Background: An increasing number of reports is challenging the notion that the antitumor potential of the selective COX-2 inhibitor celecoxib (Celebrex) is mediated primarily via the inhibition of COX-2. We have investigated this issue by applying two different analogs of celecoxib that differentially display COX-2-inhibitory activity: the first analog, called unmethylated celecoxib (UMC), inhibits COX-2 slightly more potently than its parental compound, whereas the second analog, 2,5-dimethyl-celecoxib (DMC), has lost the ability to inhibit COX-2.
Results: With the use of glioblastoma and pancreatic carcinoma cell lines, we comparatively analyzed the effects of celecoxib, UMC, and DMC in various short-term (< or =48 hours) cellular and molecular studies, as well as in long-term (< or =3 months) focus formation assays. We found that DMC exhibited the most potent antitumor activity; celecoxib was somewhat less effective, and UMC clearly displayed the overall weakest antitumor potential in all aspects. The differential growth-inhibitory and apoptosis-stimulatory potency of these compounds in short-term assays did not at all correlate with their capacity to inhibit COX-2, but was closely aligned with their ability to trigger endoplasmic reticulum stress (ERS), as indicated by the induction of the ERS marker CHOP/GADD153 and activation of the ERS-associated caspase 7. In addition, we found that these compounds were able to restore contact inhibition and block focus formation during long-term, chronic drug exposure of tumor cells, and this was achieved at sub-toxic concentrations in the absence of ERS or inhibition of COX-2.
Conclusion: The antitumor activity of celecoxib in vitro did not involve the inhibition of COX-2. Rather, the drug's ability to trigger ERS, a known effector of cell death, might provide an alternative explanation for its acute cytotoxicity. In addition, the newly discovered ability of this drug to restore contact inhibition and block focus formation during chronic drug exposure, which involved neither ERS nor COX-2, suggests a novel, as yet unrecognized mechanism of celecoxib action.