Cyclooxygenase-2 inhibition increases lipopolysaccharide-induced atherosclerosis in mice

Abstract

Aims: The risk of adverse cardiovascular events in humans is increased with chronic use of cyclooxygenase-2 (COX-2) inhibitors. However, the role of COX-2 in animal models of cardiovascular disease has been controversial. In humans and animal models, cardiovascular disease is increased by bacterial infection of the supporting tissue of the teeth, a condition known as periodontal disease. Periodontal disease may result in chronic exposure to pro-inflammatory mediators, such as bacterial lipopolysaccharide (LPS), thereby producing a systemic inflammatory response. The current study examined the role of COX-2 in atherosclerosis induced by LPS derived from the periodontal disease pathogen Porphyromonas gingivalis (P. gingivalis).

Methods and results: Porphyromonas gingivalis LPS was administered by chronic infusion for 28 days and atherosclerosis development was examined in the aortic root of ApoE (apolipoprotein E)-deficient mice. The extent of atherosclerosis was compared between mice receiving control diet or diet containing the COX-2 inhibitor celecoxib. The role of COX-2 in P. gingivalis LPS-induced inflammatory cell activation was examined in peritoneal macrophages. Porphyromonas gingivalis LPS infusion significantly increased atherosclerosis development. In mice infused with P. gingivalis LPS, administration of the COX-2 inhibitor celecoxib further increased the extent of atherosclerotic lesion area. In peritoneal macrophages, P. gingivalis LPS increased the expression of COX-2 mRNA (messenger ribonucleic acid) and the production of prostaglandin (PG) E(2) (PGE(2)), the latter of which was inhibited by celecoxib. Porphyromonas gingivalis LPS-induced expression of tumour necrosis factor alpha (TNFalpha) was enhanced by inactivation of COX-2 and was attenuated by treatment with PGE(2).

Conclusion: The inhibition of COX-2-derived PGE(2) may enhance P. gingivalis LPS-induced atherosclerosis by increasing macrophage production of TNFalpha.

Figure 1

Chronic infusion of Porphyromonas gingivalis (P. gingivalis) lipopolysaccharide (LPS) increases atherosclerosis in ApoE (apolipoprotein E)-deficient mice. Quantification of aortic root lesion area stained positive for oil red O in control saline-infused mice (clear bar) and mice infused with P. gingivalis LPS. The mean ± SEM is shown for each group. n = 10–12 mice/group, ***P < 0.0001, unpaired Student t-test.

Figure 2

COX-2 (cyclooxygenase-2) inhibition increases Porphyromonas gingivalis (P. gingivalis) lipopolysaccharide (LPS)-induced atherosclerosis development in apolipoprotein E (ApoE)-deficient mice. Representative lesion development in the aortic root of P. gingivalis LPS-infused ApoE-deficient mice following treatment with (A) control diet or (B) celecoxib containing diet (scale bar 100 µm). (C) Quantification of aortic root lesion area stained positive for oil red O in Control (clear bar) and celecoxib-treated (grey bar) mice infused with P. gingivalis LPS. Data are reported as lesion area determined from oil red O staining. (D) Nanograms of PGE₂ (prostaglandin E₂) per mL of plasma detected in blood from mice treated with control diet or diet containing celecoxib. The mean ± SEM is shown for each group. n = 8–10 mice/group, *P < 0.05, ***P < 0.0001, unpaired Student t-test.

Figure 3

Cyclooxygenase-2 (COX-2) expression co-localizes with macrophages in Porphyromonas gingivalis (P. gingivalis) lipopolysaccharide (LPS)-induced atherosclerotic lesions. (A) Oil red O staining of lesion in the aortic root of ApoE (apolipoprotein E)-deficient mice following P. gingivalis LPS infusion. Immunohistochemical analysis of (B) COX-2 or (C) macrophage marker (Mac-3) expression in the aortic root (scale bar 100 µM). Dark brown staining (DAB reagent) indicates (B) COX-2 or (C) Mac-3 expression and sections are counterstained with haematoxylin (blue).

Figure 4

Porphyromonas gingivalis (P. gingivalis) lipopolysaccharide (LPS) induces peritoneal macrophage expression of cyclooxygenase-2 (COX-2) messenger ribonucleic acid (mRNA) and production of prostaglandin E₂ (PGE₂). (A) The expression of COX-2 mRNA in peritoneal macrophages following vehicle (Control) or P. gingivalis LPS (LPS) treatment for 6 h. COX-2 mRNA was normalized to HPRT (hypoxanthine phospho-ribosyl transferase) mRNA levels. (B) PGE₂ in macrophages incubated with media containing vehicle (Control), P. gingivalis LPS (LPS) or P. gingivalis LPS and 1 µM celecoxib (LPS + Celecoxib). Data are reported as mean ± SEM. n = 6–8 samples/group, **P < 0.01, unpaired Student t-test.

Figure 5

Inactivation of cyclooxygenase-2 (COX-2) enhances Porphyromonas gingivalis (P. gingivalis) lipopolysaccharide (LPS)-induced tumour necrosis factor alpha (TNFα) expression. (A) The expression of TNFα mRNA (messenger ribonucleic acid) following 6 h of P. gingivalis LPS stimulation was compared between peritoneal macrophages isolated from COX-2 wild-type (clear bars) and COX-2-deficient (grey bars) mice. For the LPS + PGE₂ samples, prostaglandin E₂ (PGE₂) was added to the culture media for the 6 h incubation with P. gingivalis LPS. (B) TNFα protein expression following 24 h of vehicle (Control) or P. gingivalis LPS (LPS) stimulation was compared between COX-2 wild-type (clear bars) and COX-2-deficient (grey bars) peritoneal macrophages. For LPS + PGE₂, PGE₂ was added to the culture media during the 24 h incubation with P. gingivalis LPS. (C) TNFα protein expression following treatment for 24 h with vehicle (Control), P. gingivalis LPS (LPS), or P. gingivalis LPS plus TLR2 blocking antibody (LPS+TLR2 Ab). (D) TNFα protein expression following treatment for 24 h with vehicle (Control), P. gingivalis LPS (LPS), P. gingivalis LPS plus TLR4 blocking antibody (LPS + TLR4 Ab), or non-specific immunoglobulin G (IgG) antibody (LPS + IgG Ab). Data are reported as mean ± SEM. n = 6–8 samples/group, significantly different from Control, ***P < 0.0001, **P < 0.01, unpaired Student t-test.