Using and improving whole-cell models to investigate bacterial minimal genomes

Abstract

We are approaching meaningful incorporation of in silico whole-cell modelling approaches and in vivo custom genome design. The whole-cell model of Mycoplasma genitalium, published in 2012, has permitted investigation into the organism’s metabolism and genome that has been impossible in vivo. The ongoing development of a whole-cell model for Escherichia coli, a well-annotated model organism, has presented us with the opportunity to test in silico predictions from a whole-cell model in vivo for the first time.
In this thesis, I investigate in silico predictions made by both the M. genitalium and the E. coli whole-cell models. Through gene ontology term analysis of eight minimal gene sets found in the literature, I identified the biological functions that M. genitalium cells can dispense with in silico to produce dividing cells. My results clarify which biological functions should be included in any minimal gene set for M. genitalium. I investigated the results of 1214 single gene knockouts in the E. coli whole-cell model and compared these results with the assessment of essentiality in the Keio collection. At least 68.8% of knockouts agreed with the Keio collection over two generations, indicating the E. coli whole-cell model can predict gene essentiality with some accuracy. I also identify ways to develop the model to improve predictions. I prepared to use the no-SCAR system of genomic engineering in our laboratories to perform a proof-of-concept deletion in E. coli and enable in vivo investigation of E. coli whole-cell model predictions. I then worked with the Covert Lab, Stanford, to continue to develop the E. coli whole-cell model, curating data on the thiosulfate sulfurtransferase reaction and Lrp transcriptional regulation from the literature.
Despite interruptions caused by the Covid-19 pandemic, the work reported in this thesis represents progress towards the clarification and application of whole-cell models to investigate bacterial genomes.

Date of Award	28 Sept 2021
Original language	English
Awarding Institution	University of Bristol
Supervisor	Claire S Grierson (Supervisor) & Lucia Marucci (Supervisor)