Western blot analysis of lung Tie2 and endothelial NO synthase (eNOS) protein revealed a reduction of ?50% in expression levels in Tie2^+/? compared with WT mice (P < 0.01 and 0.05, respectively; Fig. 1, A and B). Interestingly, basal protein levels of Ang1 in the lung were slightly increased in Tie2^+/? compared with WT mice (P < 0.01), whereas Ang2 levels were reduced (P < 0.01), resulting in a more than twofold increase in the ratio of lung Ang1 to Ang2 levels (P < 0.05; Fig. 1 C). Under basal conditions, there was a modest but significant difference in RVSP between WT and Tie2^+/? mice (P < 0.05), with 13% of the Tie2^+/? mice demonstrating a RVSP measurement >37 mmHg (two standard deviations above mean RVSP of WT mice; ?², P < 0.05 compared with WT mice; Fig. 2 A); however, no significant change in baseline systemic systolic blood pressure (SBP; Fig. 2 B) or heart rate (unpublished data) was observed. A marked reduction in microvascular perfusion in the lungs of Tie2^+/? compared with WT mice was revealed after fluorescent microangiography (Fig. 2 C). Furthermore, this perfusion abnormality was more pronounced in mice with higher RVSPs. To further characterize their susceptibility to PAH, WT and Tie2^+/? mice were subjected to 1 wk of hypoxic exposure (8?10% O₂). Chronic hypoxia resulted in a significant increase in RVSP and right ventricle (RV) hypertrophy (P < 0.05 and 0.01, respectively, compared with normoxic controls), but there was no significant difference between the hypoxic WT and Tie2^+/? mice observed (Fig. S1).

To determine the effect of well-known pulmonary hypertensive stimuli, RVSP (Fig. 3 A) and RV hypertrophy (Fig. 3 B) were assessed in WT and Tie2^+/? mice after 1 wk of 5-HT infusion and 1 or 2 wk of IL-6 delivery. Treatment with 5-HT had no significant effect on RVSP in WT mice, although there was a trend (P < 0.1) toward an increase. However, 5-HT?treated Tie2-deficient mice exhibited a substantial increase in RVSP compared with their respective saline-treated controls (40 ? 4 vs. 25 ? 1 mmHg; P < 0.01). This increase was also significantly greater than that seen in WT mice exposed to 5-HT (27 ? 2 mmHg; P < 0.05). Although no significant effect on RVSP was observed in any group after 1 wk of IL-6 delivery, 2 wk of exposure to IL-6 induced a significant elevation in RVSP in Tie2^+/? mice compared with IL-6?treated WT mice (31 ? 3 vs. 22 ? 1 mmHg; P < 0.05). A trend (P < 0.1) toward an increase in RVSP was also observed in Tie2^+/? mice compared with their saline-treated controls. No significant change in RV hypertrophy was observed in either the WT or Tie2^+/? mice during the short period of treatment with 5-HT. However, a significant increase in RV hypertrophy was observed in Tie2^+/? compared with WT mice after 2 wk of exposure to IL-6 (P < 0.05).

Western blot analysis was performed on lung homogenates from WT and Tie2^+/? mice to determine whether treatment with 5-HT or IL-6 altered endogenous Ang1 levels (Fig. 4, A and B). After 1 wk of exposure to 5-HT or IL-6, Ang1 levels were significantly and similarly decreased in WT and Tie2^+/? mice compared with their respective saline-treated control groups (P < 0.05). To further establish the association between 5-HT and Ang1, human pulmonary artery smooth muscle cells (SMCs) were exposed for 24 h to 5-HT (10^?8?10^?5 M) or a combination of equal concentrations of IL-6 and a soluble IL-6R (sIL-6R; 10, 50, and 100 ng/ml), because past studies (^{Modur et al., 1997}; ^{Ammit et al., 2007}) have demonstrated that the membrane-bound receptor may not be expressed in SMCs under in vitro conditions and that the soluble receptor is required to allow for IL-6 trans-signaling (Fig. S2). Ang1 secretion, as determined by ELISA performed on conditioned media, was reduced to ?60% in response to 5-HT or IL-6/sIL-6R (P < 0.01; Fig. 4, C and D). To investigate the implications of these observations on in vivo Tie2 activity, Western blot analysis of phosphorylated Tie2 was performed on lung homogenates (Fig. 5, A and B). Although in WT mice no change in Tie2 receptor activation was observed after 5-HT infusion or IL-6 injections, Tie2^+/? mice displayed a significant decrease in lung Tie2 receptor activation compared with their saline-treated groups, in parallel to the decreases in Ang1 levels demonstrated in Fig. 4 (A and B) (P < 0.05). Because baseline levels of eNOS were significantly down-regulated in Tie2^+/? mice (Fig. 1 B), we examined eNOS protein expression in WT and Tie2^+/? mice after 5-HT infusion (Fig. S3). Interestingly, although there was no significant reduction in lung eNOS protein levels after 5-HT exposure, a trend toward a decrease (P < 0.1) was observed in the Tie2-deficient group.

