Alterations of cellular bioenergetics in pulmonary artery endothelial cells

Clinical Characteristics.

Altogether, six controls, nine IPAH, and four PAH subjects were studied. PAEC were derived from donor lungs not used in transplantation and five IPAH and four PAH explanted lungs obtained at transplantation. Positron-emission tomography (PET)-computed tomography (CT) scans were conducted in three healthy female controls and four IPAH female patients. Clinical characteristics among subjects were similar [age (years): IPAH 45 ± 3, PAH 40 ± 4, control 35 ± 6; sex (female/male): IPAH 8/1, PAH 2/2, control 5/1; race (Caucasian/African American/Hispanic): IPAH 9/0/0, PAH 3/1/0, control 6/0/0]. Pulmonary hypertension was diagnosed by right heart catheterization performed for clinical care [pulmonary artery pressures (mmHg, 1 mmHg = 133 Pa): IPAH, systolic 94 ± 6, diastolic 41 ± 4, mean 59 ± 2; PAH, systolic 97 ± 12, diastolic 44 ± 14, mean 50 ± 6]. PAH was secondary to congenital heart diseases, except for one case of sarcoidosis. IPAH individuals were receiving vasodilators, anticoagulants, diuretics, digitalis, and/or oxygen by nasal cannula.

Decreased Oxygen Consumption in IPAH PAEC.

To initially evaluate cellular respiration (21, 22), oxygen consumption in IPAH and control PAEC was determined. Compared with control cells (n = 3), IPAH PAEC (n = 4) had less oxygen consumption for state 3 and state 4 respirations with glutamate-malate or succinate as substrate (Table 1). However, the coupling between oxygen consumption and ATP production, the respiratory control index, was similar among IPAH and controls.

Activity and Expression of Mitochondrial Complexes.

To determine causes of lower oxygen consumption, the activities of Complexes III and IV and the ability of mitochondria to reduce 3-(4,5-dimethylthiazolyl-2)-2,5-diphenyltetrazolium bromide (MTT) were determined (Fig. 1A and B). Activity of Complex IV, the terminal enzyme complex of the respiratory chain that catalyzes the transfer of electrons from reduced cytochrome c to molecular oxygen, was significantly lower in IPAH PAEC (n = 2) than in control (n = 2), whereas Complex III activity was similar among the controls and IPAH cells (nanomoles of cytochrome c per minute per 10⁶ cells: Complex IV, IPAH 9.83 ± 0.47, control 12.33 ± 0.73, P < 0.05, n = 9 replicate experiments, cells derived from each of two lungs from IPAH and controls in these series of experiments; Complex III, IPAH 5.65 ± 0.19, control 5.97 ± 0.22, P > 0.05, n = 4 replicate experiments) (Fig. 1A). MTT reduction by IPAH PAEC was strikingly lower than control, indicating less mitochondrial dehydrogenase activity (absorbance at 570 nm per 60,000 cells: IPAH 0.57 ± 0.03, control 0.75 ± 0.02, P < 0.05, n = 9 replicate experiments) (Fig. 1B).

Quantitation of mitochondria in electron microscopy images of cells (Fig. 1C) and Southern blot analysis of mtDNA (Fig. 1 D and E) revealed decreased mitochondrial numbers in IPAH PAEC compared with control (mitochondrial number per cell, IPAH 69 ± 7, control 191 ± 23, P < 0.01; mtDNA relative units per microgram of total DNA, IPAH 1 ± 0.16, control 2.29 ± 0.64, P < 0.05) but no discernible difference in mitochondrial morphology (Fig. 1F).

NO Production in IPAH PAEC.

IPAH patients have low levels of exhaled NO (3, 11), and IPAH PAEC produce lower total amounts of NO products than control cells in vitro (10). Here, NO production by IPAH PAEC (n = 3) detected by measuring nitrite (NO₂⁻) in the culture supernatants after ionomycin stimulation was also significantly lower than controls (n = 3, IPAH 0.30 ± 0.27, controls 3.91 ± 1.38 pmol/min per 10⁶ cells, P < 0.05). Ionomycin-stimulated nitrite (NO₂⁻) concentrations of IPAH and control cells were correlated to oxygen consumption for state 3 and state 4 respirations with glutamate-malate substrates (all P < 0.01) (Fig. 1G).

