A new computational method for deducing quantitative structure-activity relationships (QSARs) using structural data from ligand-macromolecule complexes is presented. First, the ligand-macromolecule interaction energy is computed for a set of ligands using molecular mechanics calculations. Then, by selecting and scaling components of the ligand-macromolecule interaction energy that show good predictive ability, a regression equation is obtained in which activity is correlated with the interaction energies of parts of the ligands and key regions of the macromolecule. Application to the interaction of the human synovial fluid phospholipase A2 with 26 inhibitors indicates that the derived QSAR has good predictive ability and provides insight into the mechanism of enzyme inhibition. The method, which we term comparative binding energy (COMBINE) analysis, is expected to be applicable to ligand-receptor interactions in a range of contexts including rational drug design, host-guest systems, and protein engineering.

Thymidine phosphorylase (TP), also known as "platelet-derived endothelial cell growth factor" (PD-ECGF), is an enzyme, which is upregulated in a wide variety of solid tumors including breast and colorectal cancers. TP promotes tumor growth and metastasis by preventing apoptosis and inducing angiogenesis. Elevated levels of TP are associated with tumor aggressiveness and poor prognosis. Therefore, TP inhibitors are synthesized in an attempt to prevent tumor angiogenesis and metastasis. TP is also indispensable for the activation of the extensively used 5-fluorouracil prodrug capecitabine, which is clinically used for the treatment of colon and breast cancer. Clinical trials that combine capecitabine with TP-inducing therapies (such as taxanes or radiotherapy) suggest that increasing TP expression is an adequate strategy to enhance the antitumoral efficacy of capecitabine. Thus, TP plays a dual role in cancer development and therapy: on the one hand, TP inhibitors can abrogate the tumorigenic and metastatic properties of TP; on the other, TP activity is necessary for the activation of several chemotherapeutic drugs. This duality illustrates the complexity of the role of TP in tumor progression and in the clinical response to fluoropyrimidine-based chemotherapy.

A comparative binding energy (COMBINE) analysis (Ortiz et al. J. Med. Chem. 1995, 38, 2681-2691 has been performed on a training set of 33 HIV-1 protease inhibitors, and the resulting regression models have been validated using an additional external set of 16 inhibitors. This data set was originally reported by Holloway et al. (J. Med. Chem. 1995, 38, 305-317), who showed the usefulness of molecular mechanics interaction energies for predicting the activity of novel HIV-1 protease inhibitors within the framework of the MM2X force field and linear regression techniques. We first used the AMBER force field on the same set of threedimensional structures to check up on any possible force-field dependencies. In agreement with the previous findings, the calculated raw ligand-receptor interaction energies were highly correlated with the inhibitory activities (r 2 ) 0.81), and the linear regression model relating both magnitudes had an acceptable predictive ability both in internal validation tests (q 2 ) 0.79, SDEP cv ) 0.61) and when applied to the external set of 16 different inhibitors (SDEP ex ) 1.08). When the interaction energies were further analyzed using the COMBINE formalism, the resulting PLS model showed improved fitting properties (r 2 ) 0.89) and provided better estimations for the activity of the compounds in the external data set (SDEP ex ) 0.83). Computation of the electrostatic part of the ligand-receptor interactions by numerically solving the Poisson-Boltzmann equation did not improve the quality of the linear regression model. On the contrary, incorporation of the solvent-screened residue-based electrostatic interactions and two additional descriptors representing the electrostatic energy contributions to the partial desolvation of both the ligands and the receptor resulted in a COMBINE model that achieved a remarkable predictive ability, as assessed by both internal (q 2 ) 0.73, SDEP cv ) 0.69) and external validation tests (SDEP ex ) 0.59). Finally, when all the inhibitors studied were merged into a single expanded set, a new model was obtained that explained 91% of the variance in biological activity (r 2 ) 0.91), with very high predictive ability (q 2 ) 0.81, SDEP cv ) 0.66). In addition, the COMBINE analysis provided valuable information about the relative importance of the contributions to the activity of individual residues that can be fruitfully used to design better inhibitors. All in all, COMBINE analysis is validated as a powerful methodology for predicting binding affinities and pharmacological activities of congeneric ligands that bind to a common receptor.

Limited data are available on the genotypic and phenotypic resistance profile of the ␣-(1-2)mannose oligomer-specific prokaryotic lectin cyanovirin (CV-N). Therefore, a more systematic investigation was carried out to obtain a better view of the interaction between CV-N and human immunodeficiency virus type 1 (HIV-1) gp120. When HIV-1-infected CEM cell cultures were exposed to CV-N in a dose-escalating manner, a total of eight different amino acid mutations exclusively located at N-glycosylation sites in the envelope surface gp120 were observed. Six of the eight mutations resulted in the deletion of high-mannose type N-glycans (i.e., at amino acid positions 230, 332, 339, 386, 392, and 448). Two mutations (i.e., at position 136 and 160) deleted a complex type N-glycan in the variable V1/V2 domain of gp120. The level of phenotypic resistance of the mutated virus strains against CV-N generally correlated with the number of glycan deletions in gp120, although deletion of the glycans at N-230, N-392, and N-448 generally afforded a more pronounced CV-N resistance than other N-glycan deletions. However, the extent of the decrease of antiviral activity of CV-N against the mutated virus strains was markedly less pronounced than observed for ␣(1-3)-and ␣(1-6)-mannose-specific plant lectins Hippeastrum hybrid agglutinin (HHA) and Galanthus nivalis agglutinin (GNA), which points to the existence of a higher genetic barrier for CV-N. This is in agreement with a more consistent suppression of a wider variety of HIV-1 clades by CV-N than by HHA and GNA. Whereas the antiviral and in vitro antiproliferative activity of CV-N can be efficiently reversed by mannan, the pronounced mitogenic activity of CV-N on peripheral blood mononuclear cells was unaffected by mannan, indicating that some of the observed side effects of CV-N are unrelated to its carbohydrate specificity/activity.

Trabectedin (Yondelis) is a potent antitumor drug that has the unique characteristic of killing cells by poisoning the DNA nucleotide excision repair (NER) machinery. The basis for the NER-dependent toxicity has not yet been elucidated but it has been proposed as the major determinant for the drug's cytotoxicity. To study the in vivo mode of action of trabectedin and to explore the role of NER in its cytotoxicity, we used the fission yeast Schizosaccharomyces pombe as a model system. Treatment of S. pombe wild-type cells with trabectedin led to cell cycle delay and activation of the DNA damage checkpoint, indicating that the drug causes DNA damage in vivo. DNA damage induced by the drug is mostly caused by the NER protein, Rad13 (the fission yeast orthologue to human XPG), and is mainly repaired by homologous recombination. By constructing different rad13 mutants, we show that the DNA damage induced by trabectedin depends on a 46-amino acid region of Rad13 that is homologous to a DNA-binding region of human nuclease FEN-1. More specifically, an arginine residue in Rad13 (Arg961), conserved in FEN1 (Arg314), was found to be crucial for the drug's cytotoxicity. These results lead us to propose a model for the action of trabectedin in eukaryotic cells in which the formation of a Rad13/DNAtrabectedin ternary complex, stabilized by Arg961, results in cell death. (Cancer Res 2006; 66(16): 8155-62)