Adenovirus-Mediated Gene Transfer of Human Platelet-Activating Factor–Acetylhydrolase Prevents Injury-Induced Neointima Formation and Reduces Spontaneous Atherosclerosis in Apolipoprotein E–Deficient Mice
Background—Atherosclerosis is characterized by an early inflammatory response involving proinflammatory mediators such as platelet-activating factor (PAF)-like phospholipids, which are inactivated by PAF-acetylhydrolase (PAF-AH). The effect of adenovirus-mediated expression of PAF-AH on injury-induced neointima formation and spontaneous atherosclerosis was studied in apolipoprotein E–deficient mice.
Methods and Results—Intravenous administration of an adenovirus (5×108 plaque-forming units) directing liver-specific expression of human PAF-AH resulted in a 3.5-fold increase of plasma PAF-AH activity at day 7 (P<0.001); this was associated with a 2.4- and 2.3-fold decrease in malondialdehyde-modified LDL autoantibodies and the lysophosphatidylcholine/phosphatidylcholine ratio, respectively (P<0.001 for both). Non-HDL and HDL cholesterol levels in PAF-AH–treated mice were similar to those of control virus-treated mice. Seven days after virus injection, endothelial denudation of the common left carotid artery was induced with a guidewire. Neointima formation was assessed 18 days later. PAF-AH gene transfer reduced oxidized lipoproteins by 82% (P<0.001), macrophages by 69% (P=0.006), and smooth muscle cells by 84% (P=0.002) in the arterial wall. This resulted in a 77% reduction (P<0.001) of neointimal area. Six weeks after adenovirus-mediated gene transfer, spontaneous atherosclerotic lesions in the aortic root were analyzed. PAF-AH gene transfer reduced atherosclerotic lesions by 42% (P=0.02) in male mice, whereas a nonsignificant 14% reduction was observed in female mice. Basal and PAF-AH activity after gene transfer were higher in male mice than in female mice (P=0.01 and P=0.04, respectively).
Conclusions—Gene transfer of PAF-AH inhibited injury-induced neointima formation and spontaneous atherosclerosis in apolipoprotein E–deficient mice. Our data indicate that PAF-AH, by reducing oxidized lipoprotein accumulation, is a potent protective enzyme against atherosclerosis.
Atherosclerotic cardiovascular diseases, including myocardial infarction and ischemic stroke, are one of the main causes of morbidity and mortality in Western societies.1 Atherosclerosis is characterized by an early inflammatory response involving macrophage activation2 and oxidative modifications leading to the generation of platelet-activating factor (PAF) and oxidized phospholipids with PAF-like bioactivity in LDL.3
PAF, a potent lipid mediator, is generated by endothelial cells in response to oxidant injury. It can induce macrophages to produce superoxide anions and thus contribute to the progression of atherosclerosis.4 5 PAF-like activity is inactivated by PAF-acetylhydrolase (PAF-AH), a Ca2+-independent enzyme that hydrolyzes the sn-2 group of PAF, converting it into lyso-PAF.6 PAF-AH is released by monocytes, macrophages, platelets, erythrocytes, and spleen and liver cells,7 and it has anti-inflammatory properties.8 Human PAF-AH is mainly associated with both LDL and HDL.9 In mice, PAF-AH is predominantly associated with HDL.10 Oxidative modification of LDL, a major feature of the atherogenic process,11 involves generating PAF-like oxidized phospholipids that are inactivated to lyso-PAF-like compounds by the LDL-associated AH, because this enzyme also possesses high phospholipase-A2 activity toward phosphatidylcholines (PC), with either oxidized or short-chain fatty acids in the sn-2 position.12 A decrease in the levels of autoantibodies to malondialdehyde-modified LDL (MDA-LDL) and in the lysophosphatidylcholine (LPC)/PC ratio, may account for a decrease in oxidative stress.13 14
Transgenic apolipoprotein (apoE)–deficient, apoAI-overexpressing mice have increased plasma PAF-AH activity and reduced MDA-LDL autoantibodies.13 Human-like HDL, generated by adenovirus-mediated apoAI gene transfer in apoE-deficient mice, protected against neointima formation.15 Reduced neointima formation was associated with decreased accumulation of oxidized lipoproteins in the injured vessels. Adenovirus-mediated gene transfer of PAF-AH resulted in increased PAF-AH activity, decreased phospholipid oxidation, and reduced recruitment of macrophages to lesion-prone sites in the aortic root of apoE-deficient mice.13 These data suggested that PAF-AH may protect against atherosclerosis.
