- Email: [email protected]
- Mice
Journal of Surgical Research 146, 135–142 (2008) doi:10.1016/j.jss.2007.04.040
The Select Cyclooxygenase-2 Inhibitor Celecoxib Reduced the Extent of Atherosclerosis in Apo E-/- Mice Shushan Jacob, M.D.,* Lisa Laury-Kleintop, Ph.D.,† and Susan Lanza-Jacoby, Ph.D.*,1 *Department of Surgery, Thomas Jefferson University, Philadelphia, Pennsylvania and †Lankenau Institute for Medical Research, Wynnewood, Pennsylvania Submitted for publication February 1, 2007
Many investigators have suggested that immune activation may trigger the atherosclerotic process. The benefits of aspirin in preventing myocardial infarction have been attributed, in part, to its anti-inflammatory effects. Several reports have documented that cyclooxygenase (COX)-2 is up-regulated in human and mouse atherosclerotic lesions. To clarify the role of COX-2 in atherosclerosis, we conducted a study to test whether the COX-2 inhibitor, celecoxib, prevents the development and progression of the atherosclerotic process. We have used the apo E-/- mouse, a relevant animal model of atherosclerosis that develops fibrofatty lesions similar to human atherosclerosis. Treatment of 4-wk old apo E-/- mice with a standard rodent no. 5020 diet supplemented with 900 ppm of celecoxib for 16 wk led to an 81% reduction in lesion size. The mean lesion area per section (mean ⴞ SD) of proximal aorta from the apo E-/- mice fed the diet with celecoxib (33,991 ⴞ 7863 m 2, P < 0.001) was significantly less than that of the untreated apo E-/- mice (183,401 ⴞ 36,212 m2). There were no lesions detected in the C57B1/6 mice. Immunohistochemistry of the ileum revealed that there was 80% reduction in staining for intercellular adhesion molecule and 60% reduction in staining for vascular cell adhesion molecule in the celecoxib treated mice. The protective effect of celecoxib was not maintained when the mice were switched after feeding the celecoxib-supplemented diet to the control 5020 diet for an additional 10 wk. These findings demonstrate that selective inhibition of the COX-2 enzyme with celecoxib prevented the development of atherosclerotic lesions in the proximal aortas from apo E-/- mice. One of the possible mechanisms is reduction in expression of the endothelial cell adhesion 1
To whom correspondence and reprint requests should be addressed at Lankenau Institute for Medical Research, 100 Lancaster Ave., Wynnewood, PA 19096. E-mail: Susan.Lanza-Jacoby@jefferson. edu.
cell molecules intercellular adhesion molecule and vascular cell adhesion molecule, which plays a key role in the recruitment of inflammatory cells during the early stages of atherogenesis. © 2008 Elsevier Inc. All rights reserved. Key Words: atherosclerosis; cyclooxigenase-2; adhesion molecules; apo E-/- mice; cyclooxygenase-2 inhibitors. INTRODUCTION
Current data support the hypothesis that atherosclerosis is an inflammatory disease [1, 2]. The beneficial effect of aspirin in reducing the risk of myocardial infarction is, in part, attributable to its antiinflammatory action. Increased inflammation in the plaque is associated with plaque rupture, leading to thrombosis in the coronary vessel [3]. Nonsteroidal anti-inflammatory drugs (NSAIDS) inhibit both isoforms of cyclooxygenase (COX), the enzymes controlling the conversion of arachidonic acid into prostaglandins (PG). COX-1 is the constitutive form of the enzyme involved in normal cell functioning and COX-2 is induced by cytokines, bacterial infections, shear stress, or growth factors. The inhibition of COX-1 by NSAIDS prevents synthesis of protective prostaglandins that synthesize the mucous barrier of the gastric mucosa and blocks the formation of vasodilator, prostacyclin (PGI 2) by kidney cells, leading to the side effects of gastric ulcers and renal toxicity. COX-2 has been found to have particular relevance to inflammation and atherosclerosis in two ways. The first is the suggestion that activation of COX-2 in human monocytes may be able to generate prostaglandin PGF2␣ [4], which is mitogenic, leading to cellular proliferation and vasoconstriction [5] and, thus, may play a role in the genesis of atherosclerosis. The second is that COX-2 with its nuclear loca-
135
0022-4804/08 $34.00 © 2008 Elsevier Inc. All rights reserved.
136
JOURNAL OF SURGICAL RESEARCH: VOL. 146, NO. 1, MAY 1, 2008
tion may produce eicosanoids that are active within the nucleus, and these eicosanoids may modulate transcriptional events [6]. Several reports have documented that COX-2 is upregulated in human and mouse atherosclerosis [7–11]. The studies with COX-2 inhibitors in animal models of atherosclerosis have produced variable results depending on the age when starting treatment, the length of treatment, the strain of mice, and the difference in COX-2 inhibitors [12–18]. The purpose of this study was to evaluate the effect of the COX-2 inhibitor, celecoxib, on early development of atherosclerosis and on the progression of established lesions. We also investigated whether any potential beneficial effects of early intervention would be maintained when celecoxib treatment was removed from the diet. One of the earliest detectable cellular responses in the formation of atherosclerotic lesions is leukocyte adherence to the endothelium by expression of specific leukocyte adhesion molecules expressed by the endothelium. To date there are no studies on the effects of the select COX-2 inhibitors on the adhesion molecules in the vascular tissues. In this study, we investigated whether celecoxib would be effective in preventing the progression of atherosclerosis in the apo E-/- mouse, a transgenic mouse model prone to develop atherosclerosis. We also evaluated whether this protection is associated with decreased expression of the adhesion proteins, intercellular adhesion molecule (ICAM)-1, and mouse vascular cell adhesion molecule (VCAM).
ficed at 26 wk of age. The heart and aortic tree were perfused through the left ventricle with iced phosphate buffered 0.9% saline (PBS) for 10 min. The heart was carefully dissected and removed. The upper half of the heart containing the aortic origin was separated and embedded in Tissue-Tek (Fischer Scientific, Newark, DE) OCT compound contained in cryomolds. These were frozen at ⫺70°F and sequential 10 m thick sections were cut from the base toward the apex of the heart until the aortic valve leaflets appeared. From this point, sequential sections, each 10 microns thick were cut from each mold and stained with 0.5% Oil Red O in propylene glycol for 4 h and counterstained in Mayer’s hematoxylin for 1 min. Quantification of atherosclerosis was performed using a computer assisted image analysis. A video camera mounted to a Nikon upright light microscope (Nikon Instruments, Melville, NY) was used to view the aortic sections. The image was captured and analyzed by NIH Image software. Data were collected from each of the 14 sections taken from the proximal aortic root, averaged, and expressed as m 2 per section.
METHODS
Following perfusion of heart and aortic tree with PBS, the superior mesenteric vein was incised and the ileum was washed free of blood by perfusion with PBS flowing through the aortic tree and out through the superior mesenteric vein. The ileum was then perfused with iced 4% paraformaldehyde in PBS for 20 min. A 3- to 4-cm segment of ileum was isolated from the perfused intestine and fixed in 4% paraformaldehyde for 90 min at 4°C. The ileum was then cut into smaller rings, and the tissue dehydrated in graded acetone washes at 4°C. The ileum rings were embedded in plastic (Immunobed; Polysciences Inc., Warrington, PA), and 4 m thick sections were cut and transferred to Vectabond-coated slides (Vector Laboratories, Burlingame, CA). Immunohistochemical localization of ICAM-1 and VCAM-1 was investigated by using the avidin-biotin immunoperoxidase technique (Vectastain ABC Reagent; Vector Laboratories) according to a previously described method [20]. Tissue sections were treated with 0.25% trypsin to improve reagent penetration. Blocking serum was applied to the tissue sections for 30 min to reduce non-specific binding, and then the tissue sections were incubated for 24 h with specific primary antibodies. ICAM-1 and VCAM-1 were identified with the specific mouse monoclonal IgG2a at a dilution of 1/100. The tissue was then incubated with the biotinylated secondary antibody, and peroxidase staining was carried out using 3,3=-diaminobenzidine. Control preparations consisted of omission of either the primary or the secondary antibody. Expression of adhesion molecules was determined by microscopic observation of the brown peroxidase reaction product on the microvascular endothelium of the tissue sections. Positive staining was defined as a vessel displaying brown reaction product on ⬎50% of the circumference of its endothelium. Fifty ileal venules per tissue section were examined in each of 20 sections, and the percentage of positive staining vessels was tallied.
