From bacteria to cancer exploiting metabolic networks for industrial and therapeutic purposes

Tomi Andrino, Claudio (2023) From bacteria to cancer exploiting metabolic networks for industrial and therapeutic purposes. PhD thesis, University of Nottingham.

[img] PDF (From bacteria to cancer exploiting metabolic networks for industrial and therapeutic purposes) (Thesis - as examined) - Repository staff only - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Available under Licence Creative Commons Attribution.
Download (15MB)


Metabolic modelling has been used to facilitate industrial microbiology projects and identify potential drug targets for the last 15 years. Although originally modest, the reliability of simulations has greatly improved by the progressive inclusion of chemical, biological and medical information in the models. While the multidisciplinary nature of systems biology entails rich and novel perspectives to solve problems, it is accompanied by a general lack of standardisation and consensus on methodology. Therefore, establishing robust workflows requires (i) assessing and comparing the predictive capabilities of current approaches against experimental data, and (ii) combining the strong points of different strategies to minimize their limitations and generate useful data in a timely manner.

In this study, metabolic networks for a bacterium, a parasite and a type of cancer were inspected by means of flux balance analysis and network analysis. Firstly, a readily available toolbox integrating thermodynamics information into flux balance analysis was modified to increase the number of physicochemical parameters including more suitable equations to physiological conditions. Thermodynamics constraints are closely related to the quantitation of metabolites in the cell. However, the lack of a “one size fits all” extraction and quantitation method prevents from performing a systems-wide analysis, forcing the researcher to select ca. 50 compounds for a targeted method. Exploring the topology of the network allowed identifying a set of important metabolites with a high constraining power, thus providing a rationale to such selection.

Secondly, metabolic modelling was used to facilitate drug repurposing efforts. There is a plethora of neglected diseases or conditions yet to be efficiently treated, so screening readily available drugs for novel uses has been deemed a cheaper and faster option than developing drugs from scratch. Therefore, systems biology has been exploited to reduce the list of potential candidates to be tested, as well as to identify novel potential drug targets. Specifically, a novel topological feature was introduced and combined with multiple metabolic states predicted by flux balance analysis in a model for the parasite causing sleeping sickness. Comparing against the literature validated the predictive capabilities of such approach and identified an antiviral whose potential lethality was tested in vitro.

Thirdly, a more information-rich workflow was developed based on the previous results. Transcriptomics data for a brain tumour was exploited to generate contextualised human metabolic models. Network analyses pinpointed important reactions and topological features of interest, greatly reducing the list of potential drug targets to be considered. Consequently, an automated search of chemical gene interactions and published results allowed identifying five compounds that had proven anti-proliferative and anti invasion effects on other cancer types. Finally, in vitro testing on patient-derived cancer cell lines showed their potential for further studies and generated new research questions.

This study provides considerations and approaches to increase the reliability of metabolic modelling predictions, as well as a novel workflow to identify relevant potential drug targets yet to be explored and prioritise chemicals to assess their suitability as drugs in a timely manner.

Item Type: Thesis (University of Nottingham only) (PhD)
Supervisors: Kim, Dong Hyun
King, John
Winzer, Klaus
Keywords: metabolic modelling, metabolic networks, drug targeting
Subjects: R Medicine > RM Therapeutics. Pharmacology
Faculties/Schools: UK Campuses > Faculty of Science > School of Pharmacy
Item ID: 72462
Depositing User: Tomi Andrino, Claudio
Date Deposited: 31 Aug 2023 08:57
Last Modified: 31 Aug 2023 08:57

Actions (Archive Staff Only)

Edit View Edit View