Dissecting quantitative, analogue, antisense-mediated transcriptional control

One of the most fundamental questions in molecular biology is how quantitative gene expression is achieved. Traditionally, such regulation is ascribed to sequence specific transcription factors that bind to regulatory DNA elements. According to the concentration of the transcription factors, gene expression can then be quantitatively up or down regulated. While in many systems, such regulation undoubtedly occurs, it has become abundantly clear in recent years that this paradigm is fundamentally incomplete. This is especially so in eukaryotes where transcription has to occur in the context of chromatin. There is now substantial data showing that quantitative regulation of transcription comes from modulation of the local chromatin environment of a gene. One possibility is that alteration of the chromatin environment could permit so-called kinetic regulation by, for example, altering the ability of RNA polymerase to progress through a gene.

A further important feature of eukaryotic genes is non-coding transcription, often in an anti-sense direction. Whole genome studies have clearly shown that such transcription is pervasive, yet its role in regulating gene expression is obscure and hotly debated. Moreover, answering such mechanistic questions is difficult at the level of whole genomes. It is therefore imperative to focus on specific target loci, thoroughly dissect their mechanisms of regulation, and then leverage this knowledge to properly interpret whole genome datasets. We propose to implement this plan of action at a plant gene called FLC.

FLC functions as a repressor of the transition to flowering, and as such is a gene whose expression is under tight, quantitative control. Our preliminary data indicates that FLC expression in warm temperature conditions is controlled in an analogue fashion, like a molecular dimmer switch. This is in contrast to its behaviour after exposure to prolonged cold where expression at individual loci is switched off permanently in an all or nothing digital fashion. Our goal here is to understand how analogue control is quantitatively achieved through a mechanism that appears to function through non-coding antisense transcription. Answering this question at a deep level will require a fusion of advanced multidisciplinary techniques from molecular biology and imaging to mathematical modelling, but will allow us to get to the mechanistic heart of quantitative transcriptional control.

How gene expression can be controlled by non-coding antisense RNA is still poorly understood. We propose to elucidate such control by an in depth study at the Arabidopsis gene FLC using both experimental single cell assays and mathematical modelling. This combination of expertise will put us in a stronger position to tease apart the underlying control processes. FLC is an ideal target for quantitative investigation into multicellular eukaryotic gene expression due to a wealth of knowledge about the component parts that regulate its expression. Moreover, in plant roots, regular files of growing and dividing cells occur, allowing us to more easily probe the effects of cell division and of the cell cycle on FLC transcription control. Information gained at FLC can then be used to properly interpret data coming from whole genome experiments, which provide big data but without necessarily having the focus to resolve underlying mechanisms.

At FLC, our preliminary data supports an analogue model of gene regulation for FLC in warm environmental conditions, where expression can be dialled up or down at a single locus. This is in contrast to a digital mode of repression that occurs at FLC after prolonged cold exposure, with FLC fixed into either an on or off expression state at a locus. Our hypothesis is that the analogue mode of regulation is controlled via a kinetic coupling between antisense polymerase elongation and polyadenylation. We will explore this scenario first by assaying FLC dynamics in single cells via confocal microscopy and smFISH. We will then fit this data to a mechanistically unbiased model of sense transcription. Potential kinetic coupling will then be probed through intronic and antisense smFISH measurements, as well as by assaying the local chromatin state. Finally, we will fuse this data into a fully mechanistic, predictive model of antisense-mediated FLC transcriptional control.

The primary impact of this research will be academic. However, the fundamental nature of the work, together with the wide conservation of the components involved, will ensure that our work will have implications all across transcriptional control, ranging from biotechnology and synthetic biology all the way to human health. For example, our work will expand the toolbox for regulated gene expression and will therefore facilitate the construction of genetic circuits with precise functionality. Such an ability would be a powerful tool for synthetic biology in many biotechnological applications. Accordingly, we will constantly monitor our results for commercially exploitable output. Nevertheless, the time horizon for translating our results into directly exploitable commercial outcomes may be rather long. We will also use the project to enhance the status of interdisciplinary collaboration, including mathematical modelling, emphasising its importance in uncovering fundamental mechanism to undergraduate and graduate physical science students. Such modelling approaches are still hugely under-utilised in biology as compared to bioinformatics methods. We will also continue our outreach activities in schools by delivering an in-class teaching aid illustrating quantitative gene expression. This will be developed together with the Norfolk Teacher Scientist Network (TSN), which specialises in one-to-one teacher-scientist partnerships. Lastly, we will also develop, and then update, on-line teaching materials relating to our research findings describing the mechanistic basis of antisense-mediated transcriptional control. This is an exciting area of research, as well as being of fundamental importance. Furthermore, the areas of non-coding RNA and chromatin regulation are likely to constitute larger and larger fractions in undergraduate courses as the years go by. We therefore believe that such teaching material will be very widely used.