General Research Design Male apo E-/- and C57B1/6 mice were used in this study. The original breeding pairs of apo E-/- mice were obtained from the Gladstone Institute (San Francisco, CA); breeders for the C57B1/6 mice were purchased from the The Jackson Laboratory (Bar Harbor, ME). Apo E-/- mice were generated on a C57B1/6 background. At 4 wk of age, apo E-/- and C57B1/6 mice were assigned to the following groups: Group-1: apo E-/- fed no. 5020 diet (Purina, St. Louis, MO); Group-2: apo E-/- fed no. 5020 ⫹ 900 ppm celecoxib; and Group-3: C57B1/6 fed no. 5020 diet. The dose of celecoxib is based on previous reports that this amount was well tolerated by the mice when fed for over 1 y [19]. Diets were made weekly in our laboratory and stored at ⫺20°C. Fresh diet was provided every third day. Three mice were placed in one metabolic cage; food intake and weight gain were monitored weekly. At 16 wk of age, 12 mice from each of Groups 1, 2, and 3 were sacrificed. The remaining 10 mice in Group 1 were switched to the no. 5020 ⫹ celecoxib diet and the remaining 10 mice in Group 2 were switched to the no. 5020 diet. Groups 1 and 2 were re-designated as Groups 4 and 5, respectively, and fed an additional 10 wk on the new diet before being sacrificed. At the time of sacrifice, blood was obtained from the retro-orbital plexus for determination of the concentration of cholesterol and prostaglandin E 2 (PGE 2) in the serum.
Analysis of Atherosclerotic Lesions Groups 1, 2, and 3 were sacrificed at 16 wk of age and the newly designated Groups 4 and 5 (previously Groups 1 and 2) were sacri-
Immunohistochemical Analysis on Mouse Aorta Pieces of aorta were fixed in 10% neutral buffered formalin for 24 h, embedded in paraffin, and sectioned. For immunohistochemistry, 10% rabbit serum was added to unstained, frozen sections from the proximal aorta, to block nonspecific activity. This was followed by addition of rabbit polyclonal antibodies to mouse COX-1 (Santa Cruz), mouse COX-2 (Cayman), mouse ICAM-1 and mouse VCAM (Santa Cruz). The sections are then treated with biotinylated rabbit anti-mouse antibody IgG followed by streptavidin-horseradish peroxidase. The peroxidase reaction was developed with diaminobenzidine and counterstained with hematoxylin. The sections were viewed under the Nikon light microscope and sections from the different groups were compared for difference in intensity of the dark brown staining between antigen and antibody. The images were analyzed using NIH Image software.
Immunohistochemical Analysis on Mouse Ileum
JACOB, LAURY-KLEINTOP, AND LANZA-JACOBY: THE SELECT COX-2 INHIBITOR CELECOXIB
137
FIG. 1. (A) Representative photomicrograph of Oil-Red-O stained aortic root from Apo E-/- mice treated with celecoxib. Mice were fed the no. 5020 diet supplemented with 900 ppm celecoxib for 16 wk. Sections are shown at ⫻40 magnification areas of lipid accumulation. (B) The extent of atherosclerotic lesions was quantified with Oil-Red-O staining and analyzed by NIH Image software. Data were collected from each of the 14 sections taken from the proximal aortic root, averaged, and expressed as m 2 per section. Bars indicate means ⫾ SD for n ⫽ 12, *P ⬍ 0.001 by Student’s t-test. (Color version of figure is available online.)
Determination of PGE 2
RESULTS
Enzyme-linked immunosorbent assay kits (R and D Systems, Minneapolis, MN) were used for determination of PGE 2 concentration in the serum. Samples were diluted 1:2 in assay buffer. The assay was performed in accordance with the manufacturer’s protocols in a microplate reader at a wavelength of 405 nm. Limits of detection for these assays were ⬍15.9 pg/mL (PGE 2). Samples were assayed in duplicate.
We evaluated whether celecoxib treatment early in life would prevent the development of aortic atherosclerosis in the apo E-/- mice. Providing celecoxib in the diet at a dose of 126 mg/kg beginning at 4 wk of age prevented the development of atherosclerosis in apo E-/- mice. The mean area of aortic lesions from the apo E-/- mice fed the no. 5020 ⫹ celecoxib diet (33,991 m 2 ⫾ 7863) was 81% (P ⬍ 0.001) less than the apo E-/- mice fed no. 5020 diet (183,401 m 2 ⫾ 36,212) (Fig. 1A and B). The protective effect of celecoxib was not associated with a reduction of serum cholesterol levels in both apo E-/- and C57B1/6 mice (data not included).
Statistical Analysis Results were expressed as mean ⫾ SD. The extent of atherosclerosis was analyzed by analysis of variance and subsequently by Student’s unpaired two-tailed t-test.
138
JOURNAL OF SURGICAL RESEARCH: VOL. 146, NO. 1, MAY 1, 2008
FIG. 2. Celecoxib does not alter the progression of atherosclerosis in apo E-/- mice. Mean lesion size in the proximal aorta was quantified after 16 wk of feeding apo E-/- mice with the no. 5020 diet. Then a subgroup of these mice were switched to the no. 5020 diet supplemented with 900 ppm celecoxib for an additional 10 wk and mean lesion size was quantified. Bars indicate means ⫾ SD for n ⫽ 12.
We also questioned whether inhibition of COX-2 by celecoxib would alter the progression of atherosclerosis. After feeding the no. 5020 diet for 16 wk some mice in Group 1 were switched to a diet supplemented with celecoxib. Figure 2 shows that lesion size continued to progress when celecoxib was started after feeding the control diet for 16 wk. The lesion size increased 52% from 183, 401 ⫾ 36,212 m 2 at the 16th wk of feeding the control diet, to 279,624 ⫾ 43,599 m 2 after an additional 10 wk of feeding the celecoxib diet. To determine whether continuous exposure to celecoxib was required to maintain the protective effects, celecoxib-treated apo E-/- mice were switched at the end of the 16 wk feeding period to the control diet for an additional 10 wk. Lesion size increased from 33,992 m 2 ⫾ 7863 at 16 wk to 235,965 m 2 ⫾ 1434 at 26 wk (Fig. 3). Immunohistochemical analysis demonstrated that the aorta of the apo E-/- mice stained positive for COX-1 and COX-2. Staining for both COX-1 and COX-2 antigen appear to be proportional to the lesion size with less staining in the apo E-/- mice fed celecoxib (Fig. 4A). Staining for ICAM and VCAM also appeared to be less in aortas from the apo E-/- mice fed celecoxib (Fig. 4B). To determine whether the reduction in ICAM and VCAM was due to a direct effect of the celecoxib treatment or the decrease in plaque size, we analyzed the protein expression in the ileum. Twenty sections were studied in each mouse and 50 vessels were studied in each section. The expression of adhesion molecules was determined by microscopic observation of the brown peroxidase reaction product on the microvascular endothelium of the tissue sections. Positive staining was defined as a vessel displaying a brown reaction product
on ⬎50% of the circumference of its endothelium. Immunohistochemical analysis showed an 80% reduction in staining for ICAM and 60% reduction in staining for VCAM in apo E-/- mice fed celecoxib compared with the mice fed the control diets (Fig. 5A and B). Since COX-2 is increased in plaque [7–11] and COX-2derived PGE 2 may facilitate arterial thrombosis [21], we evaluated the PGE2 concentration in the serum. The PGE 2 concentration of the serum from the celecoxib-fed mice (4.4 pg/mL) was significantly reduced in comparison to the control apo E-/- mice (12.7 pg/mL), (n ⫽ 6), P ⬍ 0.01).