Therefore, the aim of the present study was to estimate the effect of increasing the plasma levels of PAF-AH on injury-induced neointima formation and spontaneous atherosclerosis in susceptible apoE-deficient mice.
All experimental procedures in animals were performed in accordance with protocols approved by the Institutional Animal Care and Research Advisory Committee. ApoE-deficient mice were backcrossed for 10 generations into the C57BL/6J background and thus had 99.9% C57BL/6J background.16 Mice were fed normal chow ad libitum.
Adenovirus-Mediated Gene Transfer of Human PAF-AH
Human PAF-AH cDNA was a gift from Dr P. Wolkhof (Bayer Pharma, Wuppertal, Germany). It was subcloned in the shuttle plasmid pACpLpA downstream of the Cytomegalovirus promoter/enhancer. PAF-AH recombinant adenovirus (AdPAF-AH) was generated as previously described.13 The control recombinant adenovirus AdRR5 has been described elsewhere.17 A total of 5×108 plaque forming units of AdPAF-AH or AdRR5 was injected into the tail vein of apoE-deficient mice. Citrated blood was collected from the retrobulbar plexus at indicated time points.
Injury-Induced Neointima Formation
Endothelial denudation of the left common carotid artery in female apoE-deficient mice (n=35) was performed 7 days after virus injection using a guidewire, as described elsewhere.18 Eighteen days after injury, mice at 3 months of age were killed, and both the left and right common carotid arteries were dissected and embedded, as previously described15 (Figure 1⇓).
Six weeks after virus injection, mice (n=24; male, n=12; female, n=12) at 6 months of age were killed, and hearts and aortas were fixed and embedded as described previously19 (Figure 1⇑). In both guidewire injury and atherosclerosis models, total blood was collected by puncture of the inferior caval vein. Livers were immediately frozen in liquid nitrogen and stored at −80°C until use.
Morphometric analysis of carotid arteries and hearts was performed in a blinded manner using the Quantimet 600 image analyser (Leica), as described previously.15 19 In AdRR5-treated mice, 58±2 sections per carotid artery and 12±1 sections per heart were analyzed; 60±4 and 12±1 sections, respectively, were analyzed in AdPAF-AH-treated mice.
RNA Extraction and Real-Time Reverse-Transcription Polymerase Chain Reaction
Total RNA was extracted from mouse livers by a single-step method (TRIzol, Gibco Life Technologies) that was based on the guanidinium isothiocyanate acid phenol method. First-strand cDNA was generated from liver total RNA by reverse transcription using random primers from Takara and Superscript RNase H-reverse transcriptase (Gibco Life Technologies). The DNA was then subjected to quantitative real-time polymerase chain reaction (PCR) according to the supplier protocols (Perkin-Elmer). Human PAF-AH was amplified using a forward primer (5′-TCCTGTTGCCCATA- TGAAATCA-3′), a reverse primer (5′-GGCCAAAGCTTGCAG- CA-3′), and a probe that was 3′-labeled with the fluorescent quencher 6-carboxy-tetramethyl rhodamine and 5′-labeled with the indicator dye 6-carboxyfluoresceine (5′-FAM-AGCATGGGT- CAACAAAATACAAGTACTGATGGC-TAMRA-3′). The copy numbers were calculated from plasmid cDNA standards containing the reverse-transcription PCR amplicon. PAF-AH mRNA was expressed as copy number per 1000 copies of hypoxanthine transferase, which was amplified as described elsewhere.13
Plasma PAF-AH activity was measured in plasma as previously described9 and expressed in nmol · mL–1 · min–1.
