Date Published: January 23, 2017
Publisher: Public Library of Science
Author(s): Shan Wang, Ina Willenberg, Michael Krohn, Tanja Hecker, Sven Meckelmann, Chang Li, Yuanjiang Pan, Nils Helge Schebb, Pablo Steinberg, Michael Telamon Empl, Aamir Ahmad.
Although resveratrol exerts manifold antitumorigenic effects in vitro, its efficacy against malignancies in vivo seems limited. This has been increasingly recognized in recent years and has prompted scientists to search for structurally related compounds with more promising anticarcinogenic and/or pharmacokinetic properties. A class of structurally modified resveratrol derivatives, so-called resveratrol imine analogs (IRA’s), might meet these requirements. Therefore, the biological activity of five of these compounds was examined and compared to that of resveratrol. Firstly, the antiproliferative potency of all five IRA’s was investigated using the p53 wildtype-carrying colorectal carcinoma cell line HCT-116wt. Then, using the former and a panel of various other tumor cell lines (including the p53 knockout variant HCT-116p53-/-), the growth-inhibiting and cell cycle-disturbing effects of the most potent IRA (IRA 5, 2-[[(2-hydroxyphenyl)methylene]amino]-phenol) were studied as was its influence on cyclooxygenase-2 expression and activity. Finally, rat liver microsomes were used to determine the metabolic stability of that compound. IRA 5 was clearly the most potent compound in HCT-116wt cells, with an unusually high IC50-value of 0.6 μM. However, in the other five cell lines used, the antiproliferative activity was mostly similar to resveratrol and the effects on the cell cycle were heterogeneous. Although all cell lines were affected by treatment with IRA 5, cells expressing functional p53 seemed to react more sensitively, suggesting that this protein plays a modulating role in the induction of IRA 5-mediated biological effects. Lastly, IRA 5 led to contradictory effects on cyclooxygenase-2 expression and activity and was less glucuronidated than resveratrol. As IRA 5 is approximately 50 times more toxic towards HCT-116wt cells, exerts different effects on the cyclooxygenase-2 and is metabolized to a lesser extent, it shows certain advantages over resveratrol and could therefore serve as basis for additional chemical modifications, potentially yielding compounds with more favorable biological and pharmacokinetic features.
Since Jang et al.  published a study linking the natural stilbenoid resveratrol (Fig 1A) to cancer chemoprevention in the mid 1990’s, a plethora of studies have been performed to investigate this connection in more detail . Up to now, a high number of published studies have reported that this polyphenol exerts manifold biological effects in vitro, strongly suggesting that it might inhibit or prevent the onset of cancer [2, 3]. For instance, the potential cancer-repressing effects investigated in vitro include anti-oxidative, anti-inflammatory, growth-inhibiting, pro-apoptotic, and anti-metastatic properties (reviewed in ). In addition, numerous animal studies suggest that resveratrol might indeed be able to inhibit carcinogenesis in vivo (reviewed in  and ). Nevertheless, not all animal studies have rendered promising results (see references  and  for a comprehensive listing of performed animal studies), and the outcomes of the few clinical trials conducted in human cancer patients are far from showing that resveratrol is notably helpful in preventing or treating cancer [6–8]. For example, in multiple myeloma patients, this compound even induced adverse effects . Moreover, there is a rather vast discrepancy between resveratrol concentrations biologically active in cellular models in vitro (up to 500 μM but mostly in the 20–100 μM range; reviewed in ) and the maximum plasma concentrations (967 ng/ml = approx. 4 μM) achievable in humans after oral administration of very high doses (i.e. 5 g; ). The inconsistency between resveratrol concentrations that can be reached in vivo and those that are efficient in vitro as well as the absence of a clearly demonstrated in vivo efficacy can mostly be explained by the fast metabolization (i.e. glucuronidation and sulfonation) of this compound ( and reviewed in ). This results in a very low overall bioavailability, although the absorption of orally administered resveratrol is relatively high (reviewed in  and ). Consequently, it is not surprising that a number of studies proposing the search for molecules more suited for use in cancer therapy or chemoprevention and/or investigating the anticarcinogenic/chemopreventive efficacy as well as metabolic stability of natural or synthetic compounds related to resveratrol have been published (e.g. [15–22]).
This is the first study investigating the anticarcinogenic activity/chemopreventive potential of five different resveratrol imine analogs on a panel of different human tumor cell lines. While four of these compounds did not significantly inhibit the growth in an initial cytotoxicity screen using HCT-116wt colorectal carcinoma cells, IRA 5 proved to be unusually potent in this regard, with an IC50 value approximately 50 times lower than that of resveratrol. However, in other tumor cell lines, the growth-inhibitory effects where mostly similar to those of resveratrol. Moreover, a cell line-dependent cell cycle dysregulation followed by an activation of apoptosis in certain cell lines upon treatment with IRA 5 was observed, yet mostly at high concentrations. Based on the results obtained regarding the IRA 5-related effects on the expression and activity of COX-2, this polyphenol can be considered an inhibitor of this enzyme. Lastly, IRA 5 was shown to be metabolically more stable than resveratrol.