Development of in-vitro models to study effects of compounds that modulate cellular functions in patients with mitochondrial disease

Thesis

Background: Adverse drug reactions caused by off-target activity can limit the use or withdrawal of otherwise effective drugs. Mitochondria have been recognised as potential target for inadvertent drug activity thus causing toxicity. It is difficult to detect drug induced mitochondrial toxicity in pre-clinical animal models and in human trials. Drug trials are carried out in healthy humans and it is not until a large population has been exposed to the drug that adverse reactions leading to mitochondrial toxicity come to light. The development of pre-clinical in vitro assays to detect drug induced mitochondrial toxicity can help reduce adverse drug reactions and drug development costs. In this project, we used different in vitro cell models (with or without mitochondrial disease) to study the effect of the anticonvulsant drug, sodium valproate (VPA) on cellular functions such as mitochondrial respiration, under both normal conditions (glucose) and energetic stress (galactose). To investigate changes to the morphology of the mitochondrial network, we developed a novel high-throughput imaging method.

Results: The results presented in this thesis indicate that sodium valproate treatment did not have the same effect on cellular activity and mitochondrial network morphology in all the in vitro models used. VPA treatment inhibited mitochondrial respiration in the cell model without mitochondrial disease. However, in the models harbouring mitochondrial disease the effect was variable. Treatment under normal and energetic stress condition caused increased cell death and a reduction in cell number in all the cell models. We also evaluated a novel high-throughput imaging method and found that VPA treatment affects mitochondrial network morphology with fragmentation observed in two of cell models. It was found that not all cell models are suitable for this type of high-throughput imaging.

Conclusion: Our results suggest that the effect of VPA under normal and energetic stress condition varies according to the cell model used. The induction of energetic stress by galactose is possibly too acute for certain cell models leading to complications in data interpretation. Depending upon the model used, reducing the glucose concentration could potentially be sufficient to detect drug induced mitochondrial toxicity.

Authors: Lodge, T

• Alternative title: Drug induced mitochondrial dysfunction
• DOI: 10.5287/ora-5jna97kbq
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English