Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Am J Physiol Lung Cell Mol Physiol 305: L725–L736, 2013. First published September 13, 2013; doi:10.1152/ajplung.00186.2013. Pulmonary veins in the normal lung and pulmonary hypertension due to left heart disease James M. Hunt,1 Brian Bethea,2 Xiang Liu,3,4 Aneta Gandjeva,1 Pradeep P. A. Mammen,5 Elvira Stacher,6,7 Marina R. Gandjeva,1 Elisabeth Parish,8 Mario Perez,1 Lynelle Smith,1 Brian B. Graham,1 Wolfgang M. Kuebler,3,9 and Rubin M. Tuder1 1 Submitted 15 July 2013; accepted in final form 8 September 2013 Hunt JM, Bethea B, Liu X, Gandjeva A, Mammen PP, Stacher E, Gandjeva MR, Parish E, Perez M, Smith L, Graham BB, Kuebler WM, Tuder RM. Pulmonary veins in the normal lung and pulmonary hypertension due to left heart disease. Am J Physiol Lung Cell Mol Physiol 305: L725–L736, 2013. First published September 13, 2013; doi:10.1152/ajplung.00186.2013.—Despite the importance of pulmonary veins in normal lung physiology and the pathobiology of pulmonary hypertension with left heart disease (PH-LHD), pulmonary veins remain largely understudied. Difficult to identify histologically, lung venous endothelium or smooth muscle cells display no unique characteristic functional and structural markers that distinguish them from pulmonary arteries. To address these challenges, we undertook a search for unique molecular markers in pulmonary veins. In addition, we addressed the expression pattern of a candidate molecular marker and analyzed the structural pattern of vascular remodeling of pulmonary veins in a rodent model of PH-LHD and in lung tissue of patients with PH-LHD obtained at time of placement on a left ventricular assist device. We detected urokinase plasminogen activator receptor (uPAR) expression preferentially in normal pulmonary veins of mice, rats, and human lungs. Expression of uPAR remained elevated in pulmonary veins of rats with PH-LHD; however, we also detected induction of uPAR expression in remodeled pulmonary arteries. These findings were validated in lungs of patients with PH-LHD. In selected patients with sequential lung biopsy at the time of removal of the left ventricular assist device, we present early data suggesting improvement in pulmonary hemodynamics and venous remodeling, indicating potential regression of venous remodeling in response to assist device treatment. Our data indicate that remodeling of pulmonary veins is an integral part of PH-LHD and that pulmonary veins share some key features present in remodeled yet not normotensive pulmonary arteries. pulmonary hypertension; pulmonary circulation; left heart failure; pulmonary veins; vessel remodeling ALTHOUGH THE ARTERIAL and venous pulmonary circulations form a communicating, integrated vascular system, pulmonary veins are developmentally, physiologically, and structurally distinct from pulmonary arteries. Pulmonary veins develop Address for reprint requests and other correspondence: R. M. Tuder, Program in Translational Lung Research, Division of Pulmonary Sciences and Critical Care Medicine, Univ. of Colorado Denver, Anschutz Medical Campus, Research 2 — 9th floor, Rm. 9001; Mail stop C-272, 12700 East 19th Ave., Aurora, CO 80045 (e-mail: [email protected]). http://www.ajplung.org chronologically after the initial segments of large pulmonary arteries form and are often identifiable in a perpendicular arrangement to that of the pulmonary arteries in the bronchoarterial sheath (5). Pulmonary veins respond physiologically to the vasodilator prostacyclin and to vasoconstrictors, such as platelet activating factor, to a larger extent than that observed with pulmonary arteries (9, 34). These anatomic and physiological distinctions are carried over to the histological level since pulmonary veins lack distinct double internal and external elastic laminae and have only a thin medial muscular layer (as compared with similar diameter arteries) (32). However, this more refined histological distinction does not persist in diseases characterized by pulmonary hypertension (PH), because pulmonary veins become more arterial in appearance, acquiring a distinct double elastic layer flanking a hypertrophic media (33). Although veins in the lung parenchyma lack unique distinguishing features, the location of larger veins in interlobular septa (alongside pulmonary lymphatics) provides for a unique anatomic landmark that aids in their histological recognition. Indeed, venous pathology within interlobular septa serves as the diagnostic criterion of a rare but characteristic form of PH due to thrombotic venous occlusion, veno-occlusive disease (VOD) (21, 25). The rarity of VOD contrasts with PH due to left heart disease (PH-LHD), which is the most common cause of PH in the United States and affects over 250,000 Americans and perhaps as many as 2 million people worldwide (11, 26, 27). Despite its high prevalence and significant impact on morbidity and mortality, the venous pathology and pathophysiology of PH-LHD are poorly understood, and insights that can guide proper recognition and translational investigations are lacking. These shortcomings also apply to the venous involvement in pulmonary vascular disease caused by the autoimmune disease scleroderma, thought to contribute to the severity of the disease (6, 7). A critical limitation to the study of pulmonary venous disease is the current inability to identify parenchymal pulmonary veins; as outlined above, these vessels lack clear histological or distinct molecular markers. A recent publication showed that the pulmonary vein endothelium in adult mice expresses the ephrin B4 receptor (EphB4), which was, however, not reactive in adult human pulmonary veins (4, 12); 1040-0605/13 Copyright © 2013 the American Physiological Society L725 Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on May 6, 2017 Program in Translational Lung Research, Division of Pulmonary Sciences and Critical Care Medicine, Anschutz Medical Campus, Aurora, Colorado; 2University of Texas Southwestern Medical Center, Department of Cardiothoracic and Vascular Surgery, Dallas, Texas; 3Institute of Physiology, Charité — Universitätsmedizin Berlin and German Heart Institute, Berlin, Germany; 4Department of Anesthesia, Shanghai East Hospital, China; 5University of Texas Southwestern Medical