|Home | About | Journals | Submit | Contact Us | Français|
Rationale: Impaired endothelial cell–dependent vasodilation, inflammation, apoptosis, and proliferation are manifestations of endothelial dysfunction in chronic obstructive pulmonary disease (COPD). Prostacyclin (PGI2) is a major product of the cyclooxygenase pathway with potent vasodilatory and antimitogenic properties and may be relevant to endothelial dysfunction in COPD.
Objectives: To determine if PGI2 expression is altered in smoking-related lung disease and if it may be protective in COPD-associated endothelial dysfunction.
Methods: We evaluated, by immunohistochemistry, Western blotting, and polymerase chain reaction, human emphysema tissue compared with normal tissue for expression of prostacyclin synthase (PGI2S). We examined the effects of cigarette smoke extract (CSE) and aldehyde components on eicosanoid expression in primary human pulmonary microvascular endothelial cells. Finally, we used a murine model of lung-specific PGI2S overexpression and in vitro studies to determine if PGI2 expression has protective effects on cigarette smoke–induced endothelial apoptosis.
Measurements and Main Results: Human emphysema lung tissue exhibited lower PGI2S expression within the pulmonary endothelium than in normal lung. In vitro studies demonstrated that CSE, and in particular the α,β unsaturated aldehyde acrolein, suppressed PGI2S gene expression, whereas CSE significantly induced the upstream mediators COX-2 and cytosolic phospholipase A2 in human pulmonary microvascular endothelial cells. Mice with lung-specific PGI2S overexpression exhibited less endothelial apoptosis after chronic smoke exposure. In vitro, iloprost exhibited protective effects on CSE-induced apoptosis.
Conclusions: PGI2 has protective effects in the pulmonary vasculature after acute and chronic cigarette smoke exposure. An imbalance in eicosanoid expression may be important to COPD-associated endothelial dysfunction.
Prostacyclin seems to be biologically relevant in endothelial dysfunction; however, the role of eicosanoids in smoking-related lung disease remains unclear.
Prostacyclin expression is decreased in chronic obstructive pulmonary disease. Prostacyclin may have protective effects in the pulmonary endothelium after cigarette smoke exposure.
Chronic obstructive pulmonary disease (COPD) can be characterized as a destruction of small airways and parenchyma resulting in a progressive impairment in pulmonary function (1). Cigarette smoke is the major pathogenic factor implicated in COPD, and pulmonary hypertension develops in approximately 6% of smokers with COPD (1). The interaction between parenchymal disease and the vasculature is often clinically evident by the observation that patients with severe COPD often have mild or moderate pulmonary hypertension at rest (2). Histopathologically and microscopically, the pulmonary vasculature in COPD is characterized by intimal thickening, with smooth muscle deposition (2) and loss of alveolar septal structures and microvasculature (3). In COPD, alveolar septal and endothelial cells undergo apoptosis, which is likely important to the pathogenesis of disease (4). The etiology of pulmonary hypertension in COPD has been considered to be chronic hypoxemia, but pulmonary hypertension is evident in milder forms of COPD where hypoxemia is absent. This finding suggests that other mechanisms are responsible for the vascular changes in COPD (2).
Cigarette smoke is considered to be a major risk factor in the development of COPD, and its effects on the lung epithelium have been well characterized (5–7). The alveolar septae and microvasculature are affected in COPD. Because endothelial dysfunction has been implicated in the pathogenesis of several chronic diseases, including coronary artery disease, peripheral vascular diseases, diabetes, and renal failure, impaired endothelial function of pulmonary arteries has been described in small pulmonary arteries from patients with COPD (8–11). Endothelial cell apoptosis has been implicated in smoking-related endothelial dysfunction (12). Researchers have also demonstrated that cigarette smoke induces necrosis of epithelial and endothelial cells (13, 14).
The role of eicosanoid expression in COPD and other smoking-related lung disease remains unclear. Previous studies have identified elevated levels of markers of oxidative stress, such as the prostaglandin (PG)F2α analog 8-isoprostane in the exhaled breath condensate of patients with COPD (15). Proinflammatory molecules, such as leukotriene B4 and prostaglandin E2, are elevated in the exhaled breath condensate of individuals with COPD (16).
