Altered carnitine homeostasis is associated with decreased mitochondrial function and altered nitric oxide signaling in lambs with pulmonary hypertension

Surgical preparations and care.

Twelve mixed-breed Western pregnant ewes (137–141 days of gestation, term = 145 days) were operated on under sterile conditions with the use of local anesthesia (2% lidocaine hydrochloride) and inhalational anesthesia (1–3% isoflorane). A midline incision was made in the ventral abdomen, and the pregnant horn of the uterus was exposed. Through a small uterine incision, the left fetal forelimb and chest were exposed, and a left lateral thoracotomy was performed in the third intercostal space. Additional fetal anesthesia consisted of local anesthesia with 1% lidocaine hydrochloride and ketamine hydrochloride (5 mg im). With the use of side-biting vascular clamps, an 8.0-mm Gore-Tex vascular graft (∼2 mm long; W.L. Gore and Associates, Milpitas, CA) was sutured between the ascending aorta and main pulmonary artery with 7.0 proline (Ethicon, Somerville, NJ), using a continuous suture technique. The thoracotomy incision was then closed in layers. This procedure was previously described in detail (53).

Two or four weeks after spontaneous delivery, these lambs and 12 age-matched control lambs were fasted for 24 h with free access to water. The lambs were then anesthetized with ketamine hydrochloride (15 mg/kg im). Under additional local anesthesia with 1% lidocaine hydrochloride, polyurethane catheters were placed in an artery and vein of a hind leg. These catheters were advanced to the descending aorta and the inferior vena cava, respectively. The lambs were then anesthetized with ketamine hydrochloride (∼0.3 mg·kg⁻¹·min⁻¹), diazepam (0.002 mg·kg⁻¹·min⁻¹), and fentanyl citrate (1.0 μg·kg⁻¹·h⁻¹), intubated with a 7.0-mm outer diameter (OD) cuffed endotracheal tube, and mechanically ventilated with 21% oxygen using a Healthdyne pediatric time-cycled, pressure-limited ventilator. With the use of strict aseptic technique, a midsternotomy incision was then performed. Patency of the vascular graft was confirmed by inspection and changes in oxygen saturation. A side-biting vascular clamp was used to isolate peripheral lung tissue from randomly selected lobes, and the incisions were cauterized. Approximately 300 mg of peripheral lung were obtained for each biopsy; four biopsies were obtained. Blood was obtained from the femoral artery.

At the end of the protocol, all lambs were killed with a lethal injection of pentobarbital sodium in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals. All protocols and procedures were approved by the Committees on Animal Research at the University of California, San Francisco, and the Medical College of Georgia.

Pulmonary arterial endothelial cell cultures.

Primary cultures of juvenile ovine pulmonary arterial endothelial cells (PAEC) were isolated as detailed previously (68). All cultures were maintained in DMEM supplemented with 10% fetal calf serum (Hyclone, Logan, UT) and antibiotics/antimycotic solution (500 IU penicillin, 500 μg/ml streptomycin, and 1.25 μg/ml amphotericin B; MediaTech, Herndon, VA) at 37°C in a humidified atmosphere with 5% CO₂ and 95% air. Cells were used at passages 9 and 10, seeded at ∼50% confluence, and utilized when fully confluent.

Western blot analysis.