Center, Division of Cardiology, Dallas, Texas; 6Institute of Pathology, Medical University of Graz, Graz, Austria; 7Ludwig Boltzmann Institute for Lung Vascular Research, Graz, Austria; 8University of Texas Southwestern Medical Center, School of Medicine, Dallas, Texas; and 9The Keenan Research Centre, St. Michael’s Hospital, Toronto, Ontario, Canada L726 PULMONARY VEINS AND LEFT HEART DISEASE MATERIALS AND METHODS Reagents. We purchased commercial antibodies to uPAR (Santa Cruz Biotechnology, Santa Cruz, CA), anti-smooth muscle actin (anti-␣-SMA) (Abcam, Cambridge, MA), heavy chain cardiac myosin (HCCM) (mouse monoclonal) IgG1 (Abcam), and EphB4 (R&D Systems, Minneapolis, MN). Commercial kits were used for all immunohistochemistry and immunofluorescence (Vector Labs, Burlingame, CA), and manufacturer’s instructions were followed. A commercial kit for the Russell-Movat pentachrome stain was purchased and the manufacturer’s instructions were followed (American MasterTech Scientific, Lodi, CA). Animals. Animal studies were conducted in accordance with the Guide for the Care and Use of Laboratory Animals, 9th Edition (2011), and all protocols were approved by the University of Colorado’s Institutional Animal Care and Use Committee. C57BL/6 male wild-type mice were purchased from The Jackson Laboratory (Bar Harbor, ME), and 2- to 3-wk-old male Sprague Dawley rats were purchased from Harlan Laboratories (Denver, CO). Humans. Study subjects were prospectively enrolled patients referred to a tertiary cardiac hospital for LVAD placement due to PH-LHD from February 2010 to June 2011. The study was approved by the University of Texas Southwestern Institutional Review Board (IRB 092010-093), and all samples were deidentified by assignment of a code at the site of collection. Briefly, wedge biopsies of the left or right lower lung were taken at the time of LVAD placement. Three patients had additional wedge biopsies collected at the time of LVAD removal and heart transplantation. Tissue was immediately fixed in 4% paraformaldehyde, processed, and embedded in paraffin blocks. Clinical data were obtained and matched to the coded lung samples (Table 1). Control subjects consisted of 20 randomly chosen failed organ donors enrolled in the Pulmonary Hypertension Breakthrough Initiative from April 2006 to August 2011 (31). Deidentified clinical data and lung tissue samples were available. The study was approved by the Colorado Multiple Institution Review Board (COMIRB 10-1440 and 11-0135). Venous backfilling. Venous backfilling was performed in anesthetized animals ventilated at tidal volumes of 6 ml/kg. A median thoracotomy was performed and the pulmonary circulation was flushed with heparin and phosphate-buffered saline until clear. The aorta was cross-clamped with a hemostat and a 1:10 bead-1% agarose mixture (3 ml for rats, 250 ml for mice) was injected under constant pressure (20 mmHg) into the left ventricle, backfilling the left atria and pulmonary veins. The beads were 1 m in diameter and were either red fluosphere polystyrene microspheres (Molecular Probes, Eugene, OR), to be later used in fresh frozen sections, or green silica Table 1. Demographic and hemodynamic details in control subjects and patients with PH-LHD No. Age, yr Sex (male), no. (%) Race, no. (%) W AA Ever smoker, % BMI, kg/m2 CHF diagnosis, no. (%) Ischemic Nonischemic Viral Idiopathic DCM Echocardiogram, no. (%) Severe LV Dysfunction RV Dilatation Normal Mild Moderate Severe RHC hemodynamics RAP, mmHg mPAP, mmHg PAOP, mmHg TPG, mmHg CO, l/min PVR, Woods units NYHA class (%) Class I Class II Class III Class IV PH-LHD Control 18* 57.4 ⫾ 12.9 (30–78) 13 (72) 19 34.5 ⫾ 17.8 (12–62) 14 (73) 18 (100) 18 (95) 1 (5) NA NA 12 (67) 28.6 ⫾ 6.3 8 (44) 8 (44) 1 (6) 1 (6) NA 18 (100) 3 (16) 5 (28) 6 (33) 1 (6) 13.2 ⫾ 6.4 41.6 ⫾ 12.3 24.0 ⫾ 6.9 18.6 ⫾ 6.7 4.2 ⫾ 1.1 4.7 ⫾ 2.4 0 0 4 (23) 14 (77) NA NA NA Data are shown as ⫾SD. AA, African American; BMI, body mass index; CHF, congestive heart failure; CO, cardiac output; DCM, dilated cardiomyopathy; LV, left ventricle; NYHA, New York Heart Association; PH-LHD, pulmonary hypertension with left heart disease; mPAP, mean pulmonary arterial pressure; PAOP, pulmonary arterial occlusion pressure; PVR, pulmonary vascular resistance; RAP, right atrial pressure; RV right ventricle; TPG, transpulmonary gradient; W, White; NA, not available. *Clinical data were not available for 4 cases of PH-LHD. particles (Kisker Biotech, Steinfurt, Germany) for use in formaldehyde-fixed paraffin-embedded tissue. Aortic banding model of PH-LHD. PH-LHD was induced in juvenile rats (92 ⫾ 9.7 g in weight) by supracoronary aortic banding (AOB) as previously described (35). Isoflurane was used as the anesthetic for all procedures. Briefly, anesthetized rats underwent a left thoracotomy and the ascending aorta was partially occluded by a titanium clip (Hemoclip, Weck Closure System, Research Triangle Park, NC) with an internal diameter of 0.8 mm. Sham-operated rats (with implantation of the clip onto the thymus) of similar body weight and age served as controls. Hemodynamic monitoring. Nine weeks after AOB or sham operations, rats underwent invasive hemodynamic monitoring. PH-LHD or control rats were anesthetized and ventilated with a tidal volume of 6 ml/kg body wt. Polyvinyl catheters with an internal diameter of 0.58 mm were introduced via the left carotid artery and the right jugular vein for continuous monitoring of mean arterial and central venous pressure, respectively. Measurement of pulmonary arterial pressure was performed through the right jugular catheter after it was advanced into the pulmonary artery via right atria and ventricle. Following a median thoracotomy the pericardium was opened and a 1.6-Fr solidstate catheter (Transonic Systems, Ithaca, NY) was introduced into the left ventricle to measure left ventricular end-diastolic pressure AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00186.2013 • www.ajplung.org Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on May 6, 2017 furthermore, rats lack an annotated EphB4 gene analog. Consequently, molecular tools to systematically study the pulmonary venous circulation and pulmonary venous disease have been largely lacking to date. Reliant on fluorescent backcasting, which is allied to laser capture