Prostacyclin (PGI2) has potent vasodilatory and antimitogenic properties and is one of the main therapies proven to improve survival in patients with severe idiopathic pulmonary arterial hypertension (17). PGI2 production occurs primarily in pulmonary vascular smooth muscle and endothelium via the cyclooxygenase prostaglandin H synthase pathway. Subsequent conversion to PGI2 is mediated by prostacyclin synthase (PGI2S), a member of the group of cytochrome P450 enzymes. Our laboratory previously identified decreased expression of PGI2S in the lungs from patients with severe idiopathic pulmonary arterial hypertension (18). Furthermore, transgenic mice with lung specific PGI2S overexpression were protected from hypoxia-induced pulmonary hypertension (19).
Cigarette smoke may contribute to endothelial dysfunction and the development of cardiovascular disease through the inhibition of PGI2 release from the endothelium (20, 21). The mechanism of inhibition of PGI2 release by cigarette smoke is unclear. Researchers have implicated arachidonic acid mobilization and downstream enzymes as potential targets of cigarette smoke (20, 21).
We hypothesized that PGI2 expression may be reduced in other smoking-related diseases, such as COPD. In addition, we hypothesized that reduced PGI2 expression may be important to the pathogenesis of the disease, and, given the decreased survival among patients with COPD who have concomitant pulmonary hypertension, PGI2 expression may be protective to the pulmonary vasculature. We examined the pulmonary endothelium of patients with emphysema for PGI2S expression and the direct effects of cigarette smoke and acrolein, a potent αβ unsaturated aldehyde found in cigarette smoke, on PGI2S expression in the pulmonary microvasculature. Finally, we determined whether PGI2 has protective effects on the pulmonary endothelium (in vivo and in vitro) after cigarette smoke exposure. Some of the results of these studies have been previously reported in the form of an abstract (22).
Human pulmonary microvascular endothelial cells (HPMVECs, passage 4–6) and endothelial growth medium-2-microvascular (EGM-2-MV) medium (containing 5% fetal bovine serum, hydrocortisone, human recombinant vascular endothelial growth factor, recombinant human fibroblast growth factor–B, recombinant insulinlike growth factor–1, human recombinant epidermal growth factor, ascorbic acid, gentamycin, and amphotericin-B) were purchased from Clonetics Corporation (Baltimore, MD). Cells were grown to confluence at 37°C in a humidified atmosphere of 21% O2, 5% CO2. N-acetylcysteine (NAC) and diphenyleneiodonium chloride (DPI) were obtained from Sigma-Aldrich (St. Louis, MO). Nω-nitro-l-arginine methyl ester (l-NAME) was obtained from Cayman Chemical (Ann Arbor, MI). Z-Asp2,6-dichlorobenzoylmethylketone (Caspase Family Inhibitor IV) was obtained from Alexis Biochemicals (San Diego, CA). A superoxide dismutase (SOD) mimetic and catalase were received as a generous gift from Russell Bowler, M.D. (National Jewish Medical Center, Denver CO). Acrolein and crotonaldehyde (αβ unsaturated aldehydes) and saturated aldehydes (acetaldehyde, propionaldehyde) were purchased from Sigma-Aldrich.
Cigarette smoke extract (CSE) was prepared by a modification of a previously published method (23). Briefly, one nonfiltered Camel cigarette (R.J. Reynolds, Winston-Salem, NC) was passed through 10 ml of phosphate-buffered saline (PBS) using a vacuum pump. This 100% CSE was adjusted to a pH of 7.4 and filtered through a 0.22-μM pore filter (Fisher, Hampton, NH) to remove particles and bacteria, and the CSE was diluted to the appropriate concentration and added to endothelial cells within 10 minutes of preparation.
Paraffin-embedded sections of human lung tissue (obtained from the University of Colorado Tissue Bank), four emphysema and three nondiseased, were deparaffinized and rehydrated with xylene and ethanol. Antigen retrieval was performed using the microwave method with citrate buffer for 20 minutes. Avidine and biotin block was performed, and endogenous peroxidase was quenched by 3% hydrogen peroxide. After blocking with 5% normal goat serum, rabbit anti-human PGI2S antibody (1:25) (a generous gift from D. Dewitt, Ph.D., Michigan State University) was applied overnight at 4°C. The sections were washed with PBS with 0.05% Tween and incubated with biotinylated goat anti-rabbit IgG. After washing, sections were stained with ABC Vectastatin reagents (PK-6101; Vector Labs, Burlingame, CA) and DAB (SK4100; Vector Labs). Sections were then scored by a pathologist in blinded fashion (number of PGI2S-positive endothelial cells/100 endothelial cells per case) for PGI2S staining in capillaries, small/medium arteries, and arterioles.