Lung protein extracts were prepared by homogenizing sheep lung tissues in lysis buffer (50 mM Tris·HCl, pH 7.6, 0.5% Triton X-100, and 20% glycerol) containing Halt protease inhibitor cocktail (Pierce, Rockford, IL). Extracts were then clarified by centrifugation (15,000 gfor 15 min at 4°C). Supernatant fractions were assayed for protein concentration using the Bradford reagent (Bio-Rad, Richmond, CA) and used for Western blot analysis. Protein extracts (25–50 μg) were separated on LongLife 4–20% Tris-SDS-HEPES gels (Frenchs Forest, Australia) and electrophoretically transferred to Immuno-Blot polyvinylidene difluoride membrane (Bio-Rad Laboratories, Hercules, CA). The membranes were blocked with 5% nonfat dry milk in Tris-buffered saline containing 0.1% Tween 20. After blocking, the membranes were probed with antibodies to CrAT (Santa Cruz Biotechnology), CPT1-B (Affinity Bioreagents), CPT2 (Affinity Bioreagents), HSP90 (BD Transduction), MnSOD (Upstate, Lake Placid, NY), and uncoupling protein-2 (UCP-2; Calbiochem, La Jolla, CA). Reactive bands were visualized using chemiluminescence (SuperSignal West Femto substrate kit; Pierce) on a Kodak 440CF image station (New Haven, CT). Band intensity was quantified using Kodak 1D image processing software. Expression of each protein was normalized by stripping the blots with Restore Western blot stripping buffer (Pierce) and then reprobing them with β-actin as a loading control (Sigma, St. Louis, MO).

Measurement of carnitine homeostasis.

Detection of carnitines was performed using an Amersham Biosciences AKTA purifier system (GE Healthcare, Piscataway, NJ) with a 5-μm OmniSpher C18 column (250 × 4.6-mm OD) equipped with a Jasco FP-2020 fluorescence detector (Tokyo, Japan). Total and free carnitines levels were quantified in lung tissue homogenates by fluorescence detection at 248 (excitation) and 418 nm (emission) as described previously (39, 43).

Sample purification and derivatization before HPLC detection.

For free carnitine [l-carnitine and acetyl l-carnitine (ALCAR)] determination, 100-μl samples, 300 μl of water, and 100 μl of internal standard (Sigma ST 1093) were mixed. For total carnitine determination, 100-μl samples were hydrolyzed with 0.3 M KOH, heated at 45°C, and pH neutralized using perchloric acid; the volume was made to 400 μl, and 100 μl of internal standard was added. All samples were purified using solid-phase extraction columns (SAX, 100 mg/ml; Varian, Harbor City, CA), derivatized using aminoanthracene in the presence of EDCI (catalyst), and kept at 30°C for 1 h to complete reaction of carnitines. Separation was carried out with an isocratic elution in 0.1 M Tris-acetate buffer (pH 3.5)-acetonitrile (68:32, vol/vol) at a flow rate of 0.9 ml/min as described previously (39, 43).

Measurement of CrAT activity.

Peripheral lung tissue was homogenized in 50 mM Tris·HCl, pH 7.5, 2 mM EDTA, 5 mM MgCl₂, 0.8 mM DTT, and 0.25 mM PMSF with protease inhibitor cocktail. The homogenates were then centrifuged at 3,500 gfor 5 min, and Coomassie protein estimation was carried out. CrAT activity was then determined using a modification of the method described by Liu et al. (38). Briefly, the assay mixture consisted of 50 mM Tris·HCl, pH 7.5, 2 mM EDTA, 25 mM malate, 0.25 mM NAD, 12.5 g/ml rotenone, 0.04% Triton X-100, 12.5 g/ml malic dehydrogenase, 50 g/ml citrate synthase, 6.25 μM CoA, 200 μM CoA, 400 μM CoA, and 2 mM ALCAR. Sample (10 μl) with a protein concentration of 0.5 mg/ml was added to the assay mixture along with CoA (400 μM) with a constant concentration of 2 mM ALCAR (the K_m value for CoA). Reactions were monitored at room temperature for 5 min in a Beckman DU series 600 spectrophotometer. The absorbance was measured at time intervals of 30 s, giving 10 readings per sample. The rate was calculated using linear regression to determine the best-fit line.

Immunoprecipitation analysis for nitrated CrAT.