FIG. 3. Continuous treatment with celecoxib is required to maintain protective effects. Mean lesion size in the proximal aorta was quantified after 16 wk of feeding apo E-/- mice with the no. 5020 diet supplemented with 900 ppm celecoxib. Then a subgroup of these mice were switched to the no. 5020 diet for an additional 10 wk and mean lesion size was quantified. Bars indicate means ⫾ SD for n ⫽ 12.
JACOB, LAURY-KLEINTOP, AND LANZA-JACOBY: THE SELECT COX-2 INHIBITOR CELECOXIB
139
FIG. 4. Immunohistochemistry in sections from proximal aorta from Apo E-/- mice treated with celecoxib. After 16 wk of treatment with the control no. 5020 and no. 5020 diet supplemented with 900 ppm celecoxib, mice were sacrificed. (A) COX-1 and COX-2 were detected by rabbit polyclonal antibodies to mouse COX-1 and COX-2 in proximal aorta from apo E-/- mice. (B) 0 VCAM and ICAM were detected by specific mouse monoclonal IgG2a. (Color version of figure is available online.)
DISCUSSION AND CONCLUSIONS
There is increased evidence to suggest that COX-2 is involved in the early development of atherosclerosis. Normal blood vessels contain COX-1 but not COX-2 [3]. In contrast, atherosclerotic lesions from humans [8 –12] and animals [12–16, 18] have high expression levels of COX-2. Recent studies show that deletion of
COX-2 in both the low-density lipoprotein receptor (LDLR)-/- mice and the apo E-/- mice significantly reduced lesion area of the proximal aorta [14, 15]. These studies also demonstrate that beginning treatment at 8 to 10 wk of age with the COX-2 inhibitors rofecoxib or NS-398 led to a significant reduction in the extent of atherosclerosis in LDLR-/- and apo E-/- mice [14, 15]. Here, we show that celecoxib treatment begun at 4 wk
140
JOURNAL OF SURGICAL RESEARCH: VOL. 146, NO. 1, MAY 1, 2008
FIG. 5. Celecoxib decreases immunohistochemical staining for ICAM and VCAM in ileal microvessels from apo E-/- mice. Percentage of vessels staining positive for ICAM and VCAM are shown for control, untreated C57Bl/6 mice, apo E-/- mice fed the no. 5020 diet, and apo E-/- mice fed the no. 5020 diet supplemented with 900 ppm celecoxib. Bar heights represent average values for three mice per group. Twenty sections were studied in each mouse and 50 vessels were evaluated in each section.
of age also prevents atherosclerotic lesion development in apo E-/- mice and that the protective effect was associated with a reduction in ICAM and VCAM protein expression. We further demonstrate that the protective effect of celecoxib is not maintained when treatment is stopped. In contrast to the protective effects of rofecoxib and celecoxib, other studies have found that the select COX-2 inhibitors, SC236 and MF Tricyclic had no effect on lesion development in apo E-/- mice when treatment was initiated at 6 and 8 wk of age, respectively [16, 17]. These studies suggest that differences in the selectivity of the COX-2 inhibitors may explain the contrasting results. In a previous study, we showed that the mean serum concentration of celecoxib was 5.2 ⫾ 0.6 M when mice were fed a similar dose of celecoxib [19], which agrees with previous reports [22]. The IC 50 for COX-1 and COX-2 inhibition was reported to be 6.7 ⫾ 0.09 M and 0.87 ⫾ 0.18 M, respectively [23]. Based upon these IC50 concentrations, the results of this study appear to be due to selective COX-2 not COX-1 inhibition. While COX-2 may have significance in the early development, its role in advanced atherosclerotic lesions is less clear. Recent clinical trials have found a 5-fold increase in atherothrombotic events after 18 months of rofecoxib use [24], which resulted in the withdrawal of rofecoxib from the market. Previous laboratories have suggested that the increase in thrombotic events associated with the use of COX-2 inhibitors may be related to the ability of COX- 2 inhibitors to block PGI 2 without altering thromboxane levels [11, 13]. PGI 2 suppresses platelet activity and increases bleeding while thromboxane increases platelet activity. However, the role of COX-2 inhibitors in plaque rupture may be more complex; a more recent study demonstrated that celecoxib
treatment reduced incidence of abdominal aortic aneurysms in angiotensin II-induced mouse model of aortic aneurysms [25]. COX-2 inhibitors appear to have no beneficial effects in reversing or altering the progression of atherosclerosis in animals with established lesions. A recent study conducted with a new nonselective COX-2 inhibitor INDO-PA was shown to reduce aortic PGE 2 levels and early atherosclerosis but had less of an effect on the progression of established lesions [26]. Another study found that there was no change in lesion area after feeding apo E-/- mice for 15 wk at dose of 75 mg/kg/d [27]. This study also showed that there was no effect of celecoxib on the composition or stability of advanced lesions indicating that celecoxib administration may not alter inflammatory mediators in established lesions. We also observed that there was no change in lesion size when a higher dose of celecoxib (126 mg/kg body weight/d) was fed to 20 wk-old apo E-/mice. These observations are consistent with the suggestion that COX-2 inhibition blocks early events essential to atherosclerosis. One of the earliest detectable cellular responses in the formation of atherosclerotic lesions is leukocyte adherence to the endothelium. Leukocyte recruitment occurs in lesion-prone areas of the arterial tree by expression of specific adhesion molecules, which persists as long as the condition of hypercholesterolemia continues [28]. VCAM-1, together with P- and E-selectin, is involved in the first step of tethering and rolling of monocytes and lymphocytes, and in the second stage mediate arrest and firm adhesion. ICAM-1 helps mediate arrest and firm adhesion of lymphocytes, monocytes, and neutrophils. Increased expression of VCAM has been shown to be associated with atherosclerotic lesion development [28, 29]. Several studies have demonstrated that NSAIDS
JACOB, LAURY-KLEINTOP, AND LANZA-JACOBY: THE SELECT COX-2 INHIBITOR CELECOXIB
reduce atherosclerosis in animal models [13, 30, 31] and also reduce lymphocyte-endothelial cell adhesion molecules [32, 33]. Ibuprofen has been shown to inhibit VCAM-1 [30] and diclofenac inhibited lipopolysaccharide-induced up-regulation of VCAM-1, ICAM-1, and E-selectin in endothelial cells [34]. Diclofenac, indomethacin, and aceclofenac also reduced adhesion of peripheral blood lymphocytes to human umbilical vein endothelial cells by decreasing expression of VCAM-1 [35]. A recent study showed that celecoxib reduced protein expression of ICAM and P-selectin in mouse brain tissue [36]. Our study is the first report that a specific COX-2 inhibitor, celecoxib, reduced protein expression of VCAM and ICAM in the vascular endothelium from the atherosclerosis prone apo E-/- mice. These findings support the role of inflammation in early lesion development. The increase in adhesion molecule expression in the lesion-prone areas represents an inflammatory response and their reduction may be one mechanism by which celecoxib prevents plaque formation. While this study further supports the premise that blocking COX-2 before the onset vessel injury will prevent plaque development, our data further demonstrate that inhibition of COX-2 must be continuous to maintain the arrest in lesion development. Once celecoxib treatment is discontinued, lesions continued to increase in size comparable to control apo E-/- mice. The time of initiating celecoxib treatment also is an important factor in arresting plaque development, as celecoxib is unable to arrest plaque growth in mice with established lesions. Our observation that celecoxib will reduce expression of ICAM and VCAM suggests that in apo E-/- atherosclerosis model, celecoxib retards the development of plaque and that this effect is paralleled by a suppression of the inflammatory factors, ICAM, and VCAM. ACKNOWLEDGMENTS The authors thank Chris Munery for her technical assistance. This study was supported in part by a grant from the American Heart Association to SL-J.