To test whether NO affects mitochondrial numbers and functions in IPAH cells, mitochondria numbers in electron microscopy images (Fig. 1C), mtDNA by Southern blot analysis (Fig. 1 D and E), and mitochondrial dehydrogenase activity were analyzed in IPAH cells (n = 3) after exposure to NO donors in culture. Mitochondrial numbers increased in IPAH PAEC after treatment with NO donors, DETA NONOate (detaNO), or S-nitroso-N-acetyl-d,l-penicillamine (SNAP) for 6 days compared with untreated cells (mitochondrial number/cell, untreated 69 ± 7, detaNO 93 ± 9, SNAP 90 ± 6, P < 0.01; mtDNA relative units per microgram of DNA, untreated 1 ± 0.16, detaNO 1.52 ± 0.47, SNAP 1.52 ± 0.40; P = 0.06 untreated vs. detaNO; P < 0.05 untreated vs. SNAP; n = 3) (Fig. 1 C–F). Mitochondrial dehydrogenase activity as measured by MTT also increased after NO donors (absorbance at 570 nm per 60,000 cells: untreated 0.42 ± 0.01, SNAP 0.49 ± 0.01, P < 0.01, n = 3), and Complex III-1 and cytochrome c protein levels were ≈2-fold greater after NO donors treatment (data not shown). Altogether, these data indicate that levels of NO in IPAH cells affect mitochondrial numbers and hence cellular bioenergetics in IPAH cells.

ATP Content in IPAH PAEC Under Normoxia (21% O₂) and Hypoxia (2% O₂).

On the basis of the lower oxygen uptake and decreased mitochondrial function, we reasoned that the IPAH cells might have lower levels of energy production (20, 26, 27). To test this hypothesis, ATP content in IPAH (n = 3) and control cells (n = 4) was quantitated (Fig. 2A). Under 21% O₂, ATP content in IPAH cells was similar to controls (P > 0.05). Under hypoxia of 2% O₂, ATP content in control cells dropped by >30% (P < 0.01), but ATP content in IPAH PAEC exhibited similar levels under normoxia (P > 0.05) (relative ATP content: normoxia, IPAH 0. 91 ± 0.03, control 1.00 ± 0.02; hypoxia, IPAH 0.81 ± 0.04, control 0.64 ± 0.03; n = 14 replicate experiments) (Fig. 2A).

Increased Glycolytic Rate in IPAH PAEC.

The greater tolerance to hypoxia indicates that IPAH PAEC have a lesser dependence on cellular respiration for energy than control cells and suggests a predominant anaerobic cellular energy source. Thus, the ³H-glucose glycolytic rate in IPAH was compared with control cells. The glycolytic rate of IPAH PAEC (n = 5) was significantly higher than controls (n = 3) (nanomoles per hour per 10⁶ cells, IPAH 15.4 ± 1.9, controls 6.2 ± 0.84, P < 0.01), and lactate release was ≈1.5-fold higher from IPAH PAEC (n = 3) compared with controls (n = 3), indicating that glucose metabolism was subserving the primary role for energy requirements. Interestingly, glycolytic rate of PAH PAEC (n = 4) was also significantly higher than controls (n = 3) (nanomoles per hour per 10⁶ cells, PAH 18.2 ± 1.3, controls 6.2 ± 0.84, P < 0.01) (Fig. 2B).

Glucose Metabolic Activities Determined by PET Scan of Lungs.