In vitro, the rate of apoptosis in primary human microvascular endothelial cells was analyzed by the Vybrant Apoptosis Assay Kit (Molecular Probes, Inc., Eugene, OR) according to the manufacturer's protocol. After treatment with CSE, the cells were washed and trypsinized. Cells (1 × 106) were suspended in 100 μl of annexin-binding buffer and fluorescein isothiocyanate annexin-V/propidium iodide. After 15 minutes of incubation, samples were analyzed by flow cytometry measuring the fluorescence emission at 530 nm and greater than 575 nm. In addition, apoptosis of HPMVECs was assessed by a second method (caspase 3/7 assay; Promega, Madison, WI). Immunohistochemical staining of PGI2S transgenic mice and wild-type littermates was conducted using the manufacturer's protocol with cleaved caspase 3 antibody (Cell Signaling Technology, Danvers, MA) to assess apoptosis. In a blinded fashion, small, medium, and large pulmonary arteries were scored for endothelial caspase staining (total of six vessels per mouse).
A total of 20 μg of protein was separated on NuPAGE Novex 4–12% Bis-Tris Gel (Invitrogen, Carlsbad, CA) and transferred to a polyvinylidene fluoride membrane. The membrane was blocked with 5% nonfat dry milk in PBS containing 0.05% Tween (PBST) for 1 hour. After washing with PBST, the membrane was incubated with rabbit polyclonal PGI2S antibody (1:1,000) overnight at 4°C. The membrane was washed four times and incubated with horseradish peroxidase–conjugated secondary antibody (anti-rabbit, 1:3,000; Santa Cruz, Santa Cruz, CA) for 1 hour. Film was developed by chemiluminescence (Perkin Elmer, Boston, MA).
Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Valencia, CA), following the manufacturer's instructions. RNA (1 μg) was reverse-transcribed using random primer and MultiScribe RT (High-Capacity cDNA Archive Kit; Applied Biosystems, Foster City, CA). Assay-on-demand gene expression probes for PGI2S:Hs00168766, cyclooxygenase (COX)-1:Hs00377721, COX-2:Hs00153133, cytosolic phospholipase A2 (cPLA2):Hs00233352, and B-actin:Hs999999903 were obtained from Applied Biosystems. Polymerase chain reactions were performed in 20-μl volumes containing 9 μl of cDNA, 10 μl of TaqMan Master Mix (Applied Biosystems), and 1 μl of assay-on-demand primer and probe. Real-time polymerase chain reaction (PCR) was conducted on the GeneAmp 5700 sequence Detection System (Applied Biosystems), and signal was detected by the GeneAmp 5700 SDS software (Applied Biosystems). All values were reported relative to β-actin expression.
We used competitive ELISA to measure 6-keto-PGF1α. Human whole lung lysates, antibody (mouse anti-PGE or rabbit anti–6-keto-PGF1α), and tracer were added to wells on plates coated with anti-rabbit (6-keto-PGF1α) antibodies. The tracer is 6-keto-PGF1α linked to acetylcholinesterase. After overnight incubation at 4°C, plates were washed, and Ellman's reagent was added (acetylthiocholine iodide and 55′dithiobis-[2-nitrobenzoic acid] in a 1-M phosphate buffer). The samples were read in a spectrophotometer at 405 nm. Antibodies and tracer were obtained from Cayman Chemicals.
Transgenic mice were developed using a construct consisting of the human surfactant protein C promoter and full-length rat prostacyclin synthase cDNA (24). The surfactant protein C promoter allows targeted expression to alveolar and distal airway epithelial cells (25). We conducted genotyping of animals by performing PCR on genomic DNA isolated from tails as described previously. Each line was propagated as heterozygotes. Transgenic mice were bred with wild-type FVB/N (Jackson Laboratories, Bar Harbor, ME) mice to produce the experimental transgenic mice and transgenic-negative littermates, which were used as controls in all of the experiments. Transgenic (n = 3) and wild-type (n = 3) mice were exposed to 6 months of mainstream cigarette smoke. All animal studies were conducted under an approved protocol from our local Institutional Animal Care and Use Committee (IACUC).