Determination of MDA-LDL Autoantibodies
Autoantibodies against MDA-LDL in mice were determined as described previously, and their levels were expressed as the MDA-LDL/native LDL ratio to account for unspecific binding to unmodified human LDL.13
Plasma Lipid, Phospholipid, and Lipoprotein Analyses
Lipoprotein fractions were separated by gel filtration. Cholesterol was extracted and quantitated by high-pressure liquid chromatography, as described previously.20 Phospholipids were extracted from the LDL fractions and separated by thin-layer chromatography,21 followed by quantitation of phosphate.22 The LPC/PC ratio was calculated.
Smooth muscle cells were immunostained with a monoclonal antibody against human smooth muscle α-actin (clone 1A4, DAKO; diluted 1:500). Oxidized LDL was detected using monoclonal antibody 4E6 (5 μg/mL) against human oxidized LDL.23 Bound antibodies were revealed using alkaline phosphatase-conjugated goat anti-mouse immunoglobulins (diluted 1:140) and a fuchsin substrate system (DAKO). Macrophages were detected in an indirect staining procedure using a cross-reacting rat biotinylated monoclonal antibody against murine Mac-3 antigen (clone M3/84, Pharmingen; diluted 1:50). Peroxidase reaction was performed using the TSA Biotin System from NEN. Blinded analysis of positive sections immunostained for oxidized LDL, smooth muscle cells, and macrophages was performed with the Quantimet 600 image analyser (Leica). A color threshold mask for immunostaining was defined to detect the red or brown color by sampling, and the same threshold was applied to all samples. The lesion area with positive color was recorded.
Differences between AdPAF-AH– and AdRR5-treated mice were tested by the Mann-Whitney nonparametric test. The time course of PAF-AH activity was tested by ANOVA. P<0.05 was considered statistically significant.
Effect of PAF-AH Gene Transfer on Plasma Lipoprotein-Associated Cholesterol
Baseline levels of β-VLDL cholesterol and HDL cholesterol in control apoE-deficient mice were 580±25 mg/dL and 37±2.3 mg/dL, respectively. PAF-AH gene transfer did not alter β-VLDL cholesterol or HDL cholesterol (Table⇓).
Time-Dependent Expression of Human PAF-AH After Adenovirus-Mediated Gene Transfer
Plasma PAF-AH activity was measured before (day 0) and 7, 14, 21, and 42 days after AdRR5 and AdPAF-AH injection in apoE-deficient mice. At day 0, PAF-AH activity was 36.8±2.5 nmol · mL–1 · min–1. AdRR5 had no significant effect on plasma PAF-AH activity. PAF-AH gene transfer resulted in a 3.5-fold increase in PAF-AH activity at day 7 (P<0.001), a 2.9-fold increase at day 14 (P<0.001), and a 2.4-fold increase at day 21 (P<0.005). Thereafter, PAF-AH activity decreased to baseline by day 42 (39.3±2.8 nmol · mL–1 · min–1; Figure 2⇓).
Real-time PCR analysis of total liver RNA was performed 7 and 25 days after AdPAF-AH injection. The PAF-AH copy number in the livers of mice treated with AdPAF-AH was 186±47 at day 7 (n=5) and 52±10 (n=7) at day 25; it was 4.8±1.1 at day 7 (n=4, P<0.001) and 1.5±0.4 at day 25 (n=7, P<0.001) in AdRR5-treated mice.
Effect of PAF-AH Gene Transfer on Oxidative Stress
The ratio of antibodies against MDA-LDL to antibodies recognizing human native LDL was 10.2±0.6 (n=20) in AdRR5-treated mice and 4.3±0.2 (n=12, P<0.001) in AdPAF-AH–treated mice, indicating a decrease of immunogenic neoepitopes in β-VLDL in the presence of high levels of HDL-associated PAF-AH.
The LPC/PC ratio was 2.3-fold lower in AdPAF-AH–treated mice than in AdRR5-treated mice (0.30±0.02, n=11, versus 0.14±0.01, n=10; P<0.001).
Effect of PAF-AH Gene Transfer on Guidewire Injury–Induced Neointima Formation
Endothelial denudation was performed 7 days after PAF-AH gene transfer, when PAF-AH activity was at its maximum. Before injury, mean intimal areas were very small and were not affected by PAF-AH gene transfer (data not shown).