No terminal deoxynucleotidyl transferase?mediated dUTP nick-end labeling (TUNEL) staining or propidium iodide (PI) uptake was detected in saline-treated WT or Tie2^+/? mice (not depicted). In contrast, substantial TUNEL positivity was seen after 1 wk of exposure to 5-HT in Tie2-deficient but not in WT mice (7 vs. 0%; P < 0.01; Fig. 6 A). This increase in apoptosis was confirmed by in vivo perfusion of the pulmonary circulation with PI, which only stains the nuclei of nonviable cells. Again, only Tie2^+/? mice exposed to 5-HT demonstrated an increased uptake of PI (Fig. S4). Similarly, after 1 wk of exposure to IL-6, only Tie2-deficient mice demonstrated TUNEL staining within peripheral regions of the lung (6%), whereas no staining was seen in WT mice (P < 0.01; Fig. 6 B).

To further define the effect of decreased Tie2 receptor expression and signaling on EC survival in response to 5-HT, Tie2 was knocked down in human pulmonary artery ECs using specific silencing siRNA. Transfection with specific siRNA reduced Tie2 protein expression by ?50% by Western analysis, compared with control, scrambled siRNA?treated ECs (P < 0.01; Fig. 7 A). Furthermore, Tie2 silencing significantly enhanced EC apoptosis induced by 10 ?M 5-HT cultured in low serum (0.2% fetal bovine serum) compared with cells treated with scrambled siRNA (P < 0.01; Fig. 7 B).

Exposure to 5-HT for 1 wk did not result in any differences in the muscularization of pulmonary vessels <50 ?m in diameter in either WT or Tie2-deficient mice (unpublished data). However, Tie2^+/? mice exposed to 2 wk of IL-6 demonstrated a significant increase in muscularization compared with saline-treated Tie2^+/? and IL-6-treated WT mice (P < 0.01 for both comparisons; Fig. 8, A and B). Preliminary studies demonstrated no overt phenotypic differences in extracellular matrix composition between WT and Tie2^+/? mice after exposure to 5-HT or IL-6 (Fig. S5).

Protein expression of the human Ang1 transgene was detected in the mouse plasma of binary transgenic (BT) mice after induction by doxycycline withdrawal (Fig. S6). To define the effects of Ang1 overexpression on the development of PAH, RVSP and RV hypertrophy were determined in 11?12-wk-old Ang1 BT mice and non-BT (NBT) littermate controls that had been released from doxycycline suppression at 3 wk of age, and then received infusions of 5-HT or saline for 1 wk before end-study assessments. As presented in Fig. 9 A, there were no significant differences in RVSP between Ang1 BT and NBT mice receiving either saline or 5-HT infusions; if anything, RVSP tended to be reduced in BT mice, but this was not significant. As well, no differences in RV hypertrophy were observed between the groups after 1 wk of saline or 5-HT treatment (Fig. 9 B).

To further establish the role of EC apoptosis in the exaggerated pulmonary hypertensive response to 5-HT in vivo, we treated WT or Tie2-deficient animals with a pan-caspase inhibitor, Z-VAD. In WT mice, Z-VAD had no effect, whereas in Tie2^+/? mice, this treatment completely prevented the exaggerated increase in RVSP induced by 5-HT (P < 0.01; Fig. 10 A). In addition, no significant change in RV hypertrophy was present at 1 wk (Fig. 10 B). Furthermore, no TUNEL-positive cells were detected in the lungs of Tie2^+/? mice receiving both 5-HT and Z-VAD (not depicted).

All animal procedures were approved by the Animal Care Committee of St. Michael's Hospital. Breeding pairs to generate Tie2^+/? mice and Ang1 BT mice were provided by D.J. Dumont (University of Toronto, Toronto, Canada). WT littermates were used as controls in all experiments. Tie2^+/? mice were generated by crossing Tie2^+/? mice with WT mice. Ang1 transgenic mice were generated by breeding Enh-Tie2 tTA driver and pTET-Ang1 responder mice. Ang1 overexpression was suppressed throughout breeding by administering mice with drinking water containing 100 ?g/ml doxycycline (Sigma-Aldrich). After weaning at 3 wk of age, mice were released from doxycycline, allowing for Ang1 overexpression. NBT littermate mice were used as controls. Offspring genotypes were determined via PCR (see Genotyping).

