Effect of inhibition of cox2 expression and celecoxib treatment in a model of murine lung adenocarcinoma

GUILLERMO D. PELUFFO; ALEJANDRO J. URTREGER; MIRIAM DIAMENT; SLOBODANKA M. KLEIN

Anaheim, EEUU.

Congreso; 96th Annual Meeting of the American Association for Cancer Research (AACR); 2005

American Association for Cancer Research (AACR)

COX2 has been postulated as a key mediator in cancer development. It was shown to modulate tumor growth in patients as well as in murine models. It is supported partly by high COX2 expression in tumor tissues and abrogation of tumor progression by specific COX2 inhibitors like celecoxib (CXB). However, it has not yet been elucidated if COX2 expression is necessary for NSAIDs to inhibit tumor progression. We investigated the effects of CXB in a murine lung adenocarcinoma cell line (LP07) that expresses COX-2 protein and was stably transfected with an antisense vector for COX-2 mRNA (LP07AS). We have first assessed the effect of CXB on tumor growth in BALB/c mice inoculated sc with LP07 cells. CXB reduced tumor growth (p< 0.001) and metastases outcome (8.8±2.3 vs. 27.1±4.2 nod./lung in control mice; p< 0.001). It also diminished cachexia (-1.1±3.1 vs. 10.5±3.1% of weight lost, p< 0.01) and leukocytosis (3.4±1.5 vs. 7.0±1.2 leuk. x 104/ml, p< 0.05), paraneoplastic syndromes associated with LP07 tumor evolution. The activity of MMP2 and uPA, two proteases involved in tumor invasion were reduced in conditioned media (CM) from CXB treated LP07C (control) cells (1.2±0.2 vs. 1.7±0.3 AU, p< 0.05 and 1.6±0.2 vs. 2.9±0.3 AU, p< 0.01, respect. vs. control group). Both activities were also reduced in COX2AS transfected cells (0.5±0.1, p< 0.05 and 2.0±0.1 AU, p< 0.05). CXB treatment did not produce further inhibition of MMP2 and uPA activities in CM from these cells. Unexpectedly, when LP07AS cells were sc inoculated in BALB/c mice, tumor growth was greater than in LP07C cell tumor bearing mice (TBM) (855±151 vs. 328±117mm3, p< 0.01). However, the growh of both LP07C and LP07AS tumors was reduced by CXB (49±21, p< 0.05 and 149±52mm3, p< 0.001), indicating that this effect was COX2 expression independent. The number of lung metastases was not significantly different LP07AS TBM compared to control group. CXB treatment diminished metastasis outcome for both cell types (p< 0.05 for LP07C and p< 0.001 for LP07AS). Leukocytosis and cachexia (weight loss) were more pronounced in LP07AS TBM than in control mice (p< 0.01 and p< 0.05 respectively), and CXB treatment resulted in a partial inhibition in both cell types (p< 0.01 for LP07C and p< 0.01 for LP07AS). In LP07 TBM, COX2 is not the target of CXB for its in vivo antitumoral effects. However, in vitro the inhibition of MMP2 and uPA activities were completely attributable to COX2 inhibition by CXB, given that the reduction of COX2 expression did not result in further inhibition of such activities. It must be noted that in vivo COX2 expression is not confined to epithelial tumor cells and that COX2 expressed by endothelial or immune infiltrating cells or fibroblasts may play a role in tumor progression. We conclude that other targets may be postulated for NSAIDs, like NFkB or PPARg that could account for the antitumoral action of CXB, and that the role of COX2 in tumor progression should be further investigated.