To search for topoisomerase II active compounds, one could use an agent with established clinical efficacy such as VP-16 or VM-26 as the seed. The best matches derived would be compounds exhibiting a pattern of tumor cell inhibition similar to that of the seed. Alternately, use of a synthetic compound as the seed has the potential to furnish a list of compounds with subtle differences in antitumor properties from those of the standard agents. Among the synthetic topoisomerase II inhibitors, there are numerous potential seeds. In the example presented here, we chose a semisynthetic demethylepipodophyllotoxin derivative only somewhat related in structure to VP-16 (77). The seed compound was one of a set of (arylamino)demethylepipodophyllotoxins that are closely related in structure (77). The synthetic compound database was searched at both the GI50 (Table XV) and TGI levels (Table XVI). At the time of the search, the database contained data for more than 20,000 compounds. At each level, only the top 40 matches from among the entire database are listed.
Among these 40 compounds at both the GI50 and TGI levels, VP-16 analogues predominate. At each level, it is noteworthy that neither VP-16 nor VM-26 appears among the top matches. The COMPARE result suggests that the pattern of cell inhibition by these analogues is different from that of VP-16.
Cheng et al. (78) have evaluated the seed compound and five of the other VP-16 analogues found by COMPARE for their activity as inhibitors of human DNA topoisomerase II in vitro and their cytotoxic efficacy against the KB cell line and its VP-16-resistant variants. As inhibitors of DNA topoisomerase II, five of these compounds were found to be 5-10-fold more potent than VP-16. Not only were the compounds cytotoxic to KB cells but also the compounds were cytotoxic to the variants that exhibited a lower DNA topoisomerase II content or overexpression of MDR1 phenotype and a decreased cellular uptake of VP-16. Thus, the absence of VP-16 from the top COMPARE matches may reflect differences in the pattern of growth inhibition of the NCI cell line panel stemming from mechanistic differences in antitumor activity by VP-16 and this group of analogues.
The only standard agents that appear in the lists are menogaril and AMSA, both DNA topoisomerase II inhibitors. At both the GI50 and TGI levels, the list was interspersed with acridine derivatives, most of which were structural relatives of AMSA. Also, at the TGI level anthracyclines and anthraquinones were found, which are structurally related to compounds well known to be topoisomerase II inhibitors.
At the TGI level, two camptothecin derivatives appeared in the list near the bottom. This is analogous to the small number of topoisomerase II inhibitors that were found at the bottom of the COMPARE list when camptothecin was employed as the seed at GI50 level. Although these results may be due to variability in the in vitro bioassay, it is also feasible that there is a mechanistic basis for the result.
In accord with any request by investigators submitting compounds to the program for testing who wish to maintain the chemical structures confidential, an illustration of examples of heretofore unknown topoisomerase II inhibitors is presently precluded. Notwithstanding, the analysis using the VP-16 analogue as the seed depicts the technique that is used for such discoveries. For selected discoveries, experimental confirmation of mechanism in the laboratory has served to stimulate the continued use of COMPARE.