To determine nitrated CrAT levels, we homogenized lung tissues in immunoprecipitation buffer [25 mM HEPES, pH 7.5, 150 mM NaCl, 1% Nonidet P-40, 10 mM MgCl₂, 1 mM EDTA, and 2% glycerol supplemented with protease inhibitor cocktail (Pierce)]. Tissue homogenates (1,000 μg of protein) were precipitated with a rabbit antibody against 3-nitrotyrosine (5 μg; Upstate Biotechnology) in 0.5-ml final volume at 4°C overnight. Protein G plus/protein A agarose (40 μl; Calbiochem) was added and rotated at 4°C for an additional 2 h. The precipitated protein was washed three times in 2× volume of immunoprecipitation buffer; the pellet was resuspended in Laemmli buffer (20 μl), boiled, and separated on a 4–20% SDS-PAGE gel (LongLife). CrAT protein levels were then detected using Western blot analysis.

Determination of lactate and pyruvate levels.

Lung tissues were homogenized in ice-cold 0.5 M perchloric acid and centrifuged at 14,000 rpm for 20 min. The supernatant were then neutralized with 3 M KHCO₃ and used for lactate and pyruvate assays. The relative changes in lactate levels were measured using a lactate assay kit (Biovision). Pyruvate levels were determined using the spectrophotometric enzymatic measurement assay at 340 nm. NADH was used as a cofactor and lactate dehydrogenase as the coenzyme. Experimental conditions were as previously described (40).

Immunoprecipitation analysis for HSP90 interaction with eNOS.

For the experiments in PAEC, cells were treated with 50 μM 2,4-dinitrophenol (2,4-DNP; Sigma) for up to 8 h. The cells were then lysed in ice-cold lysis buffer. For immunoprecipitation, cell lysates were incubated with anti-eNOS antibody (BD Transduction Laboratories) for 2 h at 4°C and then with protein G plus/protein A agarose suspension (Calbiochem) for 1 h at 4°C. The immune complexes were washed three times with the lysis buffer and boiled in SDS-PAGE sample buffer for 5 min. Agarose beads were pelleted by centrifugation, and protein supernatants were loaded and run on (4–20%) polyacrylamide gels, followed by transfer of the proteins to nitrocellulose membranes. Membrane was blocked with 2% BSA in Tris-buffered saline containing 0.05% Tween 20 (TBST) for 2 h at room temperature, incubated with anti-HSP90 (BD Transduction Laboratories) for 2 h at room temperature, washed three times with TBST (room temperature, 10 min), and then incubated with a horseradish peroxidase-conjugated secondary antibody (Pierce). The reactive bands were visualized with the SuperSignal West Femto maximum sensitivity substrate kit (Pierce) and Kodak 440CF image station. The same blot was reprobed with anti-eNOS antibody to normalize for the levels of eNOS immunoprecipitated in each sample.

To determine eNOS-HSP90 interactions in the peripheral lung, tissues were homogenized in immunoprecipitation buffer as described above for nitrated CrAT. Tissue homogenates (1,000 μg of protein) were then analyzed as described for PAEC.

Shear stress.

Laminar shear stress was applied using a cone-plate viscometer that accepts six-well tissue culture plates, as described previously (15, 67, 69). This method achieves laminar flow rates that represent physiological levels of laminar shear stress in the major human arteries, which is in the range of 5–20 dyn/cm² with localized increases to 30–100 dyn/cm².

Determination of mitochondrial dysfunction in PAEC.

We utilized two measures to identify mitochondrial dysfunction in PAEC exposed to 2,4-DNP. First, we determined the effect of 2,4-DNP (50 μM, 0–8 h) on PAEC ATP generation. ATP levels were estimated using the firefly luciferin-luciferase method with a commercially available kit (Invitrogen). Luminescence was determined using a Fluoroscan Ascent FL plate luminometer (Thermo Electron). Values for untreated cells at time 0 were set to 100%, and time-dependent changes in ATP levels were reported as a percentage of the value at time 0.