IPAH is pathologically characterized by proliferative vascular endothelial lesions in the size of ≈150 μm (1, 2). Because cellular energy metabolism in IPAH appeared reliant on glycolysis, a situation similar to alterations noted in cancer cell bioenergetics and used in cancer screening technologies, we reasoned that we might see a difference in the glucose analogue tracer [¹⁸F]fluoro-deoxy-d-glucose (FDG) uptake into pulmonary vascular endothelial cells, i.e., FDG-PET scan. The glucose analogue tracer [¹⁸F]FDG is taken up into cells by glucose transporters, phosphorylated to [¹⁸F]FDG-6-P in the presence of hexokinase, but unlike glucose, is trapped within cells with no further metabolism because of the lack of the hydroxyl group in the second carbon position (28). Here, although FDG-PET spatial resolution limits are at 6 mm³, given the extensive nature of the pulmonary vascular disease in IPAH, FDG-PET scan was used to evaluate whether glucose metabolism was increased diffusely in the lungs of four IPAH patients and three healthy controls (Fig. 2 C and D and supporting information (SI) Fig. 3). Lung standardized uptake values (SUV) of IPAH patients were significantly higher than healthy controls at [¹⁸F]FDG uptake times of 1.5 and 3 h. IPAH lung SUV increased at 3 h compared with SUV at 1.5 h (1.5-h lung SUV, IPAH 0.496 ± 0.030, control 0.397 ± 0.013, P < 0.01; 3-h lung SUV, IPAH 0.511 ± 0.023, control 0.367 ± 0.015, P < 0.01) (Fig. 2C), whereas there was no significant difference in liver SUV among IPAH and controls at either 1.5 or 3 h (1.5-h liver SUV, IPAH 2.023 ± 0.133, control 1.895 ± 0.058; 3-h liver SUV, IPAH 1.853 ± 0.072, control 1.693 ± 0.075; all P > 0.05). To distinguish between increase in glycolysis and increased endothelial cell mass, lung tissue fraction (LTF), which represents lung tissue density, was measured from lungs of IPAH patients and healthy controls. In contrast to SUV, LTF of IPAH patients was similar to healthy controls (1.5-h LTF, IPAH 0.256 ± 0.015, control 0.238 ± 0.008, P = 0.298; 3-h lung LTF, IPAH 0.265 ± 0.012, control 0.244 ± 0.007, P = 0.156). Thus, glucose metabolic activities relative to lung density, i.e., SUV normalized with LTF, were greater in IPAH than in healthy controls (SUV normalized at 3 h, IPAH 0.036 ± 0.001, control 0.025 ± 0.002, P < 0.01) (Fig. 2D).

Given the similar LTF of IPAH and control lungs but significantly higher glucose uptake in IPAH lungs and the increase of SUV only in IPAH lungs over time, the in vivo findings most likely reflect an increased glycolytic rate rather than increased endothelial cell mass, which would be consistent with the in vitro glycolytic rate studies.

Study Population.

IPAH patients were identified by the revised clinical classification of pulmonary hypertension (Venice 2003) (46). The study was approved by the Cleveland Clinic Institutional Review Board, and written informed consent was obtained from individuals.

Cell Culture.

Human PAEC were dissociated and cultured as described in SI Text. A549 cells, an epithelial cell line derived from lung adenocarcinoma, were cultured in MEM (Invitrogen Corporation, Carlsbad, CA) with 10% heat-inactivated FCS. Cells were subjected to hypoxia in a sealed chamber with 2% O₂, 5% CO₂, and 93% N₂ or placed directly in a 5% CO₂ and 95% air incubator (20% O₂) and cultured at 37°C. For NO donor treatment, 125 μM detaNO or 100 μM SNAP was added once a day to cell culture from day 1 to day 6 (15).

Detection of Endothelium-Derived NO.

To evaluate NO synthesis, nitrite (NO₂⁻) concentrations in the culture supernatants were measured in ionomycin-stimulated cell supernatant from subconfluent PAEC from IPAH lung or control lung as described (10). Nitrite concentrations were determined by using the ISO-NOP Nitric Oxide Sensor (World Precision Instruments, Sarasota, FL), an amperometric sensor specific for NO.

Mitochondrial Respiration.

Oxygen consumption of cells vs. healthy controls was measured at 37°C in circulator chambers by using an 5300A biological oxygen monitor (YSI, Yellow Springs, OH) and Clark-type polarographic oxygen electrodes as described with slight modifications (47). Briefly, reactions were conducted using 0.01% digitonin (wt/vol) (Sigma-Aldrich, St. Louis, MO) permeabilized PAEC (7.5 × 10⁶ cells in 3-ml reaction volume) in a reaction medium [250 mM sucrose, 20 mM d-glucose, 40 mM KCl, 5 mM KH₂PO₄, 3 mM MgCl₂, 0.5 mM EDTA, 30 mM Tris·HCl, pH 7.5, and protease inhibitors aprotinin (5 μg/ml), leupeptin (1 μg/ml), pepstatin (1 μg/ml), and Pefabloc SC (24 μg/ml)]. Oxygen consumption measurements were made under state 3 and state 4 respiration using either 5 mM glutamate and 2.5 mM malate or 5 mM succinate as substrate. State 3 respiration was stimulated by the addition of 1 mM ADP. The respiratory control index was calculated by dividing oxygen consumption rate at state 3 by state 4.

