Team:Austin UTexas/Description

Project Description

Description and Inspiration

A major issue around the world is the presence of unclean and unsafe drinking water due to contamination from various types of bacteria. Around 780,000 deaths on average are reported each year as a result of this. Current techniques of detecting bacteria in water, such as the EPA method, the PCR method, and the culture and colony count, can be inefficient and costly, especially in developing countries. We are working to create a better alternative by developing an onsite detection method called Phast Phage, using a T7 bacteriophage to sense bacterial contamination in water sources. Our mission is to make E. coli detection in water timely, user-friendly and portable.

Figure 1. Process of genetically engineering T7 phage to release a GFP signal. Figure made by team member Sophia Xu.

The bacteriophage that we are using is the T7 bacteriophage, which infects E. coli and is one of the most well characterized phages. Using the Pinetree simulation technology developed by the Wilke Lab, we hope to be able to accurately model expression of genes within T7 model and predict how different mutations affect protein production. Additionally, we will model GFP production and lysis time of the phage to ultimately optimize the genome to express the reporter protein. Our testing in Pinetree has involved inserting GFP into a location of the T7 genome that has high expression while decreasing lysis time to allow for faster detection. Furthermore, we hope that this project will also shed more light on future engineering projects with T7 and T7-like phages. The organization of the T7 phage genome is pictured below.

Figure 2. This diagram depicts the genetic organization of the T7 bacteriophage. Genes are transcribed and expressed from left (first) to right (last). The top horizontal line numbers the specific genes that have currently known functions. The lower horizontal line is a scale of the genome measured in kilobases. The RNAse III sites are shown on the bottom line for clarity. Derived from Molineux, 2006. Figure made by team member Harsh Madaik.


Due to COVID-19, we were unable to access a lab over the summer, so we modeled our project using a stochastic gene simulator called Pinetree developed by the Wilke lab at UT Austin. Using this simulator, we want to first determine the expression levels of GFP inserted into different locations of the T7 genome as well and its effects on burst size and lysis time. This can reveal the most optimal places to insert GFP as well as potentially other foreign genes in general. We also want to decrease lysis time of the phage, which we will achieve by moving lysis genes to earlier parts of the T7 genome so that there is an earlier accumulation of the lysis proteins, holin and lysozyme. Ideally we want the lysis genes to be shortly after the GFP gene in the genome so that the lysis genes will immediately start being produced after GFP has been produced. We will be measuring lysis time using a lysis time calculator we developed which calculates lysis time based on the amount of lysis proteins present when the T7 WT model was run in Pinetree at standard lysis time (11 minutes in reality, 1300 seconds in the model).

Figure 3. To create the input file, Breseq is used to apply mutations to a T7 genome GenBank file. The resulting mutated genome file is input into Pinetree. Pinetree simulates the T7 phage infection cycle of E. coli to see how the mutations to the genome affect the phage's ability to express various proteins, produce more phage progeny, and lyse. This data is later analyzed, visualized and compared against the data from the T7 wildtype simulations.


Heineman, R. H., Molineux, I. J., and Bull, J. J. (2005) ‘EvolutionaryRobustness of an Optimal Phenotype: Re-Evolution of Lysis ina Bacteriophage Deleted for Its Lysin Gene’,Journal of MolecularEvolution, 61: 181–91

Molineux, I.J. 2006. The T7 Group, pp. 277-301. The Bacteriophages. Oxford University Press, New York.