Genomic DNA was extracted from mouse ear-notch samples, and PCR was performed using the buffers and PCR reaction mix from the REDExtract-N-Amp tissue PCR kit (Sigma-Aldrich), according to the manufacturer's instructions. A 0.5-?m mixture of the following primers was used in the PCR to detect the endogenous Tie2 (?150 bp) and mutated Tie2 (?450 bp): endogenous forward, 5?-AGCGCGTGGACCATGCGAGC-3?; endogenous reverse, 5?-CCATTGCTCAGCGGTGCTGTCCAT-3?; and mutant forward, 5?-AGGAGCAAGCTGACTCCACAG-3?. The PCR thermal cycling parameters included an initial denaturation at 94?C for 3 min, followed by 30 cycles consisting of denaturation at 95?C for 30 s, annealing at 54?C for 40 s, and extension at 72?C for 30 s According to this, a final extension at 72?C for 10 min was performed, and the amplified DNA was resolved in a 1.5% agarose gel. Ang1 BT mice were identified via PCR using the following primer pairs to detect the tetracycline transactivator (tTA; ?500 bp) and the human Ang1 transgene (hAng1; ?350 bp): tTA forward, 5?-CTCACTTTTGCCCTTTAGAA-3?; tTA reverse, 5?-GCTGTACGCGGACCCACTTT-3?; hAng1 forward, 5?-ATAGGAACCAGCCTCCTCTCT-3?; and hAng1 reverse 5?-AAGGACACTGTTGTTGGTGGTA-3?.

Cardiac puncture was performed on anesthetized mice, and blood was collected from both NBT and Ang1 BT mice. Plasma was isolated and stored at ?80?C. Human Ang1 expression in mouse plasma was determined by ELISA according to the manufacturer's instructions (R&D Systems).

Systemic SBP was determined by inserting a 1.4F Millar catheter (Millar Instruments, Inc.) into the left ventricle of anesthetized mice (200 mg/kg ketamine:10 mg/kg xylazine i.p.). Similarly, RVSP was determined in anesthetized mice by inserting the 1.4F Millar catheter into the RV. RV hypertrophy was assessed by evaluating the mass ratio of the RV to the left ventricle plus septum.

This technique was adapted from ^{Dutly et al. (2006)}. In brief, mice were anaesthetized and the pulmonary vessels were flushed through the RV with a flushing buffer (10 mM sodium phosphate, 127 mM sodium chloride, 10 mM EDTA, 5 U/ml heparin). A 10% mixture of 0.2 ?m of fluorescent yellow-green microbeads (FluoSpheres; Invitrogen) reconstituted in low-temperature melting point agarose was injected into the pulmonary circulation via the RV. Crushed ice was then placed atop the mice to enhance gelling of the microbead?agarose mixture. After this, the lungs were inflated with cold 4% paraformaldehyde (PFA), and the heart?lung block was fixed in 4% PFA for 48 h and transferred to PBS for storage. The whole top right lobe was subjected to fluorescent microscopy (1? magnification; Discovery V8 stereo microscope equipped with an AxioCam MR3 camera) and images were acquired using AxioVision LE 4.6 software (all from Carl Zeiss, Inc.).

This method was adapted from ^{Cho et al. (1995)}. 5 ?mol/kg PI (Sigma-Aldrich) was injected via the right external jugular vein of anesthetized mice (200 mg/kg ketamine:10 mg/kg xylazine) and allowed to circulate for 15 min before the mice were euthanized. The left atrium was transected, and the systemic and pulmonary circulations were flushed with PBS. Lungs were fixed in optimum cutting temperature and stored at ?80?C overnight. The next day, 60-?m-thick cross sections of the left lobe of the lungs were cryosectioned for confocal microscopy. Cells that did not exclude the PI were considered to be nonviable.

Apoptotic cells were detected in 5-?m-thick paraffin-embedded sections using a fluorescent TUNEL assay kit (Promega) according to manufacturer's instructions. Nuclear counterstaining was performed using TO-PRO-3 (1:2,500; Invitrogen), and sections were examined by confocal microscopy (40? magnification; Radiance 2100) and images were acquired using LaserSharp software (both from Bio-Rad Laboratories). The number of TUNEL-positive nuclei was determined by examining four randomly selected microscopic fields from each of four to five mice from each experimental group. The percentage of TUNEL positivity was calculated as follows: (number of TUNEL-positive nuclei)/(total number of nuclei) ? 100%.