Respiratory Chain Complex Assays.

The activities of Complex III and IV of the mitochondrial respiratory chain were measured spectrophotometrically as described (48). Briefly, cells (20 × 10⁶ per ml) were incubated with 0.01% digitonin (Sigma-Aldrich) in PBS containing protease inhibitors [5 μg/ml aprotinin, 1 μg/ml leupeptin, 1 μg/ml pepstatin, and 24 μg/ml Pefabloc SC (Sigma-Aldrich)] (pH 7.4) for 5 min on ice. After centrifugation at 16,000 × g and 4°C for 1 min, the supernatant, containing cytosolic proteins, was removed, and the organelle pellet was resuspended in an equal amount of assay buffer. For Complex III and Complex IV activity, an organelle volume correlating to 40,000 and 10,000 cells was used per reaction, respectively. Complex III activity was measured at 550 nm and 30°C with 15 μM cytochrome c (III), 0.6 mM n-dodecyl-β-d-maltoside (EMD Biosciences, Inc., La Jolla, CA), 2 μg/ml rotenone, and 35 μM ubiquinol in assay medium [25 mM potassium phosphate, 5 mM MgCl₂, 2 mM KCl, 2.5 mg/ml BSA (fraction V), pH 7.2]. Ubiquinol was prepared by dissolving ubiquinone (10 μmol) in 1 ml of ethanol acidified to pH 2 with 6 M HCl. The quinone was reduced with excess solid sodium borohydride. Ubiquinol was extracted into diethylether/cyclohexane (2:1, vol/vol), evaporated to dryness under nitrogen gas, and redissolved in 1 ml of ethanol acidified to pH 2 with 6 M HCl. For Complex IV activity, cytochrome c was reduced with 1 M ascorbate in 0.1 M Mes (pH 7.0) for 20 min and desalted afterward by using a D-Salt Excellulose GF-5 column (Pierce Biotechnology, Inc., Rockford, IL). Complex IV activity was measured at 550 nm and 25°C with 15 μM cytochrome c (II), 0.45 mM n-dodecyl-β-d-maltoside in 20 mM potassium phosphate (pH 7.0).

MTT Assay.

Mitochondrial dehydrogenase activity in cells was determined by the ability of mitochondria to reduce MTT according to the protocol from American Type Culture Collection (Manassas, VA). Absorbance at 570 nm was measured.

ATP Content in IPAH PAEC.

ATP in cells was measured under normoxia (21% O₂) or hypoxia (2% O₂) for 4 h by using a luciferase-based luminescence assay kit (PerkinElmer, Boston, MA).

Glycolytic Rate.

Glycolytic rate of cells was measured by monitoring the conversion of 5-³H-glucose to ³H₂O (49, 50). Briefly, 2 × 10⁵ PAEC from IPAH or healthy controls were suspended in 0.2 ml of endothelial cell growth medium (EGM-2 containing 1,000 mg/L glucose; Cambrex, Walkersville, MD), and 10 μl of 5-³H-glucose (PerkinElmer Life Sciences Inc., Boston, MA) was added to each well. Samples were incubated for 1 h at 37°C in a humidified incubator under 5% CO₂. Reactions were terminated with 0.5 ml of 0.2 N HCl, and 100 μl of the cell/HCl mixture was added to open tubes, which were then placed upright in 4-ml scintillation vials containing 0.5 ml of H₂O. The vials were capped, sealed with parafilm, and incubated for 2 days at room temperature. During the incubation, ³H₂O generated by glycolysis diffused from the tube to the H₂O in the scintillation vial through evaporation and condensation. The contents of the tube were transferred to a new scintillation vial with 0.5 ml of H₂O, the tube was discarded, and scintillation fluid was added to both the original (diffused counts) and second (undiffused counts) vials. The amounts of diffused and undiffused ³H were determined by scintillation counting. Appropriate ³H-glucose-only and ³H₂O-only controls were included, enabling the calculation of ³H₂O in each sample and thus the rate of glycolysis as described (49). Lactate concentration in supernatants overlying culture cells was determined spectrophotometrically (Trinity Biotech, St. Louis, MO). Lactate was converted to pyruvate and hydrogen peroxide (H₂O₂) by lactate oxidase. H₂O₂ in the presence of a peroxidase subsequently catalyzes oxidative condensation of a precursor, producing a chromogen with an absorption maximum at 540 nm.

