Assmus, Frauke. Artificial tissue binding models : development and comparative evaluation of high - throughput lipophilicity assays and their use for PET tracer optimization. 2015, Doctoral Thesis, University of Basel, Faculty of Science.
|
PDF
10Mb |
Official URL: http://edoc.unibas.ch/diss/DissB_11224
Downloads: Statistics Overview
Abstract
The purpose of this thesis was to increase the efficiency of the Positron Emission Tomography (PET) tracer development process. Since many neuroimaging agents fail due to undesirably high non-specific binding (NSB) to brain tissue, we aimed at estimating the extent of NSB as early as possible, preferably before radioactive labeling and extensive animal testing. To this purpose we have developed, optimized and evaluated several in vitro assays with respect to their ability to predict brain tissue binding and, in particular, NSB in PET. A major goal of this thesis was the implementation of a miniaturized assay for the prediction of NSB in order to meet the demand for maximal efficiency, i.e. high throughput and minimal consumption of reagents, samples and animal tissue. Since octanol/water distribution coefficients (logDoct) are routinely measured in almost every research organization, we investigated whether logDoct is also useful for the prediction of brain tissue binding. In this context, we have developed a filter-based logDoct assay (Carrier Mediated Distribution System=CAMDIS) to overcome the drawbacks of the traditional shake flask technique, i.e. tedious phase separation and high consumption of reagents. Strategies have been developed to correct for drug adsorption to the assay construct in order to warrant both high throughput and high quality of the data. Even though the CAMDIS logDoct values were in excellent agreement with literature shake flask data, our results indicated that octanol is only a poor surrogate for tissue binding, as shown by the poor correlation between logDoct and the unbound fraction of drug in brain (fu,brain) available through equilibrium dialysis. The latter is the current industrial standard method for the measurement of tissue binding, however the technique is hampered by high consumption of animal tissue and low throughput. Apart from logDoct, another, more complex membrane surrogate system, namely the Parallel Artificial Membrane Permeation Assay (PAMPA), has found entry into many laboratories. We investigated whether the fraction of drug retained by the PAMPA barrier proves useful for the prediction of tissue binding. Since the default PAMPA setup at Roche was inappropriate in this respect, we optimized PAMPA towards better predictive power and compatibility with mass spectrometric analysis. Provided that PAMPA was conducted under optimized conditions (pH 7.4, brain polar lipids, without solubilizers), the membrane fraction was in much better agreement with tissue binding as compared to logDoct. Nevertheless, the predictive power was still unsatisfactory reflecting the fact that reverse micelles rather than lipid bilayers constitute the permeation barrier as revealed by NMR experiments. Since neither CAMDIS nor PAMPA yielded sufficiently reliable NSB estimates, we developed a miniaturized label-free Lipid Membrane Binding Assay (LIMBA) allowing for the measurement of brain tissue/water distribution coefficients at minimal consumption of brain homogenate. LIMBA was highly predictive for the binding of drugs and molecular imaging probes to brain tissue and therefore provides a viable alternative to the equilibrium dialysis technique. LIMBA thus allows for more efficient optimization of potential PET tracers and should reduce the attrition rate in the late and particularly expensive stages in the PET tracer development process.
Advisors: | Seelig-Löffler, Anna |
---|---|
Committee Members: | Ernst, Beat |
Faculties and Departments: | 05 Faculty of Science > Departement Biozentrum > Former Organization Units Biozentrum > Biophysical Chemistry (Seelig A) |
UniBasel Contributors: | Seelig-Löffler, Anna and Ernst, Beat |
Item Type: | Thesis |
Thesis Subtype: | Doctoral Thesis |
Thesis no: | 11224 |
Thesis status: | Complete |
Number of Pages: | 351 S. |
Language: | English |
Identification Number: |
|
edoc DOI: | |
Last Modified: | 02 Aug 2021 15:11 |
Deposited On: | 05 May 2015 15:05 |
Repository Staff Only: item control page