1 Increased levels of several pro-in¯ammatory cytokines including interleukin-1b (IL-1b) and tumour necrosis factor-a (TNFa) have been found in bronchoalveolar lavage¯uid from symptomatic asthmatic patients. are known to stimulate a number of cells to produce in¯ammatory mediators such as prostaglandins. Although airway smooth muscle (ASM) is known to be a rich source of prostaglandins, the regulation of cyclo-oxygenase (COX) isoforms and prostanoid production by proin¯ammatory cytokines have not been studied in human airway smooth muscle. 2 We studied the e ects of IL-1b, TNFa and IFNg on the induction of two isoforms of cyclooxygenase and its relation to prostaglandin E 2 (PGE 2 ) release and COX activity (re¯ected by PGE 2 synthesis from exogenous arachidonic acid) in human cultured airway smooth muscle cells. 3 IL-1b, but not TNFa or IFNg, caused a time-and concentration-dependent enhancement in PGE 2 and other prostanoid (6-keto-PGF 1a , PGF 2a , thromboxane B 2 (TXB 2 ) and PGD 2 ) production, with PGE 2 and 6-keto-PGF 1a as the principal products. This stimulation was accompanied by a corresponding increase in COX activity. 4 COX-2 protein measured by Western blot analysis was not detectable in untreated cells, but was increased in a time-and concentration-dependent manner by IL-1b, but not TNFa or IFNg. In contrast, no variation in the expression of COX-1 protein was observed. 5 Pretreatment with the conventional non-steroidal anti-in¯ammatory drugs (NSAIDs), indomethacin and ibuprofen, and the selective COX-2 inhibitors, NS-398 and nimesulide, completely blocked IL-1b-induced PGE 2 release and COX activity. The glucocorticosteroid dexamethasone and protein synthesis inhibitors, cycloheximide and actinomycin D, not only markedly inhibited IL-1b-stimulated PGE 2 release and COX activity but also suppressed IL-1b-induced COX-2 induction. 6 This study demonstrates that human cultured ASM cells release prostanoids in response to IL-1b stimulation and that the response is mostly mediated by the induction of COX-2 rather than COX-1 isoenzyme, implying that airway smooth muscle may be an important source of prostaglandins in human airways and that COX-2 may play an important role in the regulation of the in¯ammatory process in asthma.

We have previously reported that human airway smooth-muscle (ASM) cells produce abundant interleukin (IL)-8, a major neutrophil chemoattractant involved in asthma exacerbations. Here, we tested the effects of the beta(2)-agonists salbutamol (Salbu) and salmeterol (Salme) on IL-8 release and tumor necrosis factor (TNF)-alpha-induced IL-8 release from ASM cells. We found that TNF-alpha strongly enhanced IL-8 release in a time- and concentration-dependent manner, whereas Salbu, Salme, the direct adenylyl cyclase activator forskolin (FSK), and the cyclic monophosphate (cAMP) analogue 8-bromoadenosine 3',5'-cAMP (8-Br-cAMP) alone weakly stimulated IL-8 release. TNF-alpha (10 ng/ml)-induced IL-8 release was markedly inhibited by the steroids dexamethasone (Dex) (0.1 to 10 microM) and fluticasone (Flut) (0.01 to 1 microM) but unaffected by Salbu, Salme, FSK, or 8-Br-cAMP. However, a combination of Dex (1 microM) or Flut (0.1 microM) with Salbu (10 microM), Salme (1 microM), FSK (10 microM), or 8-Br-cAMP (10 and 100 microM) significantly enhanced the inhibition by Dex or Flut alone. Experiments with KT5720, a selective inhibitor of cAMP-dependent protein kinase A; rolipram, a selective inhibitor of type IV phosphodiesterase; and ICI-118,551, a beta(2)-receptor antagonist, suggested that the synergistic inhibition was mediated by beta(2)-receptor in a cAMP-dependent manner. This novel synergistic interaction of beta(2)-agonists and steroids may partly explain the benefits that result when these agents are combined to treat asthma.

Prostanoids may be involved in bradykinin (BK)-induced bronchoconstriction in asthma. We investigated whether cyclooxygenase (COX)-2 induction was involved in prostaglandin (PG) E2 release by BK in cultured human airway smooth muscle (ASM) cells and analyzed the BK receptor subtypes responsible. BK stimulated PGE2release, COX activity, and COX-2 induction in a concentration- and time-dependent manner. It also time dependently enhanced arachidonic acid release. In short-term (15-min) experiments, BK stimulated PGE2 generation but did not increase COX activity or induce COX-2. In long-term (4-h) experiments, BK enhanced PGE2 release and COX activity and induced COX-2. The long-term responses were inhibited by the protein synthesis inhibitors cycloheximide and actinomycin D and the steroid dexamethasone. The effects of BK were mimicked by the B2-receptor agonist [Tyr(Me)8]BK, whereas the B1 agonist des-Arg9-BK was weakly effective at high concentrations. The B2antagonist HOE-140 potently inhibited all the effects, but the B1 antagonist des-Arg9,(Leu8)-BK was inactive. This study is the first to demonstrate that BK can induce COX-2. Conversion of increased arachidonic acid release to PGE2 by COX-1 is mainly involved in the short-term effect, whereas B2 receptor-related COX-2 induction is important in the long-term PGE2 release.