Year: 2023
Type: dissertation
Abstract: In the early 2000s, when the investigations of this thesis were started, new experimental methodologies were producing increasing numbers of pharmacologically interesting drug candidate molecules.Since poor pharmacokinetics of a compound had been a major contributor to expensive late-stage failures in drug development, methodologies for early prediction of the pharmacokinetics of a compound were needed more than ever.The primary objective of this thesis was to explore methods for predicting passive molecule transport across the intestinal membrane.Novel Quantitative Structure-Property Relationship (QSPR) models to predict the kinetics of passive oral drug absorption in humans were developed using multivariate analysis.The predictivities of the developed models were compared to the predictivities of previously developed computational and experimental models.Literature data were used in the analyses.To better understand cell model predictions, the paracellular pathways of the three different cell models and human intestinal tissue were characterized.The pore sizes and porosities were ascertained by determining the permeabilities of polyethylene glycol (PEG) oligomers through the examined membranes.The results of this work demonstrated that passive human intestinal absorption could be well predicted using a combination of two or three 8 simple computational lipophilicity and polarity descriptors.The developed QSPR models were shown to predict absorption equally well as the experimental Caco-2 and 2/4/A1 cell models, and, notably, one of the QSPR models predicted absorption nearly as well as the in vivo rat model.We suggest that previously reported tendency of the Caco-2 cell model to underestimate paracellular drug transport results from differences in the porosities of the Caco-2 monolayer and the human intestine.The porosity of the Caco-2 cell model was found to be about eight times lower than the porosity of the human intestinal epithelium.The pore sizes in the Caco-2 monolayer (6 and 10 Å) and the human jejunal membrane (7 and 10 Å) were similar.The human jejunal pore sizes determined in this thesis were in line with previously determined human and rat intestinal pore sizes.Our study was the first to experimentally determine the surface area of the paracellular pores relative to the total intestinal surface area in the human intestine and cell models.In summary, the research in this thesis has provided new insight into computational and experimental methods for assessing the permeability of drug candidates during the initial phases of drug discovery.
Author: Linnankoski, Johanna
Institution: University of Eastern Finland
Language: English
Subfield: Pharmaceutical Science
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7215189022