REFERENCES 1.
Ross R. The pathogenesis of atherosclerosis: Perspective for the 1990s. Nature 1993;362:801.
2.
Alexander RW. Inflammation and coronary artery disease. N Engl J Med 1994;331:4687.
3.
Mitchell JA, Warner TD. COX isoforms in the cardiovascular system: Understanding the activities of the non-steroidal antiinflammatory drugs. Nat Rev Drug Discov 2006;5:75.
4.
van der Wal AC, Becker AE, van der Loos CM, et al. Site of intimal rupture or erosion of thrombosed coronary atherosclerotic plaques is characterized by an inflammatory process irrespective of the dominant plaque morphology. Circulation 1994; 89:36.
141
5.
Pratico D, FitzGerald GA. Generation of 8-epiprostaglandin F2alpha by human monocytes. Discriminate production by reactive oxygen species and prostaglandin endoperoxide synthase-2. J Biol Chem 1996;271:8919.
6.
Fukunaga M, Makita N, Roberts LJ, et al. Evidence for the existence of F2-isoprostane receptors on rat vascular smooth muscle cells. Am J Physiol 1993;264:C1619.
7.
Serhan CN. Signalling the fat controller. Nature 1996;384:23.
8.
Baker CSR, Hall RJC, EvansTJ, et al. Cyclooxygenase-2 is widely expressed in atherosclerotic lesions affecting native and transplanted human coronary arteries and co-localizes with inducible nitric oxide synthase and nitrotyrosine particularly in macrophages. Arterioscler Thromb Vasc Biol 1999;19:646.
9.
Schonbeck U, Sukhova GK, Graber P, et al. Augmented expression of cyclooxygenase-2 in human atherosclerotic lesions. Am J Pathol 1999;155:1281.
10.
Hong BK, Kwon HM, Lee BK, et al. Coexpression of cyclooxygenase-2 and matrix metalloproteinases in human aortic atherosclerotic lesions. Yonsei Med J 2000;41:82.
11.
Belton O, Byrne D, Kearney D, et al. Cyclooxygenase-1 and -2 dependent prostacyclin formation in patients with atherosclerosis. Circulation 2000;102:840.
12.
Stemme V, Swedenborg J, Claesson H, et al. Expression of cyclooxygenase-2 in human atherosclerotic carotid arteries. Eur J Vasc Endovasc Surg 2000;20:146.
13.
Pratico D, Tillmann C, Zhang ZB, et al. Acceleration of atherogenesis by COX-1-dependent prostanoid formation in low density lipoprotein receptor knockout mice. Proc Natl Acad Sci USA 2001;98:3358.
14.
Burleigh ME, Babaev VR, Oates JA, et al. Cyclooxygenase-2 promotes early atherosclerosis lesion formation in LDL receptor-deficient mice. Circulation 2002;105:1816.
15.
Burleigh ME, Babaev VR, Yancey PG, et al. Cyclooxygenase-2 promotes early atherosclerosis lesion formation in Apo E-/- deficient and C57BL/6 mice. J Mol Cell Card 2005;39:443.
16.
Olesen M, Kwong E, Mezli A, et al. No effect of cyclooxygenase inhibition on plaque size in atherosclerosis-prone mice. Scand Cardiovasc J 2002;36:362.
17.
Rott D, Zhu J, Burnett MS, et al. Effects of MF-tricyclic, a selective cyclooxygenase 2 inhibitor, on atherosclerosis progression and susceptility to cytomegalovirus replication in apoprotein-E knockout mice. J Am Coll Cardiol 2003;41:1812.
18.
Belton OA, Duffy A, Toomey S. Cyclooxygenase isoforms and platelet wall interations in the apolipoprotein E knockout mouse model of atherosclerosis. Circulation 2003;108:3017.
19.
Lanza-Jacoby S, Miller S, Flynn J, et al. The cyclooxygenase (COX)-2 inhibitor, celecoxib, prevents the development of mammary tumors in HER-2/neu mice. Cancer Epidemiol Biomarkers Prev 2003;12:1486.
20.
Beckstead HJ, Stenberg PE, McEver RP, et al. Immunohistochemical localization of membrane and ␣-granule proteins in human megakaryocytes: Application to plastic embedded bone marrow biopsy specimens. Blood 1986;67:285.
21.
Gross S, Tilly P, Hentsch D, et al. Vascular wall-produced prostaglandin E2 exacerbates arterial thrombosis and atherothrombosis through platelet EP3 receptors. J Exp Med 2007; 204:311.
22.
Grosch S, Tegeder I, Niederberger E, et al. COX-2 independent induction of cell cycle arrest and apoptosis in colon cancer cells by the select COX-2 inhibitor celecoxib. FASEB J 2001;15:2742.
23.
Davies NM, McLachlan AJ, Day RO, et al. Clinical pharmacokinetics and pharmacodynamics of celecoxib: A selective cyclooxygenase-2 inhibitor. Clin Pharmcokinet 2000;38:225.
142 24.
JOURNAL OF SURGICAL RESEARCH: VOL. 146, NO. 1, MAY 1, 2008 Bombardier C, Laine L, Reicin A, et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med 2000;343:120.
25. King VL, Trivedi DB, Gitlin JM, et al. Selective cyclooxygenase-2 inhibiton with celecoxib decreases angiotensin II-induced abdominal aortic aneurysm formation in mice. Arterioscler Thromb Vasc Biol 2006;26:1137. 26.
27.
28.
29.
30.
Burleigh ME, Babaev VR, Patel MB, et al. Inhibition of cyclooxygenase with indomethacin phenethylamide reduces atherosclerosis in apo E-null mice. Biochem Pharmacol 2005;70:334. Bea F, Blessing E, Bennett BJ, et al. Chronic inhibition of cyclooxygenase-2 does not alter plaque composition in a mouse model of advanced unstable atherosclerosis. Cardiovasc Res 2003;60:198. Nakashima Y, Raines EW, Plump As, et al. Up-regulation of VCAM-I and ICAM-1 at Atherosclerosis prone sites on the endothelium in the ApoE- Deficient mouse. Arteriosclerosis Thromb Vasc Biol 1998;18:842. Pei H, Wang Y, Miyoshi T, et al. Direct evidence for a crucial role of the arterial wall in control of atherosclerosis susceptibility. Circulation 2006;114:2382. Tous M, Ferre N, Vilella E, et al. Aspirin attenuates the initiation but not the progression of atherosclerosis in deficient mice
fed a high-fat, high-cholesterol diet. Basic Clin Pharmacol Toxicol 2004;95:15. 31.
Bailey JM, Butler J. Anti-inflammatory drugs in experimental atherosclerosis. I. Relative potencies for inhibiting plaque formation. Atherosclerosis 1973;17:515.
32.
Hajjar DP, Pomerantz KB. Signal transduction in atherosclerosis: Integration of cytokines and the eicosanoid. FASEB J. 1992;6:2933.
33.
Menez EJ, Burtscher HY, Kolarz G. Inhibition of cytokine production and adhesion molecule expression by ibuprofen is without effect on transendothelial migration of monocytes. Inflammation 1999;23:275.
34.
Gonzalez-Alvaro C, Munoz C, Garcia-Vicuna R, et al. Interference of nonsteroidal anti-inflammatory drugs with very late activation antigen 4/vascular cell adhesion molecule-1 mediated lymphocyteendothelial cell adhesion. Arthritis Rheum1998;41:1677.
35.
Sakai A. Diclofenac inhibits endothelial cell adhesion molecule expression induced with lipopolysaccharide. Life Sci 1996;58: 2377.
36.
Miyamoto K, Miyake S, Mizuno M, et al. Selective COX-2 inhibitor celecoxib prevents experimental autoimmune encephalomyelitis through COX-2-independent pathway. Brain 2006; 129:1984.
