Studying stochasticity in eukaryotic gene expression using novel tools of synthetic biology modelling and analytical science

Lead Research Organisation: University of Warwick
Department Name: School of Life Sciences


The information contained in the genes of living cells has to be converted into cellular components that form structures and enable biochemical reactions to take place. This process is called gene expression and it is vital to all known life. Gene expression comprises two main synthetic steps, called transcription and translation. In transcription, the information in the DNA sequences of the genes is converted into equivalent sequences in so-called messenger RNA (mRNA) molecules. In translation, the mRNA molecules are 'read' by a large molecular structure called the ribosome, which uses the information to dictate the synthesis of proteins. The proteins synthesised in this way have to fold into a number of types of three-dimensional structure in order to become active. However, processes like gene expression and signalling in biology are not entirely accurate at the molecular level. Indeed, we now realise that one of the wonders of cellular systems is how they cope with the fluctuations in the functioning of their molecular processes. Indeed, these fluctuations actually play important roles in regulation, stress responses and other processes in biology. Our project will study how the fluctuations that occur at different steps in gene expression affect the performance of the system. To do this, we will use a gene expression pathway that we have partly built ourselves using synthetic biology techniques. This pathway has been made to interact in only a very limited way with other cellular processes so that when we manipulate it, this has no adverse effects on the cells and the responses are limited to the pathway itself. We will examine how each of the steps of gene expression outlined above contribute to the fluctuations in the gene expression pathway, and also how these respective contributions add up overall. We will do this using novel analytical methods and imaging. Microfluidics will be used to separate individual cells so that we can study changes in gene expression in each cell over time. In parallel to the experimental work, we will use computational modelling to help us understand how the system works and to help us design experiments (to test our models). Finally, we will investigate whether the fluctuations in the functioning of the system can be suppressed using synthetic genetic circuits that are known to be commonly used in natural biological systems. In this way, we will learn how noise at the molecular level can be prevented from becoming damaging to the cell. The theoretical and experimental technologies developed and applied in this project will be of value to a wide spectrum of researchers in academia and industry.

Technical Summary

Our project will study how the multiple steps in a model signalling and expression pathway determine the its noise characteristics (and thus accuracy). We propose to combine tools of synthetic biology, systems bology, analytical science (microfluidics) and computational modelling to analyse the contributions to noise generation of the respective steps in the pathway. We have designed and constructed a modified version of the yeast pheromone signalling pathway in which the peptide recognition step and transcriptional activation step have both been rendered orthogonal. This means that a non-wild-type peptide can be used to switch on a pathway that induces expression of reporter genes selected by the experimentalist, leaving the various downstream genes in the wild-type pathway uninduced. We will examine the contributions to noise generation of signalling and kinase cascade components, transcriptional activation (via a modified Ste12 protein) and translation, as well as of the degradation of mRNA and protein. In parallel to the experimental work, we will use computational modelling to help us understand how the system works and to help us design experiments (to test our models). Heterogeneity across cell populations will be studied using flow cytometry to assess multiple cells and using microfluidics to enable us to analyse expression in single cells over time. Finally, we will investigate whether the stochasticity can be controlled (and how) using Feed-Forward-Loop circuits that are commonly observed in natural regulatory systems. The theoretical and experimental technologies developed and applied in this project will be of value to a wide spectrum of researchers in academia and industry.