Mean intimal areas in AdPAF-AH–treated mice were 4.3-fold (P<0.001) smaller than those in AdRR5-treated mice (Figure 3c⇑). The length of the intimal lesion was 1.75-fold smaller in PAF-AH–treated mice than in AdRR5-treated mice (16±3 versus 28±3 positive sections, P=0.0083). This resulted in a 5.2-fold (P=0.001) reduction of the volume of the intimal lesion in AdPAF-AH–treated than in AdRR5-treated mice (Figure 3d⇑). Mean medial areas were similar in AdPAF-AH– and AdRR5-treated mice (Figure 3e⇑). The mean intima/media ratio in injured arteries from AdPAF-AH–treated mice was 5.0-fold lower than in that AdRR5-treated mice (Figure 3f⇑).
Effect of PAF-AH Gene Transfer on Neointima Composition
Figure 4⇓ shows the accumulation of oxidized lipoproteins and the contribution of smooth muscle cells and macrophages to neointima formation in injured arteries of AdRR5- and AdPAF-AH–treated mice.
The neointima area stained by anti-oxidized LDL antibody in AdPAF-AH–treated mice was 5.5-fold smaller than that in AdRR5-treated mice (Figure 4c⇑). The areas occupied by smooth muscle cells and macrophages were 3.5- and 2.6-fold lower in AdPAF-AH– than in AdRR5-treated mice (Figures 4d⇑ and 4e⇑). The relative composition of neointima in oxidized LDL (40±9% versus 32±5%), smooth muscle cells (31±9% versus 21±4%), and macrophages (13±2% versus 18±3%) was similar in AdRR5- and AdPAF-AH–treated mice.
Effect of PAF-AH Gene Transfer on Spontaneous Atherosclerosis
Six weeks after adenovirus injection, when male and female apoE-deficient mice were analyzed together, a 23% and 26% nonsignificant reduction in heart lesion area and volume, respectively, was observed in AdPAF-AH–treated mice (n=10; male, n=5; female, n=5) compared with AdRR5-treated mice (n=14; male, n=7; female, n=7; Figures 5a⇓ and 5b⇓). However, when male apoE-deficient mice were analyzed separately, PAF-AH gene transfer reduced the area and the volume of the lesion by 42% and 44%, respectively (Figures 5a⇓ and 5c⇓). In contrast, PAF-AH gene transfer had no significant effect on spontaneous atherosclerosis in female apoE-deficient mice. Lesion sizes were 1.6- and 2.4-fold larger in female than in male AdRR5- and AdPAF-AH–treated mice, respectively. Figure 5⇓ also shows representative sections of hearts 42 days after AdRR5 and AdPAF-AH injection in male apoE-deficient mice (Figures 5c⇓ and 5d⇓). PAF-AH gene transfer reduced the accumulation of oxidized LDL, smooth muscle cells, and macrophages, without any effect on the relative composition of the atherosclerotic lesions (data not shown).
Basal PAF-AH activity in male mice was 24% higher than that in female mice (P=0.01). Overall PAF-AH activity after PAF-AHgene transfer was 20% higher in male than in female apoE-deficient mice (area under the curve, 3800±400 for male versus 3100±420 for female mice, P=0.04).
The aim of this study was to investigate the effect of PAF-AH gene transfer on neointima formation after endothelial denudation and on spontaneous atherosclerosis in atherosclerosis-susceptible mice. The results reveal that an increase in plasma PAF-AH activity induced a decrease in MDA-LDL autoantibodies and a decrease in the LPC/PC ratio in the β-VLDL fraction, which was associated with decreased deposition of oxidized lipoproteins and decreased accumulation of macrophages and smooth muscle cells in the arterial wall. This finally resulted in the inhibition of injury-induced neointima formation and a reduction of spontaneous atherosclerosis. These observations represent the first evidence of an inhibitory effect of PAF-AH on intimal hyperplasia and on spontaneous atherosclerosis. Decreased levels of plasma MDA-LDL autoantibodies, LPC/PC ratio in the β-VLDL fraction, and accumulation of oxidized lipoproteins in the arterial wall were observed in the absence of changes in plasma cholesterol levels. These data suggest PAF-AH has direct antioxidative activity.
