Evidence that 5-lipoxygenase and acetylated cyclooxygenase 2-derived eicosanoids regulate leukocyte-endothelial adherence in response to aspirin

Abstract

(1) Unlike other nonsteroidal anti-inflammatory drugs that inhibit formation of cyclooxygenase (COX)-dependent eicosanoids, acetylation of COX-2 by aspirin switches eicosanoid biosynthesis from prostaglandin E(2) (PGE(2)) to 15-epi-lipoxin A(4) (15-epi-LXA(4) or aspirin-triggered lipoxin, ATL). ATL formation by activated leukocytes (PMN) requires the intervention of 5-lipoxygenase (5-LOX), an enzyme that is involved in leukotriene B(4) (LTB(4)) formation. (2) In the present study, we have examined the role of acetylated COX-2 and 5-LOX in modulating antiadhesive effects of aspirin on adhesion of PMN to endotoxin (LPS)-primed human umbilical endothelial cells (HUVEC). (3) Treating PMN/HUVEC cocultures with aspirin resulted in a concentration-dependent inhibition of cell-to-cell adhesion induced by LPS. Treating HUVEC with selective COX-2 inhibitors, celecoxib and rofecoxib, caused an approximately 70% reversion of antiadhesive effect of aspirin. In contrast, inhibition of neutrophil's 5-LOX pathway with 1 micro M ZD2138, a selective 5-LOX inhibitor, 1 micro M BAY-X-1005, a FLAP inhibitor, or 100 micro M licofelone, a dual COX/5-LOX inhibitor, did not affect antiadhesive properties of aspirin. (4) Exposure to celecoxib (100 micro M) or rofecoxib (10 micro M) completely suppressed ATL formation caused by aspirin without affecting LTB(4) levels. ZD2138, licofelone and BAY-X-1005 inhibited ATL formation as well as LTB(4) generation. (5) Treatment with LXA(4) reduced PMN adhesion to HUVEC and counteracted the proadhesive effect of celecoxib. In contrast, exposure to Boc-1, an LXA(4) antagonist, counteracts the antiadhesive activities of aspirin. Exposure to U75302, an LTB(4) receptor antagonist, enhances the antiadesive effect of aspirin. (6) Reversal of antiadhesive activities of aspirin by celecoxib was associated with increased expression of LFA-1 on PMN and E-selectin on HUVEC. Addition of LXA(4), ZD2138 and U75302 inhibited these changes. (7) The present results support the notion that inhibition of ATL formation is mechanistically linked to the reversal of the antiadhesive activity of aspirin caused by selective COX-1 inhibitors and suggests that the LTB(4)/ATL balance modulates pro- and antiadhesive activity of nonsteroidal anti-inflammatory drugs at the leukocyte-endothelial cell interface.

Figure 1

Panel (a) Aspirin causes a concentration-dependent inhibition of cell to cell adhesion in PMN/HUVEC cocultures. Data are mean±s.e. of six experiments. Asterisk denotes P<0.001 versus LPS alone. Panel (b) COX-2-derived eicosanoids are required for antiadhesive effects of aspirin. Exposure of PMN/HUVEC cocultures to 100 μM celecoxib or 10 μM rofecoxib but not to the dual COX-5-LOX inhibitor, licofelone, abrogates the antiadhesive activities of aspirin. Data are mean±s.e. of six experiments. ^*P<0.001 versus control; ^**P<0.0001 versus LPS alone. Panel (c) Celecoxib, but not licofelone, causes a concentration-dependent reversal of antiadhesive properties of aspirin in PMN/HUVEC cocultures. Data are mean±s.e. of six experiments. ^*P<0.01 versus aspirin alone. Panel (d) Rofecoxib causes a concentration-dependent reversal of antiadhesive properties of aspirin in PMN/HUVEC cocultures. Data are mean±s.e.of six experiments. ^*P<0.01 versus aspirin alone.