The Select Cyclooxygenase-2 Inhibitor Celecoxib Reduced the Extent of Atherosclerosis in Apo E-/- Mice Shushan Jacob, M.D.,* Lisa Laury-Kleintop, Ph.D.,† and Susan Lanza-Jacoby, Ph.D.*,1 *Department of Surgery, Thomas Jefferson University, Philadelphia, Pennsylvania and †Lankenau Institute for Medical Research, Wynnewood, Pennsylvania Submitted for publication February 1, 2007
Many investigators have suggested that immune activation may trigger the atherosclerotic process. The benefits of aspirin in preventing myocardial infarction have been attributed, in part, to its anti-inflammatory effects. Several reports have documented that cyclooxygenase (COX)-2 is up-regulated in human and mouse atherosclerotic lesions. To clarify the role of COX-2 in atherosclerosis, we conducted a study to test whether the COX-2 inhibitor, celecoxib, prevents the development and progression of the atherosclerotic process. We have used the apo E-/- mouse, a relevant animal model of atherosclerosis that develops fibrofatty lesions similar to human atherosclerosis. Treatment of 4-wk old apo E-/- mice with a standard rodent no. 5020 diet supplemented with 900 ppm of celecoxib for 16 wk led to an 81% reduction in lesion size. The mean lesion area per section (mean ⴞ SD) of proximal aorta from the apo E-/- mice fed the diet with celecoxib (33,991 ⴞ 7863 m 2, P < 0.001) was significantly less than that of the untreated apo E-/- mice (183,401 ⴞ 36,212 m2). There were no lesions detected in the C57B1/6 mice. Immunohistochemistry of the ileum revealed that there was 80% reduction in staining for intercellular adhesion molecule and 60% reduction in staining for vascular cell adhesion molecule in the celecoxib treated mice. The protective effect of celecoxib was not maintained when the mice were switched after feeding the celecoxib-supplemented diet to the control 5020 diet for an additional 10 wk. These findings demonstrate that selective inhibition of the COX-2 enzyme with celecoxib prevented the development of atherosclerotic lesions in the proximal aortas from apo E-/- mice. One of the possible mechanisms is reduction in expression of the endothelial cell adhesion 1
To whom correspondence and reprint requests should be addressed at Lankenau Institute for Medical Research, 100 Lancaster Ave., Wynnewood, PA 19096. E-mail: Susan.Lanza-Jacoby@jefferson. edu.
cell molecules intercellular adhesion molecule and vascular cell adhesion molecule, which plays a key role in the recruitment of inflammatory cells during the early stages of atherogenesis. © 2008 Elsevier Inc. All rights reserved. Key Words: atherosclerosis; cyclooxigenase-2; adhesion molecules; apo E-/- mice; cyclooxygenase-2 inhibitors. INTRODUCTION
Current data support the hypothesis that atherosclerosis is an inflammatory disease [1, 2]. The beneficial effect of aspirin in reducing the risk of myocardial infarction is, in part, attributable to its antiinflammatory action. Increased inflammation in the plaque is associated with plaque rupture, leading to thrombosis in the coronary vessel [3]. Nonsteroidal anti-inflammatory drugs (NSAIDS) inhibit both isoforms of cyclooxygenase (COX), the enzymes controlling the conversion of arachidonic acid into prostaglandins (PG). COX-1 is the constitutive form of the enzyme involved in normal cell functioning and COX-2 is induced by cytokines, bacterial infections, shear stress, or growth factors. The inhibition of COX-1 by NSAIDS prevents synthesis of protective prostaglandins that synthesize the mucous barrier of the gastric mucosa and blocks the formation of vasodilator, prostacyclin (PGI 2) by kidney cells, leading to the side effects of gastric ulcers and renal toxicity. COX-2 has been found to have particular relevance to inflammation and atherosclerosis in two ways. The first is the suggestion that activation of COX-2 in human monocytes may be able to generate prostaglandin PGF2␣ [4], which is mitogenic, leading to cellular proliferation and vasoconstriction [5] and, thus, may play a role in the genesis of atherosclerosis. The second is that COX-2 with its nuclear loca-
135
0022-4804/08 $34.00 © 2008 Elsevier Inc. All rights reserved.
136
JOURNAL OF SURGICAL RESEARCH: VOL. 146, NO. 1, MAY 1, 2008
tion may produce eicosanoids that are active within the nucleus, and these eicosanoids may modulate transcriptional events [6]. Several reports have documented that COX-2 is upregulated in human and mouse atherosclerosis [7–11]. The studies with COX-2 inhibitors in animal models of atherosclerosis have produced variable results depending on the age when starting treatment, the length of treatment, the strain of mice, and the difference in COX-2 inhibitors [12–18]. The purpose of this study was to evaluate the effect of the COX-2 inhibitor, celecoxib, on early development of atherosclerosis and on the progression of established lesions. We also investigated whether any potential beneficial effects of early intervention would be maintained when celecoxib treatment was removed from the diet. One of the earliest detectable cellular responses in the formation of atherosclerotic lesions is leukocyte adherence to the endothelium by expression of specific leukocyte adhesion molecules expressed by the endothelium. To date there are no studies on the effects of the select COX-2 inhibitors on the adhesion molecules in the vascular tissues. In this study, we investigated whether celecoxib would be effective in preventing the progression of atherosclerosis in the apo E-/- mouse, a transgenic mouse model prone to develop atherosclerosis. We also evaluated whether this protection is associated with decreased expression of the adhesion proteins, intercellular adhesion molecule (ICAM)-1, and mouse vascular cell adhesion molecule (VCAM).
ficed at 26 wk of age. The heart and aortic tree were perfused through the left ventricle with iced phosphate buffered 0.9% saline (PBS) for 10 min. The heart was carefully dissected and removed. The upper half of the heart containing the aortic origin was separated and embedded in Tissue-Tek (Fischer Scientific, Newark, DE) OCT compound contained in cryomolds. These were frozen at ⫺70°F and sequential 10 m thick sections were cut from the base toward the apex of the heart until the aortic valve leaflets appeared. From this point, sequential sections, each 10 microns thick were cut from each mold and stained with 0.5% Oil Red O in propylene glycol for 4 h and counterstained in Mayer’s hematoxylin for 1 min. Quantification of atherosclerosis was performed using a computer assisted image analysis. A video camera mounted to a Nikon upright light microscope (Nikon Instruments, Melville, NY) was used to view the aortic sections. The image was captured and analyzed by NIH Image software. Data were collected from each of the 14 sections taken from the proximal aortic root, averaged, and expressed as m 2 per section.
METHODS
Following perfusion of heart and aortic tree with PBS, the superior mesenteric vein was incised and the ileum was washed free of blood by perfusion with PBS flowing through the aortic tree and out through the superior mesenteric vein. The ileum was then perfused with iced 4% paraformaldehyde in PBS for 20 min. A 3- to 4-cm segment of ileum was isolated from the perfused intestine and fixed in 4% paraformaldehyde for 90 min at 4°C. The ileum was then cut into smaller rings, and the tissue dehydrated in graded acetone washes at 4°C. The ileum rings were embedded in plastic (Immunobed; Polysciences Inc., Warrington, PA), and 4 m thick sections were cut and transferred to Vectabond-coated slides (Vector Laboratories, Burlingame, CA). Immunohistochemical localization of ICAM-1 and VCAM-1 was investigated by using the avidin-biotin immunoperoxidase technique (Vectastain ABC Reagent; Vector Laboratories) according to a previously described method [20]. Tissue sections were treated with 0.25% trypsin to improve reagent penetration. Blocking serum was applied to the tissue sections for 30 min to reduce non-specific binding, and then the tissue sections were incubated for 24 h with specific primary antibodies. ICAM-1 and VCAM-1 were identified with the specific mouse monoclonal IgG2a at a dilution of 1/100. The tissue was then incubated with the biotinylated secondary antibody, and peroxidase staining was carried out using 3,3=-diaminobenzidine. Control preparations consisted of omission of either the primary or the secondary antibody. Expression of adhesion molecules was determined by microscopic observation of the brown peroxidase reaction product on the microvascular endothelium of the tissue sections. Positive staining was defined as a vessel displaying brown reaction product on ⬎50% of the circumference of its endothelium. Fifty ileal venules per tissue section were examined in each of 20 sections, and the percentage of positive staining vessels was tallied.