Previously, LDL oxidation was correlated with coronary artery disease in humans.23 24 Oxidized LDL may contribute to the progression of atherosclerotic lesions by inducing endothelial activation, with increased adhesion and leukocyte recruitment to the vessel wall,11 and with smooth muscle cell migration,25 proliferation,26 and apoptosis.27 Oxidized phospholipids may be the bioactive compounds in LDL and VLDL that induce growth factor expression in smooth muscle cells, and they may contribute to smooth muscle cell migration and proliferation.28 Therefore, by inactivating oxidized phospholipids, PAF-AH may prevent the atherogenic effects of oxidized LDL.
We demonstrated that PAF-AH neutralizes oxidized lipoproteins in apoE-deficient mice, resulting in reduced adhesion molecule expression and monocyte adhesion and infiltration in vivo.13 Here, PAF-AH gene transfer decreased the oxidation of lipoproteins in the blood, as evidenced by lower titers of MDA-LDL autoantibodies and lower LPC/PC ratios of β-VLDL fractions, and in the arterial wall, resulting in an inhibition of injury-induced neointima formation and a reduction of atherosclerosis in the absence of mechanical injury.
Relevance of the Study
PAF-AH deficiency, due to a mis-sense mutation near its active site, causes asthma29 30 and is also an independent risk factor for coronary artery disease.31 Furthermore, PAF-AH activity is decreased in subjects at increased risk for atherosclerotic cardiovascular disease, including patients with insulin-dependent diabetes mellitus and hypertension and smokers.5 Decreased PAF-AH activity in these subjects may be due to oxidative inactivation of the enzyme32 33 ; this decreased PAF-AH activity would lead to uncontrolled PAF activity and, hence, to accelerated atherosclerosis.
Previously, increased PAF-AH activity was demonstrated in human and rabbit atherosclerotic lesions.34 However, it was unclear whether PAF-AH contributed to the progression of these lesions or whether the overexpression of PAF-AH suppressed the proinflammatory action of PAF-like oxidized phospholipids, thereby inhibiting lesion formation. The present study reveals that an increase of PAF-AH activity inhibits neointima formation and reduces spontaneous atherosclerosis.
Limitations of the Study
In the present study, a first-generation adenovirus driven by the Cytomegalovirus promoter was used. The major drawback of this type of adenovirus is the presence of promoter shutoff, leading to the transient expression of the transgene.35 Despite short-term expression, we observed an effect of PAF-AH gene transfer on slowly evolving atherosclerosis in the absence of mechanical injury in 6-month-old male apoE-deficient mice, which have higher basal plasma PAF-AH activity and display smaller atherosclerotic lesions than female mice.
The present study demonstrates that adenovirus-mediated gene transfer of human PAF-AH inhibits injury-induced neointima formation in apoE-deficient mice and reduces spontaneous atherosclerosis in the absence of mechanical injury. PAF-AH, by reducing oxidized lipoprotein accumulation, protects against atherosclerosis.
This work was supported by the Interuniversitaire Attractiepolen Program (P4/34) and by the Fonds voor Wetenschappelijk Onderzoek-Vlaanderen (Program G.0110.98). This study was partially supported by INSERM, Groupe Lipids et Nutrition, and by a grant from the Actions Intégrées Franco-Belges “Tournesol.” Rozenn Quarck received a postdoctoral fellowship from the Association SANOFI Thrombose pour la Recherche, Paris, France. She is presently the recipient of a fellowship from the Fonds voor Wetenschappelijk Onderzoek-Vlaanderen. Bart de Geest is a postdoctoral fellow of the Fonds voor Wetenschappelijk Onderzoek-Vlaanderen. Carine Michiels is a research associate of “Fond National de la Recherche Scientifique.” Ann Mertens is a research assistant of the Fonds voor Wetenschappelijk Onderzoek-Vlaanderen. Gregor Theilmeier is a postdoctoral fellow of the Deutsche Forschungsgemeinschaft, Bonn, Germany, and is on temporary leave from the Department of Anesthesiology, Muenster University, Germany. The authors thank Hilde Bernar and Michele Landeloos for excellent technical support.
