Small molecules to modulate riboswitch activity in vivo and in vitro

The processes of transcription (RNA synthesis) and translation (protein synthesis) are two of the key elements of cellular information transfer. Transcription and translation are normally closely coupled and regulated to ensure coordination of cellular metabolism. The cell must also be able to respond to changes in its environment (eg the availability of carbon or nutrient sources) to give it the opportunity to adapt efficiently to its new surroundings. Until recently, such alterations were assumed to be achieved by protein factors that sensed the changes and then bound to the DNA, thus preventing transcription, or to the RNA preventing translation. Within the last few years, an alternative mechanism has been found in bacteria, simple single-celled organisms with their DNA free in the cytoplasm of the cell. Several nutrients, including vitamin B1 (thiamine), vitamin B12 (cobalamin) and riboflavin, bind directly to a short region of the beginning of the RNA (called a riboswitch), causing it to take on a specific structure that affects gene expression. Researchers have started to find riboswitch sequences in genomes from a range of bacteria and from eukaryotes (more complex organisms, including plants and animals, that have their DNA contained in a nucleus). In some cases, the riboswitches have been shown to bind the nutrient, but there is still a lot we do not know about how riboswitches work in the cell. In this project, we propose to identify small chemical molecules that interfere with riboswitches, and then use these as tools to probe riboswitch organisation and function. To do this, we will generate a reporter system in which the gene for an enzyme called luciferase is under the control of a riboswitch responsive to thiamine. Luciferase activity can be monitored by the fact that it gives off light. We will test thousands of different chemicals on the expression of luciferase in the presence and absence of thiamine. Any positives from this screen will be studied for their binding characteristics to the riboswitch RNA, to determine which stage of gene expression they affect, and to determine if the chemicals bind to different riboswitches. We will carry out further chemical modifications to the chemicals to see if this changes the characteristics. In addition, we will investigate the number of riboswitches in the RNA from a cell by passing the RNA over beads onto which one of these chemicals has been attached. Only those RNA molecules that can bind to the chemical will stick, and the rest will not. We can then wash off the bound RNA, and compare the pattern of this RNA to that of the total RNA. This will allow us to determine if riboswitches are found in several genes, or just those related to synthesis of the nutrient. It will also allow the study of riboswitches in eukaryotes, where our knowledge is currently very limited. Ultimately these chemicals will provide tools for us to control gene expression very precisely in a time- and concentration-dependent manner. This will be of great benefit in the investigation of many important biological questions, such as the development of multicellular animals and plants, and in the manipulation of metabolism to produce vitamins, drugs and other complex molecules by fermentation.

Riboswitches are recently discovered sequences in the untranslated regions of mRNA, to which metabolites bind, causing alterations in secondary structure and thereby regulating gene expression directly without the involvement of proteins. Our understanding of riboswitch function is at a relatively early stage, not least because: i. in many cases the presence of riboswitches has been inferred from bioinformatics, followed by in vitro studies of ligand binding, but their role in gene expression has not been demonstrated directly ii. those studies that have been carried out in vivo have concentrated almost entirely on natural ligands iii. although riboswitches are beginning to be identified in eukaryotes, virtually all the studies have been characterised in prokaryotes. Given the more complex genetic machinery of the former, it is likely that riboswitch function will be similarly more intricate. The involvement of small molecules in riboswitches make them a natural target for chemical intervention. We propose to identify novel riboligands (ie compounds that bind to riboswitches), and use these to probe riboswitch activity. Ultimately, riboligands that can be applied exogenously will provide the means to manipulate gene expression in a time- and concentration-dependent manner. The project is organised into six interrelated workplans, each of which has substantial sections that can be carried out independently of the other workplans. We will focus on two well characterised riboswitches from E. coli, the thiC and btuB riboswitches responsive to thiamine and adenosylcobalamin (coenzyme B12) respectively. We will generate reporter constructs with the luciferase gene under the control of these riboswitches, and test the effects of the small molecules for their ability to mimic the natural riboligand (agonists) or interfere with its action (antagonists) by measuring luciferase activity. Two constructs will be used such that they provide complementary outputs, and also provide the means to eliminate general inhibitors of gene expression. A high-throughput screen will be carried out in whole cells, and a more limited screen in a cell-free coupled transcription-translation system. We will use libraries of small molecules available commercially, supplemented with compounds related to thiamine, cobalamin and purines from our own collections. We will optimise the potency of riboligands identified in the screens through iterative rounds of chemical synthesis, and determine the basis of their specificity using biophysical techniques. Riboligands will be categorised into those that are generic, those specific to a class of riboswitches, and those specific within a class of riboswitches. This will give us a level of control beyond that exerted by Nature. Armed with selective riboswitches, we will probe further levels of control, e.g. the effect of the position of the riboswitch relative to the coding sequence. We will use our chemical approach to determine the extent of the transcriptome modulated by specific riboligands. Selected riboligands will be immobilised, and used in affinity chromatography to isolate all RNA species that bind to it. After elution, the profile will compared to that of total RNA, and RNA that did not bind, using microarray analysis with spotted microarrays of E. coli, yeast and Arabidopsis genes. This approach will enable the identification of novel riboswitches, which by definition cannot be done by bioinformatics. This is a multidisciplinary project that clearly applies chemical methodology to a specific and newly discovered biological system. We believe that the tools we develop will help define this new biology and assist in many other life science investigations. We will test the potential exploitation of our novel riboligands in several areas, including regulation of biosynthetic pathways, modification of developmental pathways and interference with viral replication.