General Research Design Male apo E-/- and C57B1/6 mice were used in this study. The original breeding pairs of apo E-/- mice were obtained from the Gladstone Institute (San Francisco, CA); breeders for the C57B1/6 mice were purchased from the The Jackson Laboratory (Bar Harbor, ME). Apo E-/- mice were generated on a C57B1/6 background. At 4 wk of age, apo E-/- and C57B1/6 mice were assigned to the following groups: Group-1: apo E-/- fed no. 5020 diet (Purina, St. Louis, MO); Group-2: apo E-/- fed no. 5020 ⫹ 900 ppm celecoxib; and Group-3: C57B1/6 fed no. 5020 diet. The dose of celecoxib is based on previous reports that this amount was well tolerated by the mice when fed for over 1 y [19]. Diets were made weekly in our laboratory and stored at ⫺20°C. Fresh diet was provided every third day. Three mice were placed in one metabolic cage; food intake and weight gain were monitored weekly. At 16 wk of age, 12 mice from each of Groups 1, 2, and 3 were sacrificed. The remaining 10 mice in Group 1 were switched to the no. 5020 ⫹ celecoxib diet and the remaining 10 mice in Group 2 were switched to the no. 5020 diet. Groups 1 and 2 were re-designated as Groups 4 and 5, respectively, and fed an additional 10 wk on the new diet before being sacrificed. At the time of sacrifice, blood was obtained from the retro-orbital plexus for determination of the concentration of cholesterol and prostaglandin E 2 (PGE 2) in the serum.
Analysis of Atherosclerotic Lesions Groups 1, 2, and 3 were sacrificed at 16 wk of age and the newly designated Groups 4 and 5 (previously Groups 1 and 2) were sacri-
Immunohistochemical Analysis on Mouse Aorta Pieces of aorta were fixed in 10% neutral buffered formalin for 24 h, embedded in paraffin, and sectioned. For immunohistochemistry, 10% rabbit serum was added to unstained, frozen sections from the proximal aorta, to block nonspecific activity. This was followed by addition of rabbit polyclonal antibodies to mouse COX-1 (Santa Cruz), mouse COX-2 (Cayman), mouse ICAM-1 and mouse VCAM (Santa Cruz). The sections are then treated with biotinylated rabbit anti-mouse antibody IgG followed by streptavidin-horseradish peroxidase. The peroxidase reaction was developed with diaminobenzidine and counterstained with hematoxylin. The sections were viewed under the Nikon light microscope and sections from the different groups were compared for difference in intensity of the dark brown staining between antigen and antibody. The images were analyzed using NIH Image software.
Immunohistochemical Analysis on Mouse Ileum
JACOB, LAURY-KLEINTOP, AND LANZA-JACOBY: THE SELECT COX-2 INHIBITOR CELECOXIB
137
FIG. 1. (A) Representative photomicrograph of Oil-Red-O stained aortic root from Apo E-/- mice treated with celecoxib. Mice were fed the no. 5020 diet supplemented with 900 ppm celecoxib for 16 wk. Sections are shown at ⫻40 magnification areas of lipid accumulation. (B) The extent of atherosclerotic lesions was quantified with Oil-Red-O staining and analyzed by NIH Image software. Data were collected from each of the 14 sections taken from the proximal aortic root, averaged, and expressed as m 2 per section. Bars indicate means ⫾ SD for n ⫽ 12, *P ⬍ 0.001 by Student’s t-test. (Color version of figure is available online.)
Determination of PGE 2
RESULTS
Enzyme-linked immunosorbent assay kits (R and D Systems, Minneapolis, MN) were used for determination of PGE 2 concentration in the serum. Samples were diluted 1:2 in assay buffer. The assay was performed in accordance with the manufacturer’s protocols in a microplate reader at a wavelength of 405 nm. Limits of detection for these assays were ⬍15.9 pg/mL (PGE 2). Samples were assayed in duplicate.
We evaluated whether celecoxib treatment early in life would prevent the development of aortic atherosclerosis in the apo E-/- mice. Providing celecoxib in the diet at a dose of 126 mg/kg beginning at 4 wk of age prevented the development of atherosclerosis in apo E-/- mice. The mean area of aortic lesions from the apo E-/- mice fed the no. 5020 ⫹ celecoxib diet (33,991 m 2 ⫾ 7863) was 81% (P ⬍ 0.001) less than the apo E-/- mice fed no. 5020 diet (183,401 m 2 ⫾ 36,212) (Fig. 1A and B). The protective effect of celecoxib was not associated with a reduction of serum cholesterol levels in both apo E-/- and C57B1/6 mice (data not included).
Statistical Analysis Results were expressed as mean ⫾ SD. The extent of atherosclerosis was analyzed by analysis of variance and subsequently by Student’s unpaired two-tailed t-test.
138
JOURNAL OF SURGICAL RESEARCH: VOL. 146, NO. 1, MAY 1, 2008
FIG. 2. Celecoxib does not alter the progression of atherosclerosis in apo E-/- mice. Mean lesion size in the proximal aorta was quantified after 16 wk of feeding apo E-/- mice with the no. 5020 diet. Then a subgroup of these mice were switched to the no. 5020 diet supplemented with 900 ppm celecoxib for an additional 10 wk and mean lesion size was quantified. Bars indicate means ⫾ SD for n ⫽ 12.
We also questioned whether inhibition of COX-2 by celecoxib would alter the progression of atherosclerosis. After feeding the no. 5020 diet for 16 wk some mice in Group 1 were switched to a diet supplemented with celecoxib. Figure 2 shows that lesion size continued to progress when celecoxib was started after feeding the control diet for 16 wk. The lesion size increased 52% from 183, 401 ⫾ 36,212 m 2 at the 16th wk of feeding the control diet, to 279,624 ⫾ 43,599 m 2 after an additional 10 wk of feeding the celecoxib diet. To determine whether continuous exposure to celecoxib was required to maintain the protective effects, celecoxib-treated apo E-/- mice were switched at the end of the 16 wk feeding period to the control diet for an additional 10 wk. Lesion size increased from 33,992 m 2 ⫾ 7863 at 16 wk to 235,965 m 2 ⫾ 1434 at 26 wk (Fig. 3). Immunohistochemical analysis demonstrated that the aorta of the apo E-/- mice stained positive for COX-1 and COX-2. Staining for both COX-1 and COX-2 antigen appear to be proportional to the lesion size with less staining in the apo E-/- mice fed celecoxib (Fig. 4A). Staining for ICAM and VCAM also appeared to be less in aortas from the apo E-/- mice fed celecoxib (Fig. 4B). To determine whether the reduction in ICAM and VCAM was due to a direct effect of the celecoxib treatment or the decrease in plaque size, we analyzed the protein expression in the ileum. Twenty sections were studied in each mouse and 50 vessels were studied in each section. The expression of adhesion molecules was determined by microscopic observation of the brown peroxidase reaction product on the microvascular endothelium of the tissue sections. Positive staining was defined as a vessel displaying a brown reaction product
on ⬎50% of the circumference of its endothelium. Immunohistochemical analysis showed an 80% reduction in staining for ICAM and 60% reduction in staining for VCAM in apo E-/- mice fed celecoxib compared with the mice fed the control diets (Fig. 5A and B). Since COX-2 is increased in plaque [7–11] and COX-2derived PGE 2 may facilitate arterial thrombosis [21], we evaluated the PGE2 concentration in the serum. The PGE 2 concentration of the serum from the celecoxib-fed mice (4.4 pg/mL) was significantly reduced in comparison to the control apo E-/- mice (12.7 pg/mL), (n ⫽ 6), P ⬍ 0.01).