microdissection (LCM) enrichment of vein cells from normal lungs, we report on a set of molecules that are preferentially expressed in normal pulmonary veins. When applied to both a rodent model of PH due to left heart dysfunction and human lung samples of patients with the disease, one of these markers, urokinase plasminogen activator receptor (uPAR), became shared with pulmonary arteries, consistent with a structural and molecular convergence of arterial and venous phenotypes at in pulmonary hypertensive states. Importantly, these novel observations were validated in lungs of patients with PH-LHD, in whom the PH-LHD was attenuated after placement of a left ventricular assist device (LVAD). L727 PULMONARY VEINS AND LEFT HEART DISEASE Table 2. PCR array with the 19 top genes preferentially expressed in veins compared with arteries⫹veins Average ⌬Ct Position in PCR Array E10 E03 D03 F04 C10 C03 F02 D12 C12 E12 D10 F01 G04 F11 D04 E11 D01 Description Plasminogen activator, urokinase Matrix metallopeptidase 9 Interleukin 6 Sphingosine kinase 1 Hepatocyte growth factor Fibroblast growth factor 6 Serine (or cysteine) peptidase inhibitor, clade F, member 1 Midkine Interferon gamma Plexin domain containing 1 Leptin Prostaglandinendoperoxide synthase 1 Tissue inhibitor of metalloproteinase 1 Transforming growth factor, beta 2 Integrin alpha V Plasminogen Insulin-like growth factor 1 Gene Symbol Gene Name Veins Control group artery⫹ parenchyma 2⫺⌬Ct Fold Up- or Downregulation Group 1 veins Control group artery/ parenchyma Group 1 veins vs. control group artery/ parenchyma u-PA/uPA Plau ⫺0.62 6.71 1.536875 0.009552 160.9 AW\?\43869/\?\MM P9/Clo4b/ MMPIl-6 1110006G24Rik/Sk1/Spk1 C230052L06Rik/HGF/SF/NK1/ NK2/SF/HGF Fgf-6/HSTF-2 Mmp9 ⫺0.25 6.71 1.189207 0.009552 124.5 Il6 Sphk1 Hgf 0.01 0.14 0.06 6.71 6.66 6.52 0.993092 0.907519 0.959264 0.009552 0.009889 0.010896 103.97 91.77 88.03 Fgf6 0.8 6.71 0.574349 0.009552 60.13 AI195227/EPC-1/Pedf/Pedfl/Sdf3 Serpinf1 0.03 5.92 0.97942 0.016459 59.51 MK/Mek IFN-g/Ifg 2410003I07Rik/AI848450/ MGC130377/Tem7 ob/obese COX1/Cox-1/Cox-3/Pghs1 Mdk Ifng Plxdc1 0.74 1.11 0.58 6.62 6.71 6.09 0.598739 0.463294 0.668964 0.010132 0.009552 0.01468 59.1 48.5 45.57 Lep Ptgs1 1.28 1.05 6.71 6.23 0.411796 0.482968 0.009552 0.013276 43.11 36.38 Clgi/MGC7143/TIMP-1/Timp Timp1 1.63 6.71 0.323088 0.009552 33.82 BB105277/Tgf-beta2/Tgfb-2 Tgfb2 1.22 6.28 0.429283 0.012869 33.36 1110004F14Rik/2610028E01Rik/ CD51/D430040G12Rik AI649309/Pg C730016P09Rik/Igf-1/Igf-1 Itgav ⫺0.51 4.45 1.42405 0.045753 31.12 1.95 1.61 6.71 6.07 0.258816 0.327598 0.009552 0.014833 27.1 22.09 Plg Igf1 The full set of gene list containing 96 genes can be provided upon request. AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00186.2013 • www.ajplung.org Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on May 6, 2017 10 min, followed by 40 cycles of denaturation at 95°C for 15 s and annealing at 60°C for 1 min. Software available through the manufacturer’s website (http://pcrdataanalysis.sabiosciences.com/pcr/ arrayanalysis.php) was used to analyze the RT2 PCR data. We calculated relative abundance of mRNA expression in each RT-PCR sample as 2⫺⌬⌬Ct (Table 2). The full list of genes and the fold increase/decrease in veins vs. arteries⫹parenchyma can be provided upon request. Immunohistochemistry and immunofluorescence. We performed immunohistochemistry and immunofluorescence on 4-m-thick paraffin-embedded sections of agarose-inflated lungs from mice or rats as previously described, or on 4-m-thick paraffin-embedded sections of human lung tissue. Primary antibodies included rabbit polyclonal antibody to human uPAR (1:25, sc-10815, Santa Cruz Biotechnology), mouse monoclonal antibody to ␣-SMA (1:100, clone 1A4, Abcam), mouse monoclonal to HCCM (1:100, clone 3– 48, Abcam), and goat polyclonal antibody to mouse EphB4 (1:100, R&D Systems AF446). Primary antibodies were incubated at 4°C overnight with tissue sections. Isotype antibodies served as negative controls. We performed image quantification using Metamorph as previously described (28). Quantification of vascular remodeling and density. To determine volume density of venous and arterial intima, media, and adventitia in human lung samples, slides were first stained with Russell-Movat pentachrome stain. Slides were then scanned with the Aperio ImageScope system (Vista), and the free stereological software Stepanizer (http://www.stepanizer.com) was used to overlay a 256-point grid (for assessment of the vascular structures) with a 16-point (coarse) grid for subsampling alveolar septae as previously described (31). (LVEDP). Cardiac output was measured either by an ultrasonic flow probe (Transonic, Transonic Systems) or by intravenous injection of cardiogreen dye (Sigma, St. Louis, MO) and measured by densitometry (3). Systemic arterial, central venous, pulmonary arterial, LVEDP, and cardiac output were continuously registered by use of the software package LabChart 7. Pulmonary vascular resistance was calculated by the standard equation: (mean pulmonary arterial pressure ⫺ left ventricular end-diastolic pressure)/cardiac output. Laser capture microdissection. Fresh lung samples were inflated with OCT compound (Sakura, Torrance, CA) and immediately flash frozen in liquid nitrogen. Individual 15-m sections were cut by cryostat at ⫺20°C, adhered to LCM membrane slides, stained with crystal violet and eosin for contrast, and immediately dissected by use of LCM (Molecular Machines and Industries, Zurich, Switzerland). Approximately 20 veins, arteries, and parenchymal samples were obtained per slide. RNA purification and RT-PCR. RNA was purified from LCM dissected lung tissue by use of an Arcturus PicoPure kit and following the manufacturer’s instructions (Applied Biosystems, Grand Island, NY). Isolated RNA was further purified by DNase digestion by using a commercial kit and following the manufacturer’s instructions (Qiagen). cDNA was made by using RT First strand kit (Qiagen) for later use in the RT2 PCR arrays, or reverse-transcription iScript cDNA synthesis kit (Bio-Rad, Hercules, CA) for RT-PCR. RT2 PCR array (Mouse Angiogenesis, PAMM-024A) (SABiosciences, Valencia, CA) or gene-targeted RT-PCR using commercial primer sets to uPAR or EphB4 (Invitrogen, CA) and cyclophilin A (Applied