All statistical analyses in tissue scoring and in vitro were conducted by t test (p < 0.05)
We performed immunohistochemistry for PGI2S on human lung tissue from four individuals with emphysema and three individuals with normal lung parenchyma. Scoring of endothelial cells for PGI2S expression as a factor of total endothelial cells revealed that PGI2S was decreased primarily in the endothelial cell monolayer of arterioles in the lungs from patients with pulmonary emphysema (Figure 1A) compared with normal subjects (Figures 1B and 1E). PGI2S staining in medium-sized pulmonary vessels in emphysema and nondiseased lung was comparable (Figures 1C, 1D and 1F) Western analysis of whole-lung lysates for PGI2S (Figure 2A) confirmed that protein expression was decreased (by ~ 50% by densitometry) in lung tissue extracts from the emphysema lungs compared with normal control lung tissue extracts.
To examine whether the reduced amount of lung tissue PGI2S protein was associated with decreased amounts of PGI2, we measured (by ELISA) the stable metabolite of PGI2 (6-keto-PGF1α). The amount of 6-keto-PGF1α measured per milliliter of lung tissue homogenate was reduced in lung tissue samples from patients with emphysema by 75% (Figure 2B). Because the amount of the enzyme protein and the product of the terminal synthetic reaction were reduced in emphysema lung tissues, we questioned whether the synthesis defect was related to a reduced expression of the PGI2S gene. We examined lung tissue samples using real-time PCR and detected decreased PGI2S mRNA expression in the lungs from patients with emphysema when compared with normal lung tissue (Figure 2C).
We treated primary HPMVECs (passages 4–6) with CSE (0.5%, 1%), which decreased PGI2S gene expression in a dose-dependent manner with a maximum effect at 24 hours after incubation (Figure 3A). For the remainder of the experiments, we treated endothelial cells with 0.5% and 1% CSE and examined gene expression at 24 hours. COX-1 gene expression was decreased by 1% CSE (Figure 3B), whereas COX-2 and cPLA2 were significantly increased (tenfold and threefold, respectively) (Figures 3C and 3D).
We exposed HPMVECs to 1 and 10 μM of unsaturated aldehydes (acrolein and crotonaldehyde) and saturated aldehydes (acetaldehyde and propionaldehyde). Acrolein caused a statistically significant reduction in PGI2S gene expression 24 hours after exposure at both concentrations (Figure 4A) and PGI2S protein expression 48 hours after exposure (Figure 4B). No significant change in COX-2 gene expression was noted after acrolein exposure (Figure 4C). None of three other aldehydes (at 1 and 10 μM) tested caused a similar decrease in PGI2S expression (Figure 4D). Propionaldehyde induced PGI2S gene expression.
We examined whether experimental antioxidant strategies would protect against the CSE-related decrease of PGI2S expression in HPMVECs. Four-hour preincubation of HPMVECs with NAC (1 mM) followed by CSE treatment and examination at 24 hours did not prevent down-regulation of PGI2S gene expression by 1% CSE (Figure 5A). Pretreatment with NAC prevented the CSE-mediated induction of COX-2 gene expression (Figure 5B). Four-hour preincubation of endothelial cells with an SOD mimetic (100 U/ml) (Figures 6A and 6B), catalase (100 U/ml) (Figures 6C and D), diphenyleneiodonium chloride (0.1 μM) (nitric oxide inhibitor) (Figures 6E and 6F), or l-NAME (nitric oxide synthase [NOS] inhibitor) (1 mM) (Figures 7A and 7B) did not affect the CSE changes in PGI2S and COX-2 gene expression.