FIG. 3. Continuous treatment with celecoxib is required to maintain protective effects. Mean lesion size in the proximal aorta was quantified after 16 wk of feeding apo E-/- mice with the no. 5020 diet supplemented with 900 ppm celecoxib. Then a subgroup of these mice were switched to the no. 5020 diet for an additional 10 wk and mean lesion size was quantified. Bars indicate means ⫾ SD for n ⫽ 12.
JACOB, LAURY-KLEINTOP, AND LANZA-JACOBY: THE SELECT COX-2 INHIBITOR CELECOXIB
139
FIG. 4. Immunohistochemistry in sections from proximal aorta from Apo E-/- mice treated with celecoxib. After 16 wk of treatment with the control no. 5020 and no. 5020 diet supplemented with 900 ppm celecoxib, mice were sacrificed. (A) COX-1 and COX-2 were detected by rabbit polyclonal antibodies to mouse COX-1 and COX-2 in proximal aorta from apo E-/- mice. (B) 0 VCAM and ICAM were detected by specific mouse monoclonal IgG2a. (Color version of figure is available online.)
DISCUSSION AND CONCLUSIONS
There is increased evidence to suggest that COX-2 is involved in the early development of atherosclerosis. Normal blood vessels contain COX-1 but not COX-2 [3]. In contrast, atherosclerotic lesions from humans [8 –12] and animals [12–16, 18] have high expression levels of COX-2. Recent studies show that deletion of
COX-2 in both the low-density lipoprotein receptor (LDLR)-/- mice and the apo E-/- mice significantly reduced lesion area of the proximal aorta [14, 15]. These studies also demonstrate that beginning treatment at 8 to 10 wk of age with the COX-2 inhibitors rofecoxib or NS-398 led to a significant reduction in the extent of atherosclerosis in LDLR-/- and apo E-/- mice [14, 15]. Here, we show that celecoxib treatment begun at 4 wk
140
JOURNAL OF SURGICAL RESEARCH: VOL. 146, NO. 1, MAY 1, 2008
FIG. 5. Celecoxib decreases immunohistochemical staining for ICAM and VCAM in ileal microvessels from apo E-/- mice. Percentage of vessels staining positive for ICAM and VCAM are shown for control, untreated C57Bl/6 mice, apo E-/- mice fed the no. 5020 diet, and apo E-/- mice fed the no. 5020 diet supplemented with 900 ppm celecoxib. Bar heights represent average values for three mice per group. Twenty sections were studied in each mouse and 50 vessels were evaluated in each section.
of age also prevents atherosclerotic lesion development in apo E-/- mice and that the protective effect was associated with a reduction in ICAM and VCAM protein expression. We further demonstrate that the protective effect of celecoxib is not maintained when treatment is stopped. In contrast to the protective effects of rofecoxib and celecoxib, other studies have found that the select COX-2 inhibitors, SC236 and MF Tricyclic had no effect on lesion development in apo E-/- mice when treatment was initiated at 6 and 8 wk of age, respectively [16, 17]. These studies suggest that differences in the selectivity of the COX-2 inhibitors may explain the contrasting results. In a previous study, we showed that the mean serum concentration of celecoxib was 5.2 ⫾ 0.6 M when mice were fed a similar dose of celecoxib [19], which agrees with previous reports [22]. The IC 50 for COX-1 and COX-2 inhibition was reported to be 6.7 ⫾ 0.09 M and 0.87 ⫾ 0.18 M, respectively [23]. Based upon these IC50 concentrations, the results of this study appear to be due to selective COX-2 not COX-1 inhibition. While COX-2 may have significance in the early development, its role in advanced atherosclerotic lesions is less clear. Recent clinical trials have found a 5-fold increase in atherothrombotic events after 18 months of rofecoxib use [24], which resulted in the withdrawal of rofecoxib from the market. Previous laboratories have suggested that the increase in thrombotic events associated with the use of COX-2 inhibitors may be related to the ability of COX- 2 inhibitors to block PGI 2 without altering thromboxane levels [11, 13]. PGI 2 suppresses platelet activity and increases bleeding while thromboxane increases platelet activity. However, the role of COX-2 inhibitors in plaque rupture may be more complex; a more recent study demonstrated that celecoxib
treatment reduced incidence of abdominal aortic aneurysms in angiotensin II-induced mouse model of aortic aneurysms [25]. COX-2 inhibitors appear to have no beneficial effects in reversing or altering the progression of atherosclerosis in animals with established lesions. A recent study conducted with a new nonselective COX-2 inhibitor INDO-PA was shown to reduce aortic PGE 2 levels and early atherosclerosis but had less of an effect on the progression of established lesions [26]. Another study found that there was no change in lesion area after feeding apo E-/- mice for 15 wk at dose of 75 mg/kg/d [27]. This study also showed that there was no effect of celecoxib on the composition or stability of advanced lesions indicating that celecoxib administration may not alter inflammatory mediators in established lesions. We also observed that there was no change in lesion size when a higher dose of celecoxib (126 mg/kg body weight/d) was fed to 20 wk-old apo E-/mice. These observations are consistent with the suggestion that COX-2 inhibition blocks early events essential to atherosclerosis. One of the earliest detectable cellular responses in the formation of atherosclerotic lesions is leukocyte adherence to the endothelium. Leukocyte recruitment occurs in lesion-prone areas of the arterial tree by expression of specific adhesion molecules, which persists as long as the condition of hypercholesterolemia continues [28]. VCAM-1, together with P- and E-selectin, is involved in the first step of tethering and rolling of monocytes and lymphocytes, and in the second stage mediate arrest and firm adhesion. ICAM-1 helps mediate arrest and firm adhesion of lymphocytes, monocytes, and neutrophils. Increased expression of VCAM has been shown to be associated with atherosclerotic lesion development [28, 29]. Several studies have demonstrated that NSAIDS
JACOB, LAURY-KLEINTOP, AND LANZA-JACOBY: THE SELECT COX-2 INHIBITOR CELECOXIB
reduce atherosclerosis in animal models [13, 30, 31] and also reduce lymphocyte-endothelial cell adhesion molecules [32, 33]. Ibuprofen has been shown to inhibit VCAM-1 [30] and diclofenac inhibited lipopolysaccharide-induced up-regulation of VCAM-1, ICAM-1, and E-selectin in endothelial cells [34]. Diclofenac, indomethacin, and aceclofenac also reduced adhesion of peripheral blood lymphocytes to human umbilical vein endothelial cells by decreasing expression of VCAM-1 [35]. A recent study showed that celecoxib reduced protein expression of ICAM and P-selectin in mouse brain tissue [36]. Our study is the first report that a specific COX-2 inhibitor, celecoxib, reduced protein expression of VCAM and ICAM in the vascular endothelium from the atherosclerosis prone apo E-/- mice. These findings support the role of inflammation in early lesion development. The increase in adhesion molecule expression in the lesion-prone areas represents an inflammatory response and their reduction may be one mechanism by which celecoxib prevents plaque formation. While this study further supports the premise that blocking COX-2 before the onset vessel injury will prevent plaque development, our data further demonstrate that inhibition of COX-2 must be continuous to maintain the arrest in lesion development. Once celecoxib treatment is discontinued, lesions continued to increase in size comparable to control apo E-/- mice. The time of initiating celecoxib treatment also is an important factor in arresting plaque development, as celecoxib is unable to arrest plaque growth in mice with established lesions. Our observation that celecoxib will reduce expression of ICAM and VCAM suggests that in apo E-/- atherosclerosis model, celecoxib retards the development of plaque and that this effect is paralleled by a suppression of the inflammatory factors, ICAM, and VCAM. ACKNOWLEDGMENTS The authors thank Chris Munery for her technical assistance. This study was supported in part by a grant from the American Heart Association to SL-J.