Biosystems) were performed on an Applied Biosystems 7300 real-time PCR system. RT-PCR conditions included initial denaturation at 95°C for L728 PULMONARY VEINS AND LEFT HEART DISEASE In rat (sham and AOB) and human (control and PH-LHD) lung, tissue morphometric analyses were applied to obtain media and venous cardiac myocyte sheath total wall fractional thickness as previously described (10). These measurements were performed on circular vascular profiles by use of immunofluorescent staining to ␣-SMA, HCCM, and uPAR. The external and internal media radii and intima radius were calculated by using r ⫽ A/ based on the measured area within each perimeter. For each case, 3 to 11 arteries were examined (median 6, interquartile range 4 –9). Statistics. Statistical analyses were performed with Sigma Stat Version 2.03 (IBM, Armonk, NY). Data are presented as means ⫾ SD for normally distributed data. Differences between groups were assessed by Student’s t-test. Correlation between two variables was determined by using the Pearson’s product-moment correlation coefficient. P values less than 0.05 were considered statistically significant. Backfilling of pulmonary veins. To reliably identify parenchymal pulmonary veins, we modified a previously described fluorescent microangiography technique (8) by retrogradely infusing fluorescein-tagged beads through the left heart of mice and rats (Fig. 1A). This approach led to venous casting extending from large to postcapillary small veins and venules, without leakage into capillaries or pulmonary arteries. We confirmed the accuracy of this technique by simultaneous immunostaining for EphB4, a known and reliable marker of mouse pulmonary vein endothelium (but not of rat or human veins) (Fig. 1B). The bead marking of pulmonary veins highly correlated with expression of EphB4 (Pearson’s correlation coefficient ⫽ 0.9). The venous backfilling technique was also adapted with similar efficacy to rat lungs (Fig. 1C). The Fig. 1. A: backfilling pulmonary veins with fluorescent tags is accomplished by first flushing the pulmonary circulation with heparin and PBS. The aorta is then clamped and a mixture of 1% agarose and 1-m fluorescent beads is injected in to the left ventricle under ⱕ20 mmHg pressure. B: mouse vein (ve; arrows) backfilled with fluorescent beads and counterstained for EphB4 (arrows) (N ⫽ 5). C: rat lungs. Scale bars ⫽ 100 m. Green fluorescence ⫽ beads backfilled into the veins; blue ⫽ DAPI. AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00186.2013 • www.ajplung.org Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on May 6, 2017 RESULTS cardiomyocyte sheath extends proximally from the left atrium around pulmonary veins and envelops the ␣-SMA-positive venous smooth muscle cells of the larger veins; we relied on the identification of this sheath to aid in confirmation of venous specific casting. The extent of this sheath depends on the species and size of animal studied, but in general it extends out to vessels of 100 –150 m in diameter in rats (16, 20, 24). Laser capture microdissection and identification of pulmonary vein markers. To identify candidate molecular markers of pulmonary veins, we used LCM to specifically isolate pulmonary vein mRNA, once the veins were identified by venous backfilling in mice (Fig. 2A). Samples of pulmonary arteries and alveolar septa that contain capillaries were processed in parallel to the pulmonary veins. Notwithstanding that mRNA extraction procedures and preservation were optimized in fresh frozen lung samples, the quality of the returned lung mRNA did not comply with the requirement for reproducible microarray analyses. It could nevertheless be used for accurate mRNA quantification by RT-PCR. We therefore employed the RT2 PCR array (SABiosciences) to screen for expressed candidate genes. Though technically challenging and laborious, this approach led us to identify a number of genes that were differentially expressed by pulmonary veins when compared with arteries/parenchyma obtained from selected lung samples. Importantly, we were able to detect an increased venous expression of EphB4 and, particularly, the urokinase plasminogen activator (uPA), whereas pulmonary arteries and alveolar septae expressed preferentially ephrinB2 (Fig. 2B and Table 2) (18). These results were confirmed on the basis of specific mRNA RT-PCR using target-specific primers sets. Indeed, we also found uPAR upregulated in pulmonary veins compared PULMONARY VEINS AND LEFT HEART DISEASE L729 with arteries or lung parenchyma harvested from selected lungs (Fig. 2C). Subsequent immunofluorescence staining of uPAR in mouse lungs confirmed its specificity for the pulmonary veins over arteries (Fig. 3A). uPAR staining pattern was remarkably similar between the mouse, rat, and human lung tissues (Fig. 3B). As previously described, uPAR expression was observed in the pulmonary bronchial epithelium and in scattered cells throughout the alveolar spaces. In regard to the pulmonary circulation, uPAR was largely absent from pulmonary arteries, but strongly expressed in pulmonary veins (Fig. 3B). Moreover, uPAR expression in pulmonary veins was also strongly present in the cardiomyocyte sheath of large veins and in the intima (Fig. 3B, right top) and media of more distal veins in both rats and humans (Fig. 3B). Quantification of uPAR immunofluorescence intensity confirmed the preferential expression of uPAR in veins over arteries and was similar between human and rat lung tissue (Fig. 3C). With a leading molecular candidate to identify pulmonary veins, we then interrogated whether uPAR expression was preferentially restricted to pulmonary veins in experimental and human PH-LHD. Rat aortic banding and pulmonary vascular remodeling. The model based on AOB-induced PH-LHD has been used to describe the pulmonary venous and arterial remodeling and to identify mechanisms involved in the pathogenesis of PH following left ventricular dysfunction in rats (13, 14, 35). We therefore performed AOB or sham operations in rats, followed, 9 wk later, by invasive hemodynamic measurements and morphometric/morphological studies. AOB rats developed PH with left heart dysfunction, as indicated by significantly increased mean pulmonary arterial pressure, LVEDP, and biventricular enlargement (Fig. 