Figure 2

Effect of COX and 5-LOX inhibitors on eicosanoid formation. Data are mean±s.e. of six experiments. P<0.001 versus control; ^*P<0.01 versus cells incubated with medium alone; ^**P<0.01 versus cells incubated with LPS, ^ψP<0.01 versus cells incubated with LPS plus aspirin.

Figure 3

Panel (a) LXA₄ reverses the antiadhesive activities of coxibs. LPS-treated PMN and HUVEC were treated with 10 ng ml⁻¹ LXA in the presence of 100 μM aspirin alone or aspirin plus celecoxib or rofecoxib and adhesion assessed. Data are mean±s.e.of six to eight experiments. ^*P<0.01 versus control; ^**P<0.001 versus LPS plus aspirin; ^ψP<0.05 versus LPS plus aspirin and celecoxib or rofecoxib. Panel (b) LXA causes a concentration-dependent reversion of proadhesive effects of celecoxib and rofecoxib. Data are mean±s.e. of six to eight experiments. ^*P<0.05 versus cells incubated without LXA₄. Panel (c) Boc-1 antagonizes the antiadhesive activity of LXA₄. Data are mean±s.e. of six experiments. ^*P<0.01 versus control; ^** P<0.01 versus LPS plus LXA₄. Panel (d) Boc-1, a selective LXA₄ receptor antagonist, reverses antiadhesive activities of aspirin. Data are mean±s.e. of six experiments. ^*P<0.01 versus control. ^**P<0.001 versus LPS alone; ^φ,ψ P<0.01 versus aspirin alone.

Figure 4

LTB4 reverts antiadhesive activities of aspirin. Data are mean±s.e. of six to eight experiments.^*P<0.01 versus cells incubated with medium alone; ^**P<0.01 versus cells incubated with LPS; ^ψP<0.01 versus cells incubated with LPS plus celecoxib in combination with aspirin, ZD2138, BAY-X-1005 or U75302.

Figure 5

Effect of aspirin and celecoxib in combination with ZD2138, BAY-X-1005 and U75302 on eicosanoid generation by LPS-primed PMN-HUVEC cocultures. Data are mean±s.e. of six to eight experiments. ^*P<0.01 versus LPS alone. Panel (c)–(e). The antiadhesive effect exerted by ZD2138, BAY-X-1005 and U75302 is concentration-dependent. ^*P<0.01 versus cells incubated with LPS plus celecoxib.

Figure 6

LXA₄ and 5-LOX inhibition reverses upregulation of adhesion molecules caused by celecoxib on PMN (panel a) and HUVEC (panel b). Data are mean±s.e. of six to eight experiments. ^*P<0.01 versus control; ^**P<0.05 versus LPS+aspirin; ^ψP<0.01 versus LPS plus aspirin; ^ψψP<0.05 celecoxib versus LPS plus aspirin. Panels (c) and (d). Induction of E-selectin expression on HUVEC caused by celecoxib is concentration dependent. Data are mean±s.e. of six to eight experiments. ^*P<0.01 versus untreated HUVEC; ^**P<0.01 versus aspirin; ^ψP<0.01 versus aspirin plus celecoxib. Panel (d) Licofelone causes a concentration-dependent inhibition of E-selectin expression on HUVEC. Data are mean±s.e. of six to eight experiments. ^*P<0.01 versus HUVEC incubated with LPS plus aspirin.

Figure 7

Schematic drawing of the pathways activated by aspirin in LPS-primed PMN/HUVEC cocultures. (Inset) Exposure to LPS stimulates 5-LOX and COX-2 to generate proinflammatory (PGE₂) and proadhesive (LTB₄) eicosanoids. In the presence of aspirin, COX-2 is acetylated and eicosanoid metabolism is switched from PGE₂ to ATL and LTB₄. Inhibition of LTB₄ with celecoxib causes the reversal of antiadhesive activities of aspirin. Inhibition of 5-LOX with licofelone or ZD2138 enhances antiadhesive activity of aspirin. Addition of LXA₄ restores the antiadhesive activity of aspirin in the presence of celecoxib.