REFERENCES 1.
Ross R. The pathogenesis of atherosclerosis: Perspective for the 1990s. Nature 1993;362:801.
2.
Alexander RW. Inflammation and coronary artery disease. N Engl J Med 1994;331:4687.
3.
Mitchell JA, Warner TD. COX isoforms in the cardiovascular system: Understanding the activities of the non-steroidal antiinflammatory drugs. Nat Rev Drug Discov 2006;5:75.
4.
van der Wal AC, Becker AE, van der Loos CM, et al. Site of intimal rupture or erosion of thrombosed coronary atherosclerotic plaques is characterized by an inflammatory process irrespective of the dominant plaque morphology. Circulation 1994; 89:36.
141
5.
Pratico D, FitzGerald GA. Generation of 8-epiprostaglandin F2alpha by human monocytes. Discriminate production by reactive oxygen species and prostaglandin endoperoxide synthase-2. J Biol Chem 1996;271:8919.
6.
Fukunaga M, Makita N, Roberts LJ, et al. Evidence for the existence of F2-isoprostane receptors on rat vascular smooth muscle cells. Am J Physiol 1993;264:C1619.
7.
Serhan CN. Signalling the fat controller. Nature 1996;384:23.
8.
Baker CSR, Hall RJC, EvansTJ, et al. Cyclooxygenase-2 is widely expressed in atherosclerotic lesions affecting native and transplanted human coronary arteries and co-localizes with inducible nitric oxide synthase and nitrotyrosine particularly in macrophages. Arterioscler Thromb Vasc Biol 1999;19:646.
9.
Schonbeck U, Sukhova GK, Graber P, et al. Augmented expression of cyclooxygenase-2 in human atherosclerotic lesions. Am J Pathol 1999;155:1281.
10.
Hong BK, Kwon HM, Lee BK, et al. Coexpression of cyclooxygenase-2 and matrix metalloproteinases in human aortic atherosclerotic lesions. Yonsei Med J 2000;41:82.
11.
Belton O, Byrne D, Kearney D, et al. Cyclooxygenase-1 and -2 dependent prostacyclin formation in patients with atherosclerosis. Circulation 2000;102:840.
12.
Stemme V, Swedenborg J, Claesson H, et al. Expression of cyclooxygenase-2 in human atherosclerotic carotid arteries. Eur J Vasc Endovasc Surg 2000;20:146.
13.
Pratico D, Tillmann C, Zhang ZB, et al. Acceleration of atherogenesis by COX-1-dependent prostanoid formation in low density lipoprotein receptor knockout mice. Proc Natl Acad Sci USA 2001;98:3358.
14.
Burleigh ME, Babaev VR, Oates JA, et al. Cyclooxygenase-2 promotes early atherosclerosis lesion formation in LDL receptor-deficient mice. Circulation 2002;105:1816.
15.
Burleigh ME, Babaev VR, Yancey PG, et al. Cyclooxygenase-2 promotes early atherosclerosis lesion formation in Apo E-/- deficient and C57BL/6 mice. J Mol Cell Card 2005;39:443.
16.
Olesen M, Kwong E, Mezli A, et al. No effect of cyclooxygenase inhibition on plaque size in atherosclerosis-prone mice. Scand Cardiovasc J 2002;36:362.
17.
Rott D, Zhu J, Burnett MS, et al. Effects of MF-tricyclic, a selective cyclooxygenase 2 inhibitor, on atherosclerosis progression and susceptility to cytomegalovirus replication in apoprotein-E knockout mice. J Am Coll Cardiol 2003;41:1812.
18.
Belton OA, Duffy A, Toomey S. Cyclooxygenase isoforms and platelet wall interations in the apolipoprotein E knockout mouse model of atherosclerosis. Circulation 2003;108:3017.
19.
Lanza-Jacoby S, Miller S, Flynn J, et al. The cyclooxygenase (COX)-2 inhibitor, celecoxib, prevents the development of mammary tumors in HER-2/neu mice. Cancer Epidemiol Biomarkers Prev 2003;12:1486.
20.
Beckstead HJ, Stenberg PE, McEver RP, et al. Immunohistochemical localization of membrane and ␣-granule proteins in human megakaryocytes: Application to plastic embedded bone marrow biopsy specimens. Blood 1986;67:285.
21.
Gross S, Tilly P, Hentsch D, et al. Vascular wall-produced prostaglandin E2 exacerbates arterial thrombosis and atherothrombosis through platelet EP3 receptors. J Exp Med 2007; 204:311.
22.
Grosch S, Tegeder I, Niederberger E, et al. COX-2 independent induction of cell cycle arrest and apoptosis in colon cancer cells by the select COX-2 inhibitor celecoxib. FASEB J 2001;15:2742.
23.
Davies NM, McLachlan AJ, Day RO, et al. Clinical pharmacokinetics and pharmacodynamics of celecoxib: A selective cyclooxygenase-2 inhibitor. Clin Pharmcokinet 2000;38:225.
142 24.
JOURNAL OF SURGICAL RESEARCH: VOL. 146, NO. 1, MAY 1, 2008 Bombardier C, Laine L, Reicin A, et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med 2000;343:120.
25. King VL, Trivedi DB, Gitlin JM, et al. Selective cyclooxygenase-2 inhibiton with celecoxib decreases angiotensin II-induced abdominal aortic aneurysm formation in mice. Arterioscler Thromb Vasc Biol 2006;26:1137. 26.
27.
28.
29.
30.
Burleigh ME, Babaev VR, Patel MB, et al. Inhibition of cyclooxygenase with indomethacin phenethylamide reduces atherosclerosis in apo E-null mice. Biochem Pharmacol 2005;70:334. Bea F, Blessing E, Bennett BJ, et al. Chronic inhibition of cyclooxygenase-2 does not alter plaque composition in a mouse model of advanced unstable atherosclerosis. Cardiovasc Res 2003;60:198. Nakashima Y, Raines EW, Plump As, et al. Up-regulation of VCAM-I and ICAM-1 at Atherosclerosis prone sites on the endothelium in the ApoE- Deficient mouse. Arteriosclerosis Thromb Vasc Biol 1998;18:842. Pei H, Wang Y, Miyoshi T, et al. Direct evidence for a crucial role of the arterial wall in control of atherosclerosis susceptibility. Circulation 2006;114:2382. Tous M, Ferre N, Vilella E, et al. Aspirin attenuates the initiation but not the progression of atherosclerosis in deficient mice
fed a high-fat, high-cholesterol diet. Basic Clin Pharmacol Toxicol 2004;95:15. 31.
Bailey JM, Butler J. Anti-inflammatory drugs in experimental atherosclerosis. I. Relative potencies for inhibiting plaque formation. Atherosclerosis 1973;17:515.
32.
Hajjar DP, Pomerantz KB. Signal transduction in atherosclerosis: Integration of cytokines and the eicosanoid. FASEB J. 1992;6:2933.
33.
Menez EJ, Burtscher HY, Kolarz G. Inhibition of cytokine production and adhesion molecule expression by ibuprofen is without effect on transendothelial migration of monocytes. Inflammation 1999;23:275.
34.
Gonzalez-Alvaro C, Munoz C, Garcia-Vicuna R, et al. Interference of nonsteroidal anti-inflammatory drugs with very late activation antigen 4/vascular cell adhesion molecule-1 mediated lymphocyteendothelial cell adhesion. Arthritis Rheum1998;41:1677.
35.
Sakai A. Diclofenac inhibits endothelial cell adhesion molecule expression induced with lipopolysaccharide. Life Sci 1996;58: 2377.
36.
Miyamoto K, Miyake S, Mizuno M, et al. Selective COX-2 inhibitor celecoxib prevents experimental autoimmune encephalomyelitis through COX-2-independent pathway. Brain 2006; 129:1984.