Incubation of HPMVECs with 1 and 2% CSE increased the apoptosis rate at 24 hours from 3% in control cells to 9% (as measured for annexin V by fluorescence) in cells treated with 2% CSE (Figure 8A). We confirmed the induction of apoptosis by annexin V and caspase 3/7 analysis (Figure 8B). Given the research to support that PGI2 is important to vascular integrity and likely confers protective effects to the vasculature after stressors, we sought to determine whether PGI2 (analog iloprost) would prevent apoptosis induced by CSE. Pretreatment of HPMVECs with the prostacyclin analog iloprost (1 μM and 10 μM) for 30 minutes before 2% CSE exposure significantly reduced endothelial apoptosis as measured by caspase 3/7 activity (Figure 8C).
We have previously reported that transgenic mice with lung-specific overexpression of PGI2S were partially protected from chemically and cigarette smoke–induced lung tumors (24). We attempted to determine if lung-specific PGI2S overexpression conferred protection to the pulmonary endothelium after 6 months of mainstream cigarette smoke exposure. Wild-type (Figure 9A) and PGI2S transgenic mice (Figure 9B) (FVB background) were exposed to 6 months of cigarette smoke. Immunohistochemistry and subsequent scoring for endothelial cleaved caspase activity per total endothelial cells revealed a significant reduction in caspase 3 expression in transgenic animals (0.3946 ± 0.06583) compared with wild-type littermates (0.6521 ± 0.07224), suggesting decreased apoptosis (Figure 9C).
COPD is predominantly related to cigarette smoke exposure. COPD is associated with varying degrees of lung function abnormalities and can lead to respiratory failure. However, it is less appreciated that major causes of morbidity and mortality in these patients are pulmonary hypertension and cardiovascular manifestations leading to coronary artery disease and stroke. Impaired endothelial cell–dependent vasodilation, inflammation, apoptosis, and proliferation are important to the endothelial dysfunction observed in smoke angiopathy and perhaps in the pathogenesis in emphysema. Endothelial cell dysfunction has been implicated in the pathogenesis of COPD (9). Whether or not cigarette smoke exposure alters lung cell PGI2 expression has been minimally explored. Previously, researchers demonstrated that cigarette smoke condensate impairs PGI2 release in human umbilical vein cells (20). In that study, inhibition seemed to occur at the level of arachidonic acid mobilization rather than directly on downstream eicosanoids (20). In our study, we used a comprehensive approach incorporating human tissue analysis, in vitro study, and transgenic animal modeling to determine if PGI2 expression is biologically relevant to the pulmonary vasculature in smoking-related lung disease. We report that pulmonary endothelial PGI2S expression is decreased in the arteriolar endothelium in emphysema and that, in HPMVECs, cigarette smoke and in particular acrolein suppress PGI2S expression. Furthermore, we report that PGI2 partially prevents CSE-induced apoptosis of the pulmonary endothelium in vitro and in a murine transgenic model of lung-specific PGI2S overexpression. The observed decrease in lung tissue PGI2S gene and protein expression may be multifactorial and may be caused by oxidant stress, nitric oxide–related suppression of the PGI2S protein release, and altered transcriptional control (26).
Cigarette smoke contains over 4,000 compounds (27). Researchers have identified acrolein as playing a role in cigarette smoke–induced lung toxicity and potentially in cigarette smoke–induced lung cancer (28–30). Acrolein mediates pulmonary inflammation through the induction of inflammatory cytokines and the inhibition of neutrophil apoptosis (31). These biological changes are likely important to COPD pathogenesis and possibly lung tumorigenesis. In addition, acrolein contributes to endothelial dysfunction through the depletion of glutathione and subsequent oxidative stress (28). Experimentally, we demonstrated that CSE and acrolein reduce the expression of the PGI2S gene and that CSE increases the expression upstream mediators of the eicosanoid pathway (COX-2 and cPLA2) in HPMVECs.
The induction of oxidative stress by cigarette smoke has been well documented (32). We postulated that the observed imbalance in eicosanoid gene expression may be mediated by oxidative stress. Our first approach was to incubate HPMVECs with NAC, DPI, an SOD mimetic, or a catalase to examine whether the cells would be protected against CSE-related change in PGI2S gene repression. In addition, we used pharmacologic inhibition of endothelial NOS by preincubating the cells with l-NAME. None of these strategies, whether they were intended to increase intracellular glutathione (NAC) or inhibit the NADPH-oxidoreductase (DPI), prevented the CSE-related repression of the PGI2S gene. However, NAC did prevent the induction of COX-2 gene expression. Our findings are consistent with previous studies demonstrating that NAC interferes with COX-2 induction in other inflammatory processes (33). Although CSE provides severe oxidative stress to the cultured endothelial cells, a number of antioxidant pretreatment strategies failed to prevent CSE-related decrease in PGI2S gene expression.
