Cox2 dependent and independent effect of celecoxib on a murine lung adenocarcinoma

GUILLERMO D. PELUFFO; ALEX SCHÓNTAL; ALEJANDRO J. URTREGER; MARÍA GISELLE PETERS; SLOBODANKA M. KLEIN

Washington, EEUU.

Congreso; 97th Annual Meeting of the American Association for Cancer Research; 2006

American Association for Cancer Research (AACR)

COX-2 expression in tumor cells has been implicated in inhibition of apoptosis, tumor growth and an invasive behavior. Nonsteroidal anti-inflammatory drugs (NSAIDs) exert their antitumor effects through COX-2-dependent and independent mechanisms. The murine lung adenocarcinoma LP07 overexpressed COX-2 and is sensitive to the NSAID celecoxib. To determine whether COX-2 is involved in this response we transfected LP07 cells with a COX-2 antisense vector (LP07AS) and treated them with dimethylcelecoxib (DMC), a celecoxib derivative that lacks COX-2 inhibitory properties. LP07AS cell viability remained unaffected as no differences in growth rates were found compared to control cells. Celecoxib and DMC inhibited the growth of LP07 cells, being DMC slightly more potent (LD50 below 25uM whereas it was between 25 and 50uM for celecoxib, assayed after 96h). Both drugs retained their growth inhibitory properties in LP07AS cells. Given that COX-2 seems not to be the target of the drugs, we tested their effect on the activity of NFkB and AP-1, transcription factors described as modulated by NSAIDs. NFkB activity in LP07 cells was dose-dependently reduced by celecoxib and DMC in a reporter gene assay. At 50uM celecoxib reduced NFkB activity by 41±3% while DMC did it by 49±2%. Prostaglandin E2 at 10uM could not revert this inhibition. Interestingly, AP-1 activity was increased by both drugs in a dose-dependent manner (1.9 and 2.4-fold by celecoxib and DMC respectively). As NSAIDs affect the metastasis outcome in sc LP07 bearing BALB/c mice, we studied the effects of COX-2-expression and activity inhibition on MMP-2 and uPA, secreted proteases involved in an invasive behavior. COX-2AS cells exhibited a reduced secreted MMP-2 (39±3% of control) and uPA ( 59±2% of control) activities, that could not be further decreased by celecoxib treatment. DMC reduced both activities in LP07 cells (26±4% and 68±3% of control, respectively). This effect was even increased in LP07AS cells. Both celecoxib and DMC at 50uM inhibited the migration of LP07 cells in a wound repair assay (51±6% and 45±4% of control cells, respectively). Although the action of celecoxib and DMC on LP07 cell viability is clearly COX-2-independent, the inhibition of MMP-2 and uPA activities by celecoxib involves COX-2 and this feature is not shared by DMC, which further inhibited both activities in LP07AS cells. NFkB could be addressed as a target of celecoxib and DMC. As previously reported, NFkB could render cells more resistant to apoptosis and its inhibition may explain the reduced viability when LP07AS cells were treated with either celecoxib or DMC. Conversely, the unexpected rise in AP-1 activity indicates that in LP07 cells this transcription factor may not be mitogenic or a balance between signals that mask this effect may be present.