1105 Evaluation of Quantum Mechanical Quantitative Structure Activity Relationships for Design of Non-sensitizing Dental Materials
D.M. YOURTEE1, A. AGRAWAL1, A.J. HOLDER2, M.D. MILLER1, R.E. SMITH1, E.L. KOSTORYZ1, and J.D. EICK1, 1 University of Missouri, Kansas City, USA, 2 University of Missouri-Kansas City, USA

Sensitization reactions to dental resins have been cited. One way to aid the design of non-sensitizing components is to develop Quantum Mechanical Quantitative Structure Activity Relationships (Q-QSAR). Objective: To determine whether sensitizing agents have common fundamental chemical and/or quantum mechanical characteristics related to their activity. Method: The study employed sensitization data obtained using the local lymph node assay (LLNA) (Toxicology, 103, 177-194, 1995). These data were converted to slopes. The chemical structures of the sensitizing agents were optimized using AM1, a semiempirical quantum mechanical method. Then, the CODESSA program was used to correlate quantum chemical descriptors from information generated by the AM1 results and statistically matched these descriptors with observed properties. This was applied to 5 classifications of reference chemicals: phenylating, acylating, benzoylating, potential Michael-reactive and miscellaneous agents. The programs TOXSYS and SciQSAR were used to calculate log P. Results: Log P was not a common dependency among the chemicals associated with sensitization results. Using CODESSA, r2 values were found for each set in the study exceeding 0.90. The most important descriptors for phenylating agents were the Maximum Atomic Charge and Minimum Atomic State Energy for a carbon atom. The statistically significant descriptor for acylating agents was FNSA2, and for benzolyating the number of aromatic bonds. The descriptors for the miscellaneous classification were determined by the principal moment of inertia C, Qmax, and WNSA3, FNSA3 and PPSA2. Conclusion: This analysis supported Ashby et al in that sensitization was selective to chemical type. As the Q-QSAR identified specific chemical causes, it was suggested that Q-QSARs constructed on a multi-node basis have potential to identify structural components that should be removed to avoid sensitization in proposed biomaterials. This, in turn, has potential to reduce use of animals and the associated cost of those experiments. Supported by NIH/NIDR grant DE09696.

Seq #140 - Oral Tissues, Pharmacology
11:00 AM-12:15 PM, Friday, 14 March 2003 Henry B. Gonzalez Convention Center Exhibit Hall C

Back to the Pharmacology, Therapeutics, & Toxicology Program
Back to the 32nd Annual Meeting and Exhibition of the AADR (March 12-15, 2003)

Top Level Search