Understanding the Antitumor Activity of Novel Hydroxysemicarbazide Derivatives as Ribonucleotide Reductase Inhibitors Using CoMFA and CoMSIA Anand V. Raichurkar and Vithal M. Kulkarni* ,† Pharmaceutical Division, Institute of Chemical Technology, University of Mumbai, Matunga, Mumbai-400019, India Received January 8, 2003 Three-dimensional quantitative structure-activity relationship (3D-QSAR) studies were performed on a series of Schiff bases of hydroxysemicarbazide analogues using comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) methods with their antitumor activities against L1210 cells. The models were generated using 24 molecules, out of which one molecule was a commercially available ribonucleotide reductase (RR) inhibitor, hydroxyurea (HU), and the predictive ability of the resulting each model was evaluated against a test set of four molecules. Maximum common substructure (MCS)-based method was used for alignment and compared with the known alignment methods. The QSAR models from both methods exhibited considerable correlative and predictive properties. Inclusion of additional descriptor ClogP improved the statistics of CoMFA model significantly. Both methods strongly suggest the necessity of lipophilicity for antitumor activity. CoMFA and CoMSIA methods predicted HU optimally, indicating a similar mechanism of action for the molecules considered for generating the models and HU to inhibit the tumor cells. The analysis of CoMFA contour maps provided insight into the possible modification of the molecules for better activity. Introduction Among the devastating and pandemic diseases, cancer is a major disease and according to WHO, it is consid- ered as the fourth largest killer disease. Treatment of cancer has been one of the primary goals of medicine for the last two decades. Though several therapies namely surgery, photodynamic therapy, radiation ther- apy, chemoimmunotherapy, gene therapy, and chemo- therapy are available, chemotherapy is considered as an effective approach for combating cancer. However, in view of the lack of selectivity, increasing incidence of resistance to current drug regimens, and the fre- quency of adverse events, the development of novel, selective, potent, and safe antitumor agents that are also active against mutant cells has remained a high priority. Ribonucleotide reductase (RR) is a key enzyme, which plays a major role in the DNA synthesis and repair in all dividing cells. 1 It catalyzes the reduction of ribo- nucleotides and provides the building blocks for the de novo DNA synthesis. Various biological and experimen- tal studies have indicated the critical role of RR in neoplastic expression and tumor promotion. Therefore, RR is considered as a relevant molecular target for the design and development of antitumor agents. 2 Several compounds with hydroxyguanidine, thiosemi- carbazide, and substituted benzohydroxamic acid func- tional groups have shown promising RR inhibitory property with antitumor activity. 3-5 On the basis of previous reports, it is believed that (-C(dX)NHOH; X ) O, NH) is an essential pharmacophore for antitumor/ antiviral activities. 6 Hydroxyurea (HU) and triapine, which inhibit RR, have emerged commercially as anti- tumor agents, and recently RR inhibitors have addition- ally shown antiviral activities and hence are used in the treatment of AIDS as an adjuvant therapy. 7-10 The dual therapeutic activities, antitumor and antiviral, of RR inhibitors have made RR as one of the promising targets for the design and development of novel drugs. In the search of more potent and selective antitumor agents, Ren et al. synthesized a series of Schiff bases of hydroxysemicarbazide (SB-HSC) with the essential pharmacophore (NHCONHOH) and tested them for antitumor activity against L1210 cancer cells. 11 The compounds were synthesized with a view to understand the role of aryl group and substitution pattern that affect antitumor activity. This resulted in a data set of compounds with a wide spectrum of activities. Though the compounds were tested against L1210 cancer cells, Ren et al. hypothesized that the presence of the phar- macophore, which is also present in HU, may induce the observed activity by inhibiting the RR enzyme. However, there was no experimental evidence for this hypothesis. Therefore, to rationalize the observed vari- ance in the biological activity, to propose a possible mechanism of antitumor activity as a further support to the hypothesis of Ren et al., and to guide the syn- thesis of additional compounds, we have derived three- dimensional quantitative structure-activity relation- ship (3D-QSAR) models for SB-HSC derivatives using comparative molecular field analysis (CoMFA) and com- parative molecular similarity indices analysis (CoMSIA) methods. These methods were applied in the past to various therapeutic areas in our laboratory. 12-14 The 3D-QSAR, CoMFA method was proposed by Cramer et al. in 1988, which is extensively used in the present practice of drug discovery. 15 One of the advan- * To whom correspondence should be addressed. Tel: +91-22- 24186718. Fax: +91-22-24145614. E-Mail: vithal@biogate.com. † Present address: Bharati Vidyapeeth Deemed University, Poona College of Pharmacy, Pune-411008, India. E-mail: vmkulkarni60@ yahoo.co.in. 4419 J. Med. Chem. 2003, 46, 4419-4427 10.1021/jm030016a CCC: $25.00 © 2003 American Chemical Society Published on Web 09/13/2003