(345a) Patient-Specific Design of Polymeric Delivery Vehicles for Anti-HIV Microbicides | AIChE

(345a) Patient-Specific Design of Polymeric Delivery Vehicles for Anti-HIV Microbicides

Authors 

Whitmore, T. W. - Presenter, University of Kansas
Markey, A. - Presenter, University of Kansas
Wilson, T. - Presenter, University of Kansas
Kieweg, S. L. - Presenter, University of Kansas


The purpose of this research is to design a novel microbicide delivery mechanism to prevent the transmission of HIV/AIDS to women in developing countries that will be both readily available and affordable. This delivery vehicle, either a gel or a cream, is usually polymeric. The design of these polymeric vehicles involves optimizing the molecular structure subject to specific property requirements. Computational Molecular Design (CMD) provides the methodology for this optimization problem. In the application of CMD, Quantitative Structural-Property Relationships (QSPRs) are generated and validated to relate the structure of possible polymers to their physicochemical properties. QSPRs are generated by regressing molecular topological indices, a quantitative method to represent the structure, to measured physiochemical properties. Such property data for 14 different polymeric liquids of different chain length prepared at the same concentration have been collected. The physiochemical properties considered for the QSPRs are power-law consistency, shear-thinning index, storage modulus, loss modulus, compressibility, anti-chlamydial activity, toxicity, and osmolality. Correlations were generated relating each of these properties to the structures of the 14 distinct polymeric liquids, and were then cross-validated statistically. These property prediction models were incorporated into the objective function of an MINLP formulation. Ideal values of these physicochemical properties were obtained from separate in vitro and in vivo experiments as well as review of the relevant literature. These data were then incorporated into the objective function, and the MINLP formulation was solved for local optima. This generated novel structures possessing physical property values near the ideal values. This problem was solved several times, with different ideal property values corresponding to differences in size and mechanical properties of vaginal tissue among different groups of women. Examples of these novel structures are provided.