Southern Blot Analysis.

Total DNA was extracted from cultured cells (Qiagen Inc., Valencia, CA). Primers (5′-TGATCAGAGGATTGAGTAAACGG-3′ and 5′-GGTACCCTAACCGTGCAAAGGTA-3′) were used to amplify a 1,090-bp DNA fragment (2,574–3,663 bp; GenBank accession no. AY339547) from 16,569-bp human mitochondrial genomic DNA (51). The PCR products were purified, sequenced, and used as a probe for Southern blot analysis. For Southern blot analysis, the amount of total DNA from cultured cells was digested with restriction enzyme PvuII, electrophoresed through a 0.8% agarose gel, and transferred to Duralon-UV membranes (Stratagene, Cedar Creek, TX). The filters were hybridized with the ³²P-labeled PCR-generated mitochondrial DNA (mtDNA) probe and visualized with autoradiography.

Ultrastructural Analyses.

Cells were fixed at 4°C for >1 h in 0.1 M sodium cacodylate buffer (pH 7.4) containing 2.5% glutaraldehyde and 4% formaldehyde. After washing three times in the same buffer, cells were postfixed with 1% aqueous osmium tetroxide for 1 h at 4°C, washed in sodium cacodylate buffer, followed by maleate buffer (pH 5.1), and then dehydrated with ascending grades of ethanol and propylene oxide. Samples were embedded in Eponate12 kit, polymerized at 70°C for 48 h, trimmed, sectioned at 70–90 nm, poststained in saturated uranyl acetate and lead citrate, and examined with a transmission electron microscope (Philips CM12; Philips, Eindhoven, The Netherlands).

PET Scan.

The study subjects were imaged sequentially by using x-ray CT and PET on a combined hybrid PET/CT system (Biograph 16 PET/CT; Siemens Medical Solutions, Knoxville, TN) (52). The system is calibrated so that the CT and PET data are coregistered, resulting in fused anatomic and functional images.

Subjects were fasting for at least 6 h before and during the study. For each subject, a 370-MBq (10-mCi) dose of FDG was administered through i.v. injection. With the subject in a resting state, time was allowed for uptake of FDG before commencing imaging. For four of the seven subjects, imaging was performed at 1.5 and 3 h after injection. For the other three of the seven subjects, imaging was performed at 3 h after injection. Differing uptake times were used with the goal of optimizing the lung tissue uptake relative to background activity to enhance the differences between study and control subjects. The underlying hypothesis was that a longer uptake time would allow for more clearance of activity from the blood, whereas phosphorylated FDG would remain trapped in tissue cells.

Image data were analyzed by a nuclear medicine radiologist (D.N.) by using an image fusion workstation (MSViewer; CPS Innovations, Knoxville, TN) with region-of-interest (ROI) measuring tools. Identical ROIs were placed on the several regions on fused CT and PET images, and the mean CT Hounsfield units (HU) and mean PET SUV were recorded. A total of 10 regions were drawn in the lungs, at upper/mid/lower levels and anterior/posterior of each lung. Regions were also drawn in the liver (upper/mid/lower levels) and within the aorta (ascending/arch/descending). LTF, which represents lung tissue density, was calculated by using the following formula: LTF = (1,000 + HU of lung)/1,000. To account for variations in lung tissue density and FDG distribution (e.g., muscular uptake) between patients, SUV of each region of the lung was then normalized for LTF of lung and HU of liver by using the following formula: SUV normalized = SUV of each region of the lung/LTF lung of same region/mean HU of liver.