Second, we determined the effect on mitochondrial superoxide production using the MitoSOX red mitochondrial superoxide indicator (Molecular Probes), a fluorogenic dye for selective detection of superoxide in the mitochondria of live cells. MitoSOX red reagent is live-cell permeant and is rapidly and selectively targeted to the mitochondria. Once in the mitochondria, MitoSOX red reagent is oxidized by superoxide and exhibits bright red fluorescence upon binding to nucleic acids. Briefly, after treatment with 2,4-DNP, cells were washed with fresh medium and then incubated in medium containing MitoSOX red (2 μM) for ∼10 min at 37°C in dark conditions. Cells were washed with fresh serum-free medium and imaged using fluorescence microscopy at an excitation of 510 nm and an emission of 580 nm. A personal computer-based imaging system consisting of the following components was used for the fluorescent analyses: an Olympus IX51 microscope equipped with a charge-coupled device camera (Hamamatsu Photonics) was used for acquisition of fluorescent images. The average fluorescence intensities (to correct for differences in cell number) were quantified using ImagePro Plus version 5.0 imaging software (Media Cybernetics).

Detection of NO_x levels in PAEC.

NO generated by PAEC in response to shear was measured using an NO-sensitive electrode with a 2-mm-diameter tip (ISO-NOP sensor; WPI) connected to an NO meter (ISO-NO Mark II; WPI) as described previously (70).

Measurement of tissue NO_x levels.

To quantify bioavailable NO, we determined levels of NO and its metabolites in the peripheral lung tissue from shunt and age-matched control lambs at 4 wk of age. In solution, NO reacts with molecular oxygen to form nitrite and with oxyhemoglobin and superoxide anion to form nitrate. Nitrite and nitrate are reduced using vanadium(III) and hydrochloric acid at 90°C. NO is purged from solution, resulting in a peak of NO for subsequent detection by chemiluminescence (NOA 280; Sievers Instruments, Boulder, CO). The detection limit is 1 nM per milliliter of nitrate.

Electron paramagnetic resonance spectroscopy and spin trapping.

To detect superoxide generation in intact cells, we performed electron paramagnetic resonance (EPR) measurements as described previously (70). After overnight serum starvation of the cells, 20 μl of spin-trap stock solution consisting of 1-hydroxy-3-methoxycarbonyl-2,2,5,5-tetramethylpyrrolidine·HCl (CMH; Alexis Biochemicals, San Diego, CA) at 20 μM in Dulbecco's PBS (DPBS) plus 25 μM desferrioxamine (Calbiochem) and 5 μM diethyldithiocarbamate (Alexis Biochemicals, Lausen, Switzerland) plus 2 μl of DMSO were added to each well before shear stimulation. Adherent cells were trypsinized and pelleted at 500 g after 45 min of incubation at 37°C following shear to allow entrapment of superoxide by the spin trap. Cell pellet was washed and suspended in a final volume of 35 μl of DPBS (plus desferrioxamine and diethyldithiocarbamate), loaded into a 50-μl capillary tube, and analyzed with a MiniScope MS200 EPR (Magnettech, Berlin, Germany) at a microwave power of 40 mW, modulation amplitude of 3,000 mG, and modulation frequency of 100 kHz. EPR spectra were analyzed measured for amplitude using ANALYSIS software (version 2.02; Magnettech), and experimental groups were compared using the statistical analysis described below.

Measurement of superoxide levels in peripheral lung tissue.

Approximately 0.2 g of peripheral lung tissue was sectioned from fresh frozen tissue and immediately immersed in either normal EPR buffer [PBS supplemented with 5 μM diethydithiocarbamate (Sigma-Aldrich) and 25 μM desferrioxamine (Sigma-Aldrich)] or EPR buffer supplemented with 100 U/ml polyethylene glycol-superoxide dismutase (PEG-SOD; Sigma) or 100 μM 3-ethylisothiourea (ETU; Sigma), an inhibitor of human NO synthases (21). Samples were incubated for 30 min on ice. During incubation, samples were analyzed for protein content using Bradford analysis (Bio-Rad). Sample volumes were then adjusted with EPR buffer plus 25 mg/ml CMH hydrochloride (Axxora) to achieve equal protein content and a final CMH concentration of 5 mg/ml. Samples were homogenized for 30 s with a VWR PowerMAX AHS 200 tissue homogenizer, incubated for 60 min on ice, and then centrifuged at 14,000 g for 15 min at room temperature. Supernatant (35 μl) was loaded into a 50-μl capillary tube and analyzed for superoxide generation as described above.

