The molecular mechanisms underlying tissue regeneration and wound healing are still poorly understood despite their importance. In this paper we develop a bioinformatics approach, combining biology and network theory to drive experiments for better understanding the genetic underpinnings of wound healing mechanisms and for selecting potential drug targets. We start by selecting literature-relevant genes in murine wound healing, and inferring from them a Protein-Protein Interaction (PPI) network. Then, we analyze the network to rank wound healing-related genes according to their topological properties. Lastly, we perform a procedure for in-silico simulation of a treatment action in a biological pathway. The findings obtained by applying the developed pipeline, including gene expression analysis, confirms how a network-based bioinformatics method is able to prioritize candidate genes for in vitro analysis, thus speeding up the understanding of molecular mechanisms and supporting the discovery of potential drug targets.

Parkinson's disease (PD) is an incurable neurodegenerative motor disorder caused by the inexorable loss of dopamine neurones from the substantia nigra pars compacta. Cell loss is characterised by the perturbation of multiple physiological processes (including mitochondrial function, autophagy and dopamine homeostasis) and much of this pathophysiology can be reproduced in vitro using the mitochondrial toxin MPP+ (1-methyl-4-phenylpyridinium). It was hypothesised that MPP+ toxicity could be modelled using protein-protein interaction networks (PPIN) in order to better understand the interplay of systems-level processes that result in eventual cell death in MPP+ models and PD. Initially, MPP+ toxicity was characterised in the human, dopamine-producing cell line BE(2)-M17 and it was confirmed that the neurotoxin resulted in time and dose dependent apoptosis. A radio-label pulse-chase assay was developed and demonstrated that MPP+ induced decreased autophagic flux preceded cell death. Autophagic dysfunction was consistent with lysosome deacidification due to cellular ATP depletion. Pertinent PPINs were sampled from publically available data using a seedlist of proteins with validated roles in MPP+ toxicity. These PPINs were subjected to a series of analyses to identify potential therapeutic targets. Two topological methods based on betweenness centrality were used to identify target proteins predicted to be critical for the crosstalk between mitochondrial dysfunction and autophagy in the context of MPP+ toxicity. Combined knockdown of a subset of target proteins potentiated MPP+ toxicity and the combined resulted in cellular rescue. Neither of these effects was observed following single knockdown/overexpression confirming the need for multiple interventions. Cellular rescue occurred via an autophagic mechanism; prominent autophagosomes were formed and it was hypothesised that these structures allowed for the sequestration of damaged proteins. This thesis demonstrates the value of PPINs as a model for Parkinson's disease, from network creation through target identification to phenotypic benefit.

No description available.