfoxide).
Quantification of p120-Catenin, VE-Cadherin, Vimentin, Phospho-Vimentin, Twist-1, and BMPR2 Lung Expression by Western Blot The full description of the Western blot procedure is available in the online-only Data Supplement.
Real-Time Quantitative Polymerase Chain Reaction The full description of the real-time quantitative polymerase chain reaction procedure is available in the online-only Data Supplement and Figure I in the online-only Data Supplement.
Statistical Analysis Owing to small sample sizes, we used nonparametric statistical analyses: Wilcoxon rank sum test (Mann-Whitney U test) for comparing data between 2 independent groups, Wilcoxon signed-rank test for comparing data between paired observations, and Kruskal-Wallis test for comparing data among 3 or more independent groups. Probability values of