4A, D–G) compared with sham animals. Overall, AOB rats closely approximated the hemodynamic characteristics observed in cohorts human PH-LHD patients, including our own patient lung samples (Table 1). To quantify the extent of pulmonary vascular remodeling, AOB and sham rats also underwent pulmonary venous backfilling at the time of tissue harvest. Marked vascular remodeling was observed in the lungs of AOB rats, specifically media thickening of pulmonary arteries and veins (Fig. 5A). Quantification of arterial and venous media fractional thickness confirmed these observations (Fig. 5, B and C). Notably, the venous cardiomyocyte sheath fractional thickness was not significantly increased in AOB animals (Fig. 5D). Given uPAR’s previously described role in vascular remodeling and cellular proliferation, we assessed uPAR expression AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00186.2013 • www.ajplung.org Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on May 6, 2017 Fig. 2. A: mouse lung demonstrating pulmonary venous backfilling (top, left), allowing reliable identification of pulmonary veins (top, right). Identified veins are outlined (top, right) then excised and isolated by laser capture microdissection (bottom sequence). Note absence of beads in bronchioles (br) and arteries (ar). Scale bar ⫽ 100 m. Green and red fluorescence ⫽ beads backfilled into the veins; blue ⫽ DAPI. B: qRT2 PCR array comparing mouse pulmonary veins to a composite of arteries and parenchyma (n ⫽ 1 each). C: RT-PCR of urokinase plasminogen activator receptor (uPAR) and EphB4 expression in mouse pulmonary veins vs. arteries and parenchyma (N ⫽ 2 each). uPA, urokinase plasminogen activator; EphB4, venous developmental marker; EphrinB2, arterial developmental marker. L730 PULMONARY VEINS AND LEFT HEART DISEASE in the pulmonary vasculature of AOB rats. Increased uPAR staining of pulmonary arteries was observed in AOB arteries compared with controls (Fig. 6, A and B). Quantification of immunofluorescent staining intensity confirmed these observations and also demonstrated that overall uPAR expression appeared unchanged in the pulmonary veins of AOB compared with sham controls (Fig. 6B). PCNA staining was significantly increased in the pulmonary veins of AOB rats and associated with uPAR expression, suggesting that uPAR, despite not changing in expression in pulmonary veins, appears associated with increased cellular proliferation and possible vascular remodeling in PH-LHD (Fig. 6, C and D). Similar results were obtained with assessment of proliferating vascular arterial cells (Fig. 6E). Vascular remodeling in PH-LHD. To expand on the relevance of our experimental data with AOB rats, we collected lungs from patients with PH-LHD when subjected to LVAD placement. Patient characteristics are described in detail in Table 1. To determine the nature of vascular remodeling in patients with PH-LHD, we evaluated volume of media, intima, and adventitia relative to alveolar septae (used as reference space) in pulmonary arteries and veins of PH-LHD and control subjects (Fig. 7). We identified pulmonary veins using their typical topographical location in interlobular septa and in parenchymal branches communicating with interlobular vessels. We were not able to use uPAR expression in these human samples to specifically identify pulmonary veins since, as observed in AOB-rats, uPAR was similarly expressed between remodeled pulmonary veins and arteries in the setting of PH-LHD (Fig. 7, A and B). Media volume density of control cases was significantly higher in arteries than veins (Fig. 7C, P ⬍ 0.05). Intima and adventitia volume density, however, did not appear significantly different in veins and arteries of controls (Fig. 7, D and E). Media, intima, and adventitia volume density was significantly increased in both arteries and veins of PH-LHD cases compared with controls (Fig. 7, C–E). Notably, there were no significant differences in volume densities of media, intima, or adventitia in veins compared with arteries of PH-LHD patients. Pulmonary arteries of PH-LHD patients typically demonstrated increased medial and intimal thickness, nearly obliterating the vessel lumen in selected cases (Fig. 8, A and B). Venous histopathological patterns in PH-LHD lungs typified “arterialization”; i.e., greater delineation of the medial, intimal, and adventitial layers, presence of a double elastic lamina, muscularization of the vessel, and increased medial and intimal thickness (Fig. 8, C and D). Three PH-LHD/LVAD subjects went on to receive a second lung biopsy at the time of heart transplantation. Pre/post LVAD hemodynamics and vascular remodeling parameters for these three patients are summarized AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00186.2013 • www.ajplung.org Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on May 6, 2017 Fig. 3. A: quantification of uPAR-positive veins and arteries in normal mouse lungs. B: representative images of uPAR immunofluorescence in lung vessels of control rats (top row) and humans (middle and bottom rows). uPAR expression (arrows) is similar between rat and human samples and is present in bronchiolar epithelium and veins but absent in arteries (A). Detailed high magnification shown in bottom row for human lungs, with uPAR in red, CD31 or SMA in green, and DAPI in blue. C: uPAR IF intensity in arteries and veins of sham-operated rats (N ⫽ 4) and control human tissue (N ⫽ 4) (*P ⬍ 0.05). SMA, ␣-smooth muscle actin. Scale bar: 100 m. PULMONARY VEINS AND LEFT HEART DISEASE L731 in Fig. 9. These preliminary data suggest that, in some patients, improvement in hemodynamic parameters upon treatment may parallel decreases of venous remodeling. DISCUSSION In the present study, we found that pulmonary veins can be reliably identified by venous backfilling. On the basis of this methodological tool and using laser capture microdissection, we isolated venous RNA for identification of vein markers; this approach led us to identify and subsequently confirm the expression of uPAR as a novel vascular marker of normal mouse, rat, and human pulmonary veins. To study uPAR and venous remodeling in pulmonary vascular disease, we evaluated vascular remodeling and uPAR expression in lung tissue samples from PH-LHD patients and the AOB rat model of