Quantitative Structure-property Relationship Models Of Drug Release From PH-sensitive Block Copolymer Micelles

The pH sensitive drug-loaded block copolymer micelles,a promising drug delivery system,can respond to the pH difference between the cancer tissues and normal tissues,and achieve the target release of drug around tumor.Controlling the release with a slow and uniformly rate around the deceased tissues is one of the key objectives for drug delivery system design.In this article,we developed the quantitative structure-property relationship(QSPR)models for drug release performance of pH sensitive micelles,which could be used for release mechanism explanation and quantitative prediction,offering the theory guidance for the design and improvement of polymer structure to control the drug release.The quantitative structure-property relationship models between cumulative drug release percentage of four/six-arm star copolymer micelles and the copolymer structures were studied at two periods(stage I: 0~12 h,and stage II: 12~96 h)of pH 7.4 and 5.0 respectively,based on the multiple linear regression and genetic function approximation method.Three types of descriptors were compared,including the number-weighted,mass-weighted,and block unit autocorrelation(BUA)descriptors.The results indicated that BUA descriptor is the most suitable for characterizing the copolymer structures.The drug diffusion relates with inter-block forces and micellar erosion with intra-block forces.At pH 5.0 and stage I,the drug release performance is affected by the hydrophobic property and stage II by the local polarity of copolymer molecules.At pH 7.4 the molecular surface area and volume of block unit are the main factor influencing drug release.Thus the amount of drug released in acid solution is related with the electronic structural factors,and neutral solution with the geometrical shapes of micelles.Also the ability of accepting electrons of copolymer is also the key affecting the drug release.Based on the Korsmeyer-Peppas semi-experimental equation,the TDBR(Time-Dependent BUA Release)models were built with BUA descriptors.The result turns out that the kinetics rate constant k and the release exponent n are the leading factors on drug release of stage I and stage II,respectively.The TDBR models are of good internal and external predictive ability.The external validation result in that TDBR models can predict the drug release amount of linear copolymer at p H 7.4,but at pH 5.0 they are suitable for the release of the multi-arm block unit copolymer micelle,which presents stable and relative slow release rate.