Protein was extracted from the lungs of WT and Tie2^+/? mice and subjected to immunoblotting. Membranes were incubated overnight at 4?C with rabbit eNOS antibody (1:1,000; Cell Signaling Technology), rabbit Phospho-Tie2 antibody (1:1,000; Cell Signaling Technology), rabbit Tie2 antibody (1:1,000; Santa Cruz Biotechnology, Inc.), rabbit Ang1 antibody (1:1,000; Abcam), or goat Ang2 antibody (1:500; Santa Cruz Biotechnology, Inc.), and equivalent protein loading was demonstrated using mouse ?-actin antibody (1:5,000; Sigma-Aldrich).

5-?m-thick paraffin-embedded sections were stained with a Cy3-conjugated monoclonal ??smooth muscle actin antibody (1:200; Sigma-Aldrich), and nuclei were counterstained using TO-PRO-3 (1:2,500). Sections were examined by confocal microscopy (80? magnification; Radiance 2100) and images were acquired using LaserSharp software. The percentage of wall thickness was determined using the methodology by ^{Beppu et al. (2004)} and was as follows: % wall thickness = (WT1 + WT2)/(external diameter of vessel) ? 100%, where WT1 and WT2 refer to wall thicknesses measured at two points diametrically opposite to each other. For each vessel, the percentage of wall thickness was calculated as the average determined from four regions of measurement.

The Movat pentachrome staining was performed on 5-?m-thick paraffin-embedded sections to demonstrate extracellular matrix components such as fibrin, collagen, elastin, and proteoglycans/glycosaminoglycans. Vessels from four to five randomly selected microscopic fields (10? magnification) from each of three mice from each experimental group were analyzed using a bright-field upright light microscope (model E800) equipped with a digital camera (DXM 1200), and images were acquired using Act-1 software (all from Nikon).

Human pulmonary artery SMCs (Lonza, Ltd.) at 70?80% confluency were serum starved in basal medium with 0.1% BSA for 24 h and exposed to 5-HT (Sigma-Aldrich) at concentrations of 10^?8?10^?5 M for 24 h. In a separate experiment, serum-starved cells were exposed to a combination of human IL-6 and human sIL-6R (IL-6 + sIL-6R; Sigma-Aldrich), each at concentrations of 10, 50, and 100 ng/ml for 24 h. At the end of the experiments, supernatants were collected for ELISA and cells were harvested for protein quantification.

Protein was harvested from human pulmonary artery SMCs and human endothelial progenitor cells were cultured for 7 d (positive control). Membranes were incubated overnight at 4?C with rabbit IL-6R antibody (1:500; Santa Cruz Biotechnology, Inc.).

Human Ang1 levels from human pulmonary artery SMC supernatants were measured according to the manufacturer's instructions (R&D Systems).

siRNA-mediated inhibition of Tie2 gene expression (Santa Cruz Biotechnology, Inc.) in human pulmonary artery ECs (Lonza, Ltd.) was performed according to the manufacturer's instructions, and control cells were subjected to scrambled siRNA transfection. After siRNA treatment, cells were exposed to 10^?5 M 5-HT in basal medium with 0.2% FBS for 24 h. At the end of this treatment, cells were labeled with annexin V and PI (Annexin-V-FLUOS staining kit; Roche) and analyzed by flow cytometry to determine the percentage of cell death.

All data are presented as means ? SEM. For in vivo experiments, means of two groups were compared using either an unpaired t test, or when appropriate, the nonparametric Mann-Whitney U test. Differences among multiple means were determined by analysis of variance (ANOVA), followed by the Student-Newman-Keuls post-hoc analysis, or the nonparametric Kruskal-Wallis test followed by the Dunn's post-hoc analysis. For in vitro experiments, repeated-measures ANOVA was used when appropriate, followed by the Dunnett's post-hoc analysis.

Fig. S1 demonstrates that hypoxia did not elicit more significant PAH in Tie2^+/? compared with WT mice. Fig. S2 confirms that the membrane-bound IL-6R is not expressed in pulmonary arterial SMCs in culture. Fig. S3 demonstrates that lung eNOS protein levels are not significantly altered in WT or Tie2^+/? mice after exposure to 5-HT, although a trend toward a decrease in eNOS protein levels was present in the Tie2^+/? mice. Fig. S4 illustrates the presence of nonviable cells in the lungs of 5-HT?treated Tie2^+/? mice via in vivo PI uptake. Fig. S5 demonstrates the absence of any overt phenotypic differences after Movat pentachrome staining between WT and Tie2^+/? mice treated with 5-HT or IL-6. Fig. S6 confirms the expression of human Ang1 protein in the plasma of Ang1 BT mice. Online supplemental material is available at http://www.jem.org/cgi/content/full/jem.20090389/DC1.