PH-LHD. Unlike in control tissues, uPAR expression was increased in pulmonary arteries, similar to the level in pulmonary veins, of both PH-LHD and AOB lungs. Although uPAR expression was similar in control and AOB veins, it was significantly associated with PCNA-positive cells, suggesting a possible role of uPAR in vascular remodeling. Intima, media, and adventitia were significantly remodeled in veins and arteries to similar degrees, in both human PH-LHD and rat AOB lungs. The study of pulmonary veins has been limited by the difficulty of their reliable identification, through histological, anatomic, or molecular means, in the lung parenchyma. Recently, EphB4, a developmental venous marker, has been shown to identify pulmonary veins in adult mice (4). However, this finding may not be generalizable to other organisms: careful embryological studies in humans have demonstrated overlap of EphB4 expression between pulmonary veins and arteries (12), whereas EphB4 does not appear to be present in the pulmonary veins of the rat (data not shown, and personal communication with Slaven Crnkovic, Ludwig Boltzmann Institute for Lung Vascular Research, Graz, Austria). Cardiomyocyte markers, such as heavy chain cardiac myosin, are specific to the venous cardiomyocyte sheath, but this structure does not extend deep into the parenchyma of larger animals, especially humans (2, 24). We were also unable to detect protein of CoupTF2, an orphan nuclear receptor that suppresses Notch AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00186.2013 • www.ajplung.org Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on May 6, 2017 Fig. 4. Pulmonary hemodynamics and cardiac parameters of aortic banded (AOB) rats. Mean pulmonary arterial pressures (mPAP; A), cardiac output (CO; B), pulmonary vascular resistance (PVR; C), left ventricular end-diastolic pressure (LVEDP; D), right ventricle weight/body weight (LV/BW; E), and left ventricle weight/body weight (F). *P ⬍0.05, **P ⬍ 0.01; N ⫽ 4 – 6 per group. G: a typical heart and lungs 9 wk after aortic banding (arrow). Catheter courses through the pulmonary artery. Note the biventricular enlargement (top and top right arrowheads) and marked left (left arrowhead) and right (bottom right arrowhead) atrial dilation. L732 PULMONARY VEINS AND LEFT HEART DISEASE signaling and establishes venous identity during development (36), in the pulmonary veins of adult animals (data not shown). Consequently, we relied on an imaging with fluorescent backfilling of veins and embarked on a screening approach to identify novel pulmonary venous molecular markers. Candidate markers were identified by LCM and RT2-PCR arrays and confirmed by RT-PCR of pulmonary veins compared with arteries and parenchyma obtained from selected lungs. Several of these candidate markers, such as pigment epithelium-derived factor (PEDF), CoupTF2, uPA, and uPAR, have known biological significance in the vasculogenesis and remodeling of veins. For instance, PEDF, a secreted serpin, has antiangiogenic properties that have been described in umbilical venous endothelial cells (1). Preliminary experiments, however, indicated that uPAR protein was much more abundant in pulmonary veins compared with arteries and was consequently further investigated as a marker of pulmonary veins. In control mice, rats, and human lung tissues, uPAR expression could be detected by immunofluorescence primarily in the media and less so in the intima and was specific to pulmonary veins. Knockout of uPAR has been shown to attenuate PH due to hypoxia (23). Given the known role of uPAR in smooth muscle and endothelial cell signaling promoting proliferation, migra- tion, and vascular remodeling, we investigated uPAR expression patterns in PH-LHD (19, 22, 30). PH-LHD is the most common cause of PH in the developed world and several orders of magnitude more common than pulmonary arterial hypertension (PAH) (27, 29). It results in significant morbidity and mortality, and yet there are no available treatments that target the underlying pulmonary vascular disease. Experimental models of PH-LHD have indicated increased vascular resistance of both the pulmonary arterial and venous beds (15). Consequently, the failure of PAH medications in PH-LHD may, in part, be due to the presence of venous remodeling and the relative importance of pathogenic cellular signaling cascades in both the arterial and venous systems. Consistent with this hypothesis, and with the aforementioned observations of increased pulmonary vascular resistance in the arterial and venous compartments, we observed significant and similar remodeling in the pulmonary veins and arteries of AOB rats and patients with PH-LHD. We also found that uPAR expression remained elevated in pulmonary veins and was significantly associated with cellular proliferation. PhosphoERK 1/2 was also elevated in both the pulmonary arteries and veins of AOB and PH-LHD patients (data not shown), possibly serving as an intermediary signaling relay between uPAR AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00186.2013 • www.ajplung.org Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on May 6, 2017 Fig. 5. Pulmonary vascular remodeling in AOB rats compared with sham animals. A: media thickening (arrows) in pulmonary arteries and pulmonary veins of AOB animals. Vessels from sham animals are below for comparison. Assessments of pulmonary artery media fractional thickness (B) and pulmonary vein media fractional thickness (C) in AOB vs. sham animals. N ⫽ 6 per group. HCCM, heavy chain cardiac myosin (**P ⬍ 0.01, ***P ⬍ 0.005). D: assessment of venous cardiomyocyte sheath fractional thickness in sham and AOB rats. There was no significant difference between the 2 groups [not significant (NS); P ⫽ 0.28]. N ⫽ 4 – 6 per group. Scale bar: 100 m. PULMONARY VEINS AND LEFT HEART DISEASE L733 expression and vascular remodeling, consistent with its previously described roles in uPAR signal transduction (22). Unlike in control lungs, uPAR expression was not exclusive to the pulmonary veins, being also elevated in the pulmonary arteries of AOB rats and PH-LHD patients. Interestingly, we noted in a selected group of patients who underwent sequential biopsy that improvement in hemodynamics with LVAD placement correlated with decreased arterial and