tx060188l_si_001.pdf (11.64 kB)
Structural Determinants of Binding of Aromates to Extracellular Matrix: A Multi-Species Multi-Mode CoMFA Study
journal contributionposted on 2007-01-16, 00:00 authored by Yufen Zhang, Viera Lukacova, Vladimir Bartus, Stefan Balaz
For small molecules acting in tissues, including signaling peptides, effectors, inhibitors, and other drug candidates, nonspecific binding to the extracellular matrix (ECM) is a critical phenomenon affecting their disposition, toxicity, and other effects. A commercially available ECM mimic, forming a solidified layer at the bottom of the vials, was used to measure the association constants of 25 simple aromatic compounds to two forms of ECM proteins, solidified (s-ECM) and dissolved (d-ECM) in the buffer during the incubation. Except for small homologous series, the binding data did not correlate with the lipophilicity and acidity of the compounds, contrary to a common expectation for nonspecific binding. To elucidate the putative structures of averaged binding sites of s-ECM and d-ECM, comparative molecular field analysis (CoMFA) was applied in a modified version taking into consideration that multiple modes and multiple species may be involved. The method shapes a receptor site model from a set of grid points in which the interaction energies between a probe atom and superimposed ligands are calculated. Electrostatic and steric energies in the grid points are characterized by regression coefficients. The forward-selection nonlinear regression analysis was used to optimize the coefficients in the novel multi-species, multi-mode CoMFA models. These models showed satisfactory descriptive and predictive abilities for both s-ECM and d-ECM binding data, which were better than those obtained with the standard, one-mode CoMFA analysis. The calibrated models, defining the electrostatic and van der Waals regions of putative binding sites, are suitable for the prediction of ECM binding for untested chemicals.