Planned Impact

Who will benefit from this research? This research will benefit the fundamental research community, those researchers working on biomedical problems, and the biotech industry. How will they benefit from this research? Scientists working on the mechanisms underpinning gene expression in eukaryotic systems will benefit in the short term because the results will advance our understanding of how eukaryotic cells regulate gene expression. There will be mid-term benefit for researchers wishing to apply these results to biomedical and biotechnological problems. An enhanced (quantitative) understanding of the influence of noise means that there will be an improved understanding of the sources of expression heterogeneity in cell populations. Moreover, this work will also provide an improved basis for maximising protein yields in the biotech industry. There will also be short- and mid-term benefits in a broader sense since the technologies and strategies developed in this project will be applicable to a wide range of research problems and will find numerous applications of potential medical and commercial interest. More detailed consideration of two key benefits 1. The quantitative data generated by this project will contribute strongly to the development of a digital model of a eukaryotic organism (with direct relevance to man). By inputting the data into a new systems models of yeast gene expression, we will be able to advance considerably the level of understanding of these systems. Accurate quantitative systems models are critical to our understanding of the molecular basis of disease and to drug-development strategies. 2. This quantitative approach will be of direct benefit to biotech applications involving protein production from fungal or mammalian cells. What are the realistic timescales for the benefits to be realised? The technologies being developed and applied and the data generated using them could start to be transferred to others from 12 months into the project. What will be done to ensure that they have the opportunity to benefit from this research? We will make the results widely known through publications, talks at national and international conferences and via our website. Our reagents will be made available to the wider research community. We will be open to collaborations with other parties and will promote take-up in both the academic and commercial sectors through our frequent workshops, both those organised at the MIB and those organised by the McCarthy group at the international level. Beyond this, we will transfer knowledge and skills related to this research to younger scientists through our frequent advanced training courses (funded by FEBS and industry). The institute in which we work is highly interdisciplinary and very well networked, which means that the communication of such results is both effective and rapid. What research and professional skills will staff working on the project develop which they could apply in all employment sectors? Staff working on this project will acquire important skills in cutting edge technology that can be applied widely in academic research and in industry. They will also learn how to integrate this information into a wider picture. What will be done to ensure that they benefit from this research? They will be fully engaged with all collaborations and receive all appropriate training, and will make presentations on it at national/international conferences. Their CVs will be strengthened thrugh the training in research and technology development they receive, they will be named on publications, and they will be actively involved in discussions with potential industrial partners. Engagement of users and beneficiaries. This will be achieved as outlined above, i.e. via diverse forms of communication and outreach.


10 25 50
Description Our earlier study on in vivo rate control in the yeast translation machinery (Firczuk et al., 2013) identified eIF4G as a translation initiation factor that exerts strong control over the rate of global protein synthesis. However, overproduction of this factor in yeast suppresses global translation to levels below the physiological maximum, so that a plot of the rate of translation vs the abundance of eIF4G assumes the shape of an asymmetric, approximately bell-shaped, curve. By engineering changes in the intracellular abundance of eIF4G and measuring the intrinsic noise of expression of this factor, we have been able to show that eIF4G manifests minimum stochasticity at the mean physiological abundance value (Meng et al., Nucl. Acids Res. (2016) doi: 10.1093/nar/gkw1194); forced deviations from this value result in amplified noise. It therefore seems that rate control and noise are interdependent and have co-evolved to share an optimal physiological abundance point. This principle could have broad significance in terms of the evolution of complex biological systems.

In further work (Dacheux et al, to be submitted shortly), we have discovered that stochastic ribosomal scanning events caused by structural elements in the 5'UTR (stem-loop, poly(G) sequence, upstream ORF) can serve as an independent source of additional noise (in parallel to transcriptional noise).
Exploitation Route Our findings will influence the design of gene expression constructs in academic science and biotechnology. They also inform our understanding of the sources of noise in living cells and how these influence evolutionary processes.
Sectors Manufacturing

including Industrial Biotechology

Pharmaceuticals and Medical Biotechnology

Description Our findings on noise are now contributing to our strategies for gene expression optimisation in a project involving Ingenza on the combinatorial optimisation of the synthesis of cellulose-degrading enzymes in yeast.
First Year Of Impact 2015
Sector Manufacturing, including Industrial Biotechology,Pharmaceuticals and Medical Biotechnology
Impact Types Economic

Title Noise reporters for use in yeast. 
Description We created novel reporter constructs for the measurement of noise parameters in yeast. 
Type Of Material Biological samples 
Provided To Others? No  
Impact This will come once both of the papers reporting this work have been published. 
Description Lecture to Withington School 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Schools
Results and Impact Lecture to school students about synthetic biology.
Year(s) Of Engagement Activity 2014