venous remodeling, hence suggesting that these changes can be reversible. When considered together, these data suggest that a convergence of pathogenic processes may be involved in both pulmonary vein and arterial remodeling in PH-LHD. Future investigations to characterize the role of uPAR in this vascular remodeling may employ small molecule inhibitors to uPAR, such as IPR-803 (17), or related molecules involved in signaling and uPAR signal transduction, such as uPA, plasminogen activator inhibitor-1 (PAI-1), or ERK 1/2 (22). Given the limitations in the knowledge of pulmonary veins and of the pathology and pathobiology of venous PH, our study focused largely on discovery and validation of molecular expression pattern and structural alterations. These findings relied on integrating methods to visualize pulmonary veins, methods for assessment of gene expression, modeling of PHLHD in rats, and, finally, for the first time, a stereological assessment of PH-LHD in a unique cohort of patient samples. We identified a number of potential candidate molecular markers of pulmonary veins and confirmed uPAR’s specificity to pulmonary veins in control mouse, rat, and human lungs. We also confirmed the prior observations that vascular remodeling in both the arteries and veins, and so-called arterialization of the veins, in patients with advanced PH-LHD. Furthermore, we found that the arteries and veins share similarities in the degree of remodeling of the intima, media, and adventitia. We anticipate that the present work will form the basis of future mechanistic interventions and further delineation of molecular targets to therapeutically target pulmonary veins remodeling and pulmonary venous hypertension. GRANTS This research was funded by The Cardiovascular Medical Research and Education Fund and RC1 HL 100849 (to R. M. Tuder), HL 102478 (to P. P. A. Mammen), and KO8 HL 105536 and Parker B. Francis Fellowship award (to B. B. Graham). DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the author(s). AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00186.2013 • www.ajplung.org Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on May 6, 2017 Fig. 6. A: representative images of uPAR IF (arrows) in lung arteries and veins of AOB and sham rats. B: quantification of uPAR IF intensity in sham and AOB rat vessels. N ⫽ 3– 4 per group. C: representative image of proliferating cell nuclear antigen (PCNA) immunofluorescence (arrows) in rat lung veins. Note the high number of PCNA-positive nuclei among the HCCM negative cells. Quantification of PCNA-positive cells, or PCNA- and uPAR-positive cells, divided by total DAPI-positive cells in both sham and AOB rat pulmonary veins (D) and arteries (E) (N ⫽ 4). *P ⬍ 0.01 in D; P ⬍ 0.5 in E. Scale bar ⫽ 100 m. L734 PULMONARY VEINS AND LEFT HEART DISEASE Fig. 8. Characteristic histopathological findings in control subjects (A and C) and patients with PHLHD (B and D) (Russel-Movat pentachrome stains). A: pulmonary arteries (arrows) with normal intima and media in a control subject. B: arterial intima (arrowhead) and media (arrow) thickening with luminal narrowing in PH-LHD. C: pulmonary vein in a control subject with typical poorly organized media (arrows) and thin intima (arrowhead). D: venous intima (arrowheads) and media (arrows) thickening. Note “arterialization” of vein with double elastic lamina. Scale bars ⫽ 100 m. AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00186.2013 • www.ajplung.org Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on May 6, 2017 Fig. 7. A: representative images of uPAR IF (arrows) in lung arteries and veins of control subjects and patients with pulmonary hypertension with left heart disease (PH-LHD). B: uPAR IF intensity in PH-LHD patients and controls. *P ⬍ 0.05. Scale bar ⫽ 100 m. Assessment of media (C), intima (D), and adventitia (E) volume density relative to septa of both arteries and veins of control subjects (N ⫽ 19) and patients with PH-LHD (N ⫽ 22). Student’s t-test, *P ⬍ 0.05, ***P ⬍ 0.001. PULMONARY VEINS AND LEFT HEART DISEASE L735 AUTHOR CONTRIBUTIONS J.M.H., B.B.G., W.M.K., and R.M.T. conception and design of research; J.M.H., B.B., X.L., A.G., P.P.M., E.S., M.R.G., E.P., M.P., L.S., B.B.G., and R.M.T. performed experiments; J.M.H., B.B.G., and W.M.K. analyzed data; J.M.H., B.B.G., W.M.K., and R.M.T. interpreted results of experiments; J.M.H., B.B.G., W.M.K., and R.M.T. prepared figures; J.M.H., B.B.G., W.M.K., and R.M.T. drafted manuscript; J.M.H., B.B.G., W.M.K., and R.M.T. edited and revised manuscript; J.M.H., B.B., X.L., A.G., P.P.M., E.S., E.P., M.P., L.S., B.B.G., W.M.K., and R.M.T. approved final version of manuscript. REFERENCES 1. Aparicio S, Sawant S, Lara N, Barnstable CJ, Tombran-Tink J. Expression of angiogenesis factors in human umbilical vein endothelial cells and their regulation by PEDF. Biochem Biophys Res Commun 326: 387–394, 2005. 2. Arnstein C. Zur Kenntnis der quergestreiften Muskulatur in den Lungenvenen. Zentbl Med Wiss 692–694, 1877. 3. Coleman TG. Cardiac output by dye dilution in the conscious rat. J Appl Physiol 37: 452–455, 1974. 4. Crnkovic S, Hrzenjak A, Marsh LM, Olschewski A, Kwapiszewska G. Origin of neomuscularized vessels in mice exposed to chronic hypoxia. Respir Physiol Neurobiol 179: 342–345, 2011. 5. Demello DE, Reid LM. Embryonic and early fetal development of human lung vasculature and its functional implications. Pediatr Dev Pathol 3: 439 –449, 2000. 6. Dorfmuller P, Humbert M, Perros F, Sanchez O, Simonneau G, Muller KM, Capron F. Fibrous remodeling of the pulmonary venous system in pulmonary arterial hypertension associated with connective tissue diseases. Hum Pathol 38: 893–902, 2007. 7. Dorfmuller P, Montani D, Humbert M. Beyond arterial remodelling: pulmonary venous and cardiac involvement in patients with systemic sclerosis-associated pulmonary arterial hypertension. Eur Respir J 35: 6 –8, 2010. 8. Dutly AE, Kugathasan L, Trogadis JE, Keshavjee SH, Stewart DJ, Courtman DW. Fluorescent microangiography (FMA): an improved tool to visualize the pulmonary microvasculature. Lab Invest 86: 409 –416, 2006. 9. Gao Y, Zhou H, Ibe BO, Raj JU. Prostaglandins E2 and I2 cause greater relaxations in pulmonary veins than in arteries of newborn lambs. J Appl Physiol 81: 2534 –2539, 1996. 