RNA synthetic biology

In this project, we will develop efficient methods to predict the effects of mutations in our DNA.
This is important, due to the ongoing national investment in research to read out the mutations people carry in their DNA. We will only be able to use that information if we know what individual mutations are doing and how they may affect health and disease.
Our methods rely on cell culture models. In such models, many thousands of cells are grown together in a flask, and we engineer every cell to contain a different mutation in a selected gene. We then ask which mutations cause the cells to die, and which parts of the cell are affected by each mutation.
We can then use this information to predict which mutations cause disease. We have teamed up with biomedical researchers in Edinburgh to study mutations that cause diabetes, eye malformations, and metabolic disorders.

As large-scale sequencing projects uncover new variation in the human genome, understanding the consequences of this variation becomes increasingly important. In the next five years, we will use synthetic biology, next-generation sequencing, and computational modelling to study the relationships between gene sequence, expression, structure, and function. We will focus on studies of noncoding RNAs and synonymous mutations in protein-coding genes. We believe that a detailed understanding of genotype-phenotype relations for selected model transcripts will uncover principles applicable to many RNA and protein molecules. Specifically, we aim:
1) To understand the molecular mechanisms underlying an RNA fitness landscape.
2) To understand the effects of codon usage on various stages of gene expression.
3) To develop new applications of genotype-phenotype mapping.
In aim 1, we will use yeast U3 snoRNA as a model system to study genotype-phenotype relations. We previously constructed a library of 60,000 mutants of U3 to study the effects of mutations on fitness in wild-type yeast. To study the role of gene-gene interactions, we will express the library in a collection of strains depleted of U3-interacting proteins, and to study gene-environment interactions, we will express the mutants in a range of environmental conditions. We will also develop high-throughput assays to measure the effects of mutations on U3 RNA abundance and RNA-protein interactions.
In aim 2, we will use human cell culture models to measure the effects of synonymous mutations on transcription, RNA stability, RNA export and translation. We will then use machine learning to uncover the sequence determinants of expression, and to understand which stages of expression are most strongly influenced by mutations. In collaboration with partners in industry (ThermoFisher) and academia (Laurence Hurst, University of Bath), we will then use our data to develop new strategies for codon optimization.
In aim 3, we will collaborate with University of Edinburgh researchers to apply our genotype-phenotype mapping methods to understand human disease mutations. We will focus on mutations in genes relevant to the local research community, encoding a range of transcription factors, hormone receptors, and metabolic enzymes.