The first 2 authors contributed equally to this study.
- Received December 18, 2000.
- Revision received January 29, 2001.
- Accepted January 29, 2001.
- Copyright © 2001 by American Heart Association
Subbanagounder G, Leitinger N, Shih PT, et al. Evidence that phospholipid oxidation products and/or platelet- activating factor play an important role in early atherogenesis: in vitro and in vivo inhibition by WEB 2086. Circ Res. 1999;85:311–318.
Stafforini DM, Prescott SM, McIntyre TM. Human plasma platelet-activating factor acetylhydrolase: purification and properties. J Biol Chem. 1987;262:4223–4230.
Tselepis AD, Dentan C, Karabina SA, et al. PAF-degrading acetylhydrolase is preferentially associated with dense LDL and VHDL-1 in human plasma: catalytic characteristics and relation to the monocyte-derived enzyme. Arterioscler Thromb Vasc Biol. 1995;15:1764–1773.
Steinbrecher UP, Pritchard PH. Hydrolysis of phosphatidylcholine during LDL oxidation is mediated by platelet-activating factor acetylhydrolase. J Lipid Res. 1989;30:305–315.
Theilmeier G, De Geest B, Van Veldhoven P, et al. HDL-associated PAF-AH reduces endothelial adhesiveness in apoE-/- mice. FASEB J. 2000;14:2032–2039.
Watson AD, Navab M, Hama SY, et al. Effect of platelet activating factor-acetylhydrolase on the formation and action of minimally oxidized low density lipoprotein. J Clin Invest. 1995;95:774–782.
De Geest B, Zhao Z, Collen D, et al. Effects of adenovirus-mediated human apo A-I gene transfer on neointima formation after endothelial denudation in apoE-deficient mice. Circulation. 1997;96:4349–4356.
Piedrahita JA, Zhang SH, Hagaman JR, et al. Generation of mice carrying a mutant apolipoprotein E gene inactivated by gene targeting in embryonic stem cells. Proc Natl Acad Sci U S A . 1992;89:4471–4475.
Lindner V, Fingerle J, Reidy MA. Mouse model of arterial injury. Circ Res. 1993;73:792–796.
Holvoet P, Vanhaecke J, Janssens S, et al. Oxidized LDL and malondialdehyde-modified LDL in patients with acute coronary syndromes and stable coronary artery disease. Circulation. 1998;98:1487–1494.
Holvoet P, Van Cleemput J, Collen D, et al. Oxidized low density lipoprotein is a prognostic marker of transplant-associated coronary artery disease. Arterioscler Thromb and Vasc Biol. 2000;20:698–702.
Kohno M, Yokokawa K, Yasunari K, et al. Effect of natriuretic peptide family on the oxidized LDL-induced migration of human coronary artery smooth muscle cells. Circ Res. 1997;81:585–590.
Heery JM, Kozak M, Stafforini DM, et al. Oxidatively modified LDL contains phospholipids with platelet- activating factor-like activity and stimulates the growth of smooth muscle cells. J Clin Invest. 1995;96:2322–2330.
Jovinge S, Crisby M, Thyberg J, et al. DNA fragmentation and ultrastructural changes of degenerating cells in atherosclerotic lesions and smooth muscle cells exposed to oxidized LDL in vitro. Arterioscler Thromb Vasc Biol. 1997;17:2225–2231.
Ambrosio G, Oriente A, Napoli C, et al. Oxygen radicals inhibit human plasma acetylhydrolase, the enzyme that catabolizes platelet-activating factor. J Clin Invest. 1994;93:2408–2416.
Dentan C, Lesnik P, Chapman MJ, et al. PAF-acether-degrading acetylhydrolase in plasma LDL is inactivated by copper- and cell-mediated oxidation. Arterioscler Thromb. 1994;14:353–360.
Hakkinen T, Luoma JS, Hiltunen MO, et al. Lipoprotein-associated phospholipase A(2), platelet-activating factor acetylhydrolase, is expressed by macrophages in human and rabbit atherosclerotic lesions. Arterioscler Thromb Vasc Biol. 1999;19:2909–2917.