10. Graham BB, Mentink-Kane MM, El-Haddad H, Purnell S, Zhang L, Zaiman A, Redente EF, Riches DW, Hassoun PM, Bandeira A, Champion HC, Butrous G, Wynn TA, Tuder RM. Schistosomiasisinduced experimental pulmonary hypertension: role of interleukin-13 signaling. Am J Pathol 177: 1549 –1561, 2010. 11. Guazzi M, Arena R. Pulmonary hypertension with left-sided heart disease. Nat Rev Cardiol 7: 648 –659, 2010. 12. Hall SM, Hislop AA, Haworth SG. Origin, differentiation, and maturation of human pulmonary veins. Am J Respir Cell Mol Biol 26: 333–340, 2002. 13. Hentschel T, Yin N, Riad A, Habbazettl H, Weimann J, Koster A, Tschope C, Kuppe H, Kuebler WM. Inhalation of the phosphodiesterase-3 inhibitor milrinone attenuates pulmonary hypertension in a rat model of congestive heart failure. Anesthesiology 106: 124 –131, 2007. 14. Hoffmann J, Yin J, Kukucka M, Yin N, Saarikko I, Sterner-Kock A, Fujii H, Leong-Poi H, Kuppe H, Schermuly RT, Kuebler WM. Mast cells promote lung vascular remodelling in pulmonary hypertension. Eur Respir J 37: 1400 –1410, 2011. 15. Huang W, Kingsbury MP, Turner MA, Donnelly JL, Flores NA, Sheridan DJ. Capillary filtration is reduced in lungs adapted to chronic heart failure: morphological and haemodynamic correlates. Cardiovasc Res 49: 207–217, 2001. 16. Jones WK, Sanchez A, Robbins J. Murine pulmonary myocardium: developmental analysis of cardiac gene expression. Dev Dyn 200: 117–128, 1994. 17. Khanna M, Wang F, Jo I, Knabe WE, Wilson SM, Li L, Bum-Erdene K, Li J, Sledge W, Khanna R, Meroueh SO. Targeting multiple conformations leads to small molecule inhibitors of the uPAR. uPA protein-protein interaction that block cancer cell invasion. ACS Chem Biol 6: 1232–1243, 2011. 18. Kim YH, Hu H, Guevara-Gallardo S, Lam MT, Fong SY, Wang RA. Artery and vein size is balanced by Notch and ephrin B2/EphB4 during angiogenesis. Development 135: 3755–3764, 2008. 19. Kiyan J, Kiyan R, Haller H, Dumler I. Urokinase-induced signaling in human vascular smooth muscle cells is mediated by PDGFR-beta. EMBO J 24: 1787–1797, 2005. 20. Klika E, Zajicova A, Votavova B. [Participation of the myocardium in the formation of the pulmonary vein wall of small mammals (a quantitative study)]. Arkh Anat Gistol Embriol 82: 81–83, 1982. 21. Lantuejoul S, Sheppard MN, Corrin B, Burke MM, Nicholson AG. Pulmonary veno-occlusive disease and pulmonary capillary hemangiomatosis: a clinicopathologic study of 35 cases. Am J Surg Pathol 30: 850 –857, 2006. AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00186.2013 • www.ajplung.org Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on May 6, 2017 Fig. 9. Right atrial pressures (RAP; A) and mPAP (B) prior to left ventricular assist device (LVAD) placement and 2 days post-LVAD placement in the 3 PH-LHD patients who received a second lung biopsy at time of transplantation. Volume density of venous media (C) and intima (D) and arterial media (E) and intima (F) relative to alveoli both before LVAD placement and at time of subsequent transplant (P ⫽ NS, paired t-test). Lines represent the same individual throughout graphs: light gray, 38-yr-old male transplanted at 27 days post-LVAD; medium gray, 60-yr-old female transplanted at 41 days post-LVAD; black, 60-yr-old female transplanted 7 mo post-LVAD. L736 PULMONARY VEINS AND LEFT HEART DISEASE 30. 31. 32. 33. 34. 35. 36. Langleben D, Nakanishi N, Souza R. Updated clinical classification of pulmonary hypertension. J Am Coll Cardiol 54: S43–S54, 2009. Smith HW, Marshall CJ. Regulation of cell signalling by uPAR. Nat Rev Mol Cell Biol 11: 23–36, 2010. Stacher E, Graham BB, Hunt JM, Gandjeva A, Groshong SD, McLaughlin VV, Jessup M, Grizzle WE, Aldred MA, Cool CD, Tuder RM. Modern age pathology of pulmonary arterial hypertension. Am J Respir Crit Care Med 186: 261–272, 2012. Wagenvoort CA, Wagenvoort N. Normal circulation of the lungs. In: Pathology of Pulmonary Hypertension, edited by Wagenvoort CA, Wagenvoort N. New York: Wiley, 1977, p. 1–8. Wagenvoort CA, Wagenvoort N. Pulmonary venous hypertension. In: Pathology of Pulmonary Hypertension, edited by Wagenvoort CA, Wagenvoort N. New York: Wiley, 1977, p. 177–216. Walch L, Labat C, Gascard JP, de Montpreville V, Brink C, Norel X. Prostanoid receptors involved in the relaxation of human pulmonary vessels. Br J Pharmacol 126: 859 –866, 1999. Yin J, Kukucka M, Hoffmann J, Sterner-Kock A, Burhenne J, Haefeli WE, Kuppe H, Kuebler WM. Sildenafil preserves lung endothelial function and prevents pulmonary vascular remodeling in a rat model of diastolic heart failure. Circ Heart Fail 4: 198 –206, 2011. You LR, Lin FJ, Lee CT, DeMayo FJ, Tsai MJ, Tsai SY. Suppression of Notch signalling by the COUP-TFII transcription factor regulates vein identity. Nature 435: 98 –104, 2005. AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00186.2013 • www.ajplung.org Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on May 6, 2017 22. LaRusch GA, Mahdi F, Shariat-Madar Z, Adams G, Sitrin RG, Zhang WM, McCrae KR, Schmaier AH. Factor XII stimulates ERK1/2 and Akt through uPAR, integrins, and the EGFR to initiate angiogenesis. Blood 115: 5111–5120, 2010. 23. Levi M, Moons L, Bouche A, Shapiro SD, Collen D, Carmeliet P. Deficiency of urokinase-type plasminogen activator-mediated plasmin generation impairs vascular remodeling during hypoxia-induced pulmonary hypertension in mice. Circulation 103: 2014 –2020, 2001. 24. Ludatscher RM. Fine structure of the muscular wall of rat pulmonary veins. J Anat 103: 345–357, 1968. 25. Montani D, Kemp K, Dorfmuller P, Sitbon O, Simonneau G, Humbert M. Idiopathic pulmonary arterial hypertension and pulmonary venoocclusive disease: similarities and differences. Semin Respir Crit Care Med 30: 411–420, 2009. 26. Murali S. Pulmonary hypertension in heart failure patients who are referred for cardiac transplantation. Adv Pulm Hypertens 5: 30 –35, 2006. 27. Oudiz RJ. Pulmonary hypertension associated with left-sided heart disease. Clin Chest Med 28: 233–241, x, 2007. 28. Petrache I, Natarajan V, Zhen L, Medler TR, Richter AT, Cho C, Hubbard WC, Berdyshev EV, Tuder RM. Ceramide upregulation causes pulmonary cell apoptosis and emphysema-like disease in mice. Nat Med 11: 491–498, 2005. 29. Simonneau G, Robbins IM, Beghetti M, Channick RN, Delcroix M, Denton CP, Elliott CG, Gaine SP, Gladwin MT, Jing ZC, Krowka MJ,