To demonstrate that ETU quenched superoxide in a nonspecific manner within our experimental system, we utilized an in vitro superoxide-creating reaction of xanthine oxidase (1 U/ml; Sigma) with xanthine (1 mM; Sigma), as previously described (29) in the presence of CMH hydrochloride. Reactions were run in a total volume of 1 ml (PBS, pH 7.5) for 20 min at 25°C. Two control reactions, one lacking xanthine and the other lacking xanthine oxidase, were included to ensure the absence of nonspecific reactions of either reagent with the EPR spin trap. After incubation, 35 μl of supernatant were loaded into a 50-μl capillary tube and analyzed for superoxide generation as described above.

Statistical analyses.

Statistical analysis was performed using GraphPad Prism version 4.01 for Windows (GraphPad Software, San Diego, CA). Means ± SE were calculated for all samples, and significance was determined using the unpaired t-test (for 2 groups) or ANOVA (for ≥3 groups). A value of P < 0.05 was considered significant.

Increased pulmonary blood flow in shunt lambs.

All shunt lambs had a palpable thrill and an increase in the oxygen saturation between the right ventricle and pulmonary artery, demonstrating patency of the graft. In fact, the pulmonary-to-systemic blood flow ratio of shunt lambs was 2.9 ± 0.9. Compared with 2-wk-old control lambs, 2-wk-old shunt lambs had increased mean pulmonary arterial pressure (21.7 ± 3.2 vs. 15.3 ± 2.3 mmHg), left pulmonary blood flow (147.3 ± 23.4 vs. 55.1 ± 13.2), and left atrial pressure (6.8 ± 3.7 vs. 2.7 ± 1.6 mmHg) (P < 0.05). Right atrial pressure, heart rate, and mean systemic arterial pressure values were similar between the groups.

Reduced expression of the enzymes involved in carnitine homeostasis in shunt lambs.

Initially, we measured the expression of the three important mitochondrial enzymes involved in carnitine metabolism: CPT1-B, CPT2 (involved in the transport of long-chain fatty acids from cytosol to mitochondrial matrix), and CrAT (transesterfies short-chain acyl chains) in 2-wk old shunt and age-matched control lambs. There was a significant decrease in the expression of both CPT1-B (Fig. 1, A and B) and CPT2 enzymes (Fig. 1, C and D) in shunts compared with age-matched control lambs. Similarly, both CrAT expression (Fig. 2, A and B) and activity (control: 152.9 ± 46.1 units/mg vs. shunt: 7.43 ± 3.33 units/mg, P < 0.05 vs. control; Fig. 2C) were significantly decreased in shunt compared with age-matched control lambs.

Increased nitration decreases CrAT activity in shunt lambs.

Although our data indicated that CrAT expression was decreased ∼2-fold, the CrAT activity was decreased ∼20-fold. This suggested that a posttranslational modification was altering CrAT activity. Since we had previously found that there is an increase in oxidative stress in the shunt lambs, we focused on the potential role of enzyme nitration. The levels of nitrated CrAT were determined using immunoprecipitation with a specific antiserum raised against 3-nitrotyrosine residues and were found to be significantly increased in shunt compared with age-matched control lambs (Fig. 3, Aand B). Furthermore, we found that exposing purified CrAT to peroxynitrite (10 μM, 5 min) was sufficient to significantly reduce CrAT activity (vehicle: 99.9 ± 17.74 units/μg vs. peroxynitrite: 0.6481 ± 0.24 units/μg, P < 0.05 vs. vehicle; Fig. 4).

Decreased CrAT activity leads to alterations in carnitine metabolism.

CrAT plays an important role in maintaining carnitine metabolism (10). Therefore, we next determined total carnitine and free carnitine levels in the peripheral lung of shunt and control lambs. Using HPLC analysis, we found that total carnitine levels were unchanged in shunt lambs (60.54 ± 8.45 vs. 72.37 ± 10.14 nmol/g wet wt; Table 1). However, free carnitine levels were decreased (34.86 ± 3.35 vs. 67.57 ± 13.43 nmol/g wet wt, P < 0.05 vs. control; Table 1). Furthermore, l-carnitine levels were significantly lower in shunt compared with age-matched control lambs (20.21 ± 4.29 vs. 56.41 ± 12.27 nmol/g wet wt, P < 0.05 vs. control; Table 1). Since free carnitines and other acylcarnitines contribute to total carnitine levels, these data indicate that the percentage of carnitine present as acylcarnitine (calculated as total carnitine minus free carnitine) is significantly higher in shunt lambs (40.71 ± 9 vs. 6.2 ± 4.7%, P < 0.05 vs. control; Table 1).

Altered carnitine metabolism is associated with mitochondrial dysfunction in shunt lambs.

Since the carnitine acyltransferase pathway has been shown to be of critical importance for maintaining normal mitochondrial function, we next determined whether shunt lambs present with markers of mitochondrial dysfunction. Because loss of SOD2 (8, 13, 18, 26, 35, 44, 48) and increases in UCP-2 (2, 27, 42, 46, 62) have been shown to be markers of mitochondrial dysfunction in other systems, we examined these markers. Our data indicate that SOD2 expression was significantly decreased (Fig. 5, A and B) in shunt lambs, whereas UCP-2 expression was significantly increased (Fig. 5, C and D). In addition, we determined the lung levels of lactate and pyruvate in shunt and control lambs to quantify mitochondrial activity. Under normal mitochondrial function, pyruvate levels are higher than lactate, and thus any increases in lactate levels decreases the pyruvate-to-lactate ratio and can be extrapolated to suggest a reduction in ATP generation by the mitochondria (57). As shown in Table 2, shunt lambs had significantly decreased pyruvate levels (0.059 ± 0.007 vs. 0.132 ± 0.028 μmol/g wet wt, P < 0.05) and increased lactate levels (0.568 ± 0.838 vs. 1.348 ± 0.214 μmol/g wet wt, P < 0.05) as well as a significant reduction in the pyruvate-to-lactate ratio.

Mitochondrial dysfunction attenuates HSP90-eNOS interaction and shear-induced NO production in PAEC.

To further investigate the effect of mitochondrial dysfunction on NO signaling, we utilized cultured PAEC isolated from juvenile lambs. Mitochondrial dysfunction was induced in PAEC by using 2,4-DNP (50 μM, 0–6 h). Our data indicate that 2,4-DNP significantly decreased ATP levels in PAEC (Fig. 6A) associated with an increase in oxidative stress within the mitochondria (Fig. 6B). Furthermore, we found that this mitochondrial dysfunction was associated with a decrease in the association of HSP90 with eNOS (Fig. 7, Aand B) and shear-induced NO production (Fig. 7C) and an increase in eNOS-derived superoxide (Fig. 7D), indicating that impaired mitochondrial function reduces NO production.

Decreased HSP90-eNOS interaction and altered NO signaling in shunt lambs.

To determine the effect of decreased mitochondrial function in shunt lambs on NO signaling, we carried out immunoprecipitation studies to determine eNOS-HSP90 interactions in both 2- and 4-wk-old shunt and control lambs. Our data indicate that there was a progressive decrease in eNOS-HSP90 interactions between 2 and 4 wk of age in shunt compared with control lambs (Fig. 8, A and B) and that this was associated with a progressive decrease in relative eNOS activity (as determined by tissue NO_x levels as a fraction of calcium-dependent [³H]arginine metabolism, Fig. 8C) and increased NOS-dependent superoxide levels indicative of eNOS uncoupling (Fig. 8D). We also confirmed the specificity of ETU for NOS-derived superoxide by demonstrating that ETU did not quench the superoxide generated by a xanthine/xanthine oxidase superoxide-generating system (Fig. 8E).