We selected HPMVECs as our experimental system because small-vessel and alveolar capillary endothelial cells undergo apoptosis in human and experimental lung emphysema (1). Using a standardized CSE exposure protocol, we found that CSE differentially and in a dose-dependent manner directly affects PGI2S gene expression and causes apoptotic cell death in HPMVECs. Although CSE induced a degree of endothelial cell apoptosis, the decrease in PGI2S mRNA is unlikely the result of apoptotic cell loss. Immunohistochemical analysis suggests that the observed whole lung decrease in PGI2S expression is the result of decreased endothelial expression rather than fewer numbers of endothelial cells. We demonstrate that pretreatment with the prostacyclin analog iloprost conferred a partial protective antiapoptotic effect. Finally, lung-specific overexpression of PGI2S resulted in decreased caspase 3 activity after long-term smoke exposure in a transgenic murine model. Our findings are consistent with several recent studies demonstrating that PGI2 has antiapoptotic effects in other organ systems (34–36).
Ermert and colleagues previously identified decreased expression of PGI2S in lung tumor–associated vasculature (37), Our laboratory has observed potent antitumorigenic properties of PGI2S in murine models of lung-specific overexpression, suggesting that PGI2S may be a tumor suppressor gene (24). In addition, persistent PGI2S expression in human lung tumors seems to be an independent factor in survival (38). There are many potential mechanisms for the observed decrease expression of PGI2S, including epigenetic modification and transcriptional regulation. Our findings suggest that acrolein is one key component of cigarette smoke responsible for PGI2S suppression and may be relevant to COPD and lung tumorigenesis.
In conclusion, PGI2S gene expression is decreased in lung microvascular endothelial cells by CSE. Furthermore, CSE and acrolein have direct suppressive effects on PGI2S gene expression compared with upstream mediators. Whether these experiments reflect the mechanism of decreased PGI2S in the lungs of chronic smokers is unclear. There are several potential mechanisms for PGI2S suppression, including transcriptional regulation by methylation, promoter base pair rearrangement by oxidative stress, and alteration in transcriptional binding factors. The observation that acrolein decreased PGI2S expression, whereas several other aldehydes did not, suggests a mechanism specific to acrolein. The mechanisms by which acrolein suppresses PGI2S remain unexplored. There are data to suggest that acrolein may interfere with transcriptional regulation of many genes by preferentially binding to CpG sites (29, 39). In addition, acrolein may affect PGI2S through generation of reactive oxygen species (40). An imbalance in eicosanoid expression may be relevant to the observed vascular toxicity of acrolein. We are currently investigating potential mechanisms of transcriptional regulation of PGI2S by acrolein.
Our use of CSE as an experimental model is limited in that cigarette smoke exposure of cultured endothelial cells may not adequately represent the actual exposure in vivo (41). However, our studies demonstrate reduced PGI2S expression in the pulmonary endothelium of long-standing smokers with emphysema. To our knowledge, the present study represents the first report that PGI2 expression is diminished in the pulmonary endothelium in emphysema and that PGI2 confers antiapoptotic effects to the pulmonary endothelium after acute and chronic cigarette smoke exposure.
The authors thank Dr. Brian Freed for his thoughtful review of this manuscript and invaluable discussion during the course of preparation, Dr. David Dewitt (Michigan State) for providing PGI2S antibody, and Dr. Russ Bowler (National Jewish Medical Center) for providing antioxidant reagents. The authors also thank the University of Colorado lung tissue bank for providing the human lung tissue samples for analysis.
Supported by grants from the National Heart, Lung, and Blood Institute Mentored Scientist Award HL077717 (S.P.N.S.), a SPORE in Lung Cancer Career Development Award (S.P.N.S.), a FAMRI CIA Award (R.S.S.), a Veterans' Affairs Merit Award (R.L.K.), and the SPORE in Lung Cancer (CA58187) Project 3 (M.W.G.).
Originally Published in Press as DOI: 10.1164/rccm.